• Overview: DM rates market rallied this week driven mostly by central banks’ rhetoric suggesting patience on easing but clearly endorsing that easing will come in 2024, with only modest emphasis on the recent inflation upward surprises. Fed Chair Powell’s Congressional testimony repeated the remarks of the January post-FOMC press conference making it appropriate to start dialing back policy tightening sometime this year. ECB meeting was dovish in the forecasts with core inflation projections at 2.1% in 2025 and 2.0% in 2026, although the press conference was more balanced. President Lagarde stressed again that disinflation is a process and that more time is required to gain enough confidence, suggesting a clear and strong bias for June as the start of the easing cycle. Post ECB rhetoric on Friday endorsed broadly the message of June as date for the first cut, with hints that the cumulative easing will be larger than currently in the price, in line with our view that trough priced in are currently underestimating this upcoming easing cycle.
• Euro: Hold long 5Y Germany and 10s/15s German steepener. Keep a light intra-EMU portfolio with aim to re-scale into OW on any widening. Hold OW 10Y EU vs. Germany as medium-term trade. Keep 10s/30s France steepener vs. Germany. Keep bullish duration exposure via received in 1Yx1Y€STR, Jun24 1Y midcurve call ladder, and 1Yx5Y receiver spread vs short in OTM payer. As tactical bearish hedges, keep 1x2 Apr24 put spread, 3s/10s conditional bear flattener, and 2s/10s A/A-20 YCSO CMS floor spread (1x2). Keep steepening bias at the long end via 10s/30s conditional bull steepener and 2Yx2Y/10Yx10Y outright steepener. Take profit in paying 2Yx1Y €STR/SOFR basis. We discuss implications on FRA/OIS for potential changes from ECB’s operational framework review and decline in excess liquidity. Keep Jun24 Schatz €STR swap spread widener on valuations and as cheap risk off hedge, with further support from the ECB revision of the operational framework. Tactically neutral on Bund swap spreads with a bias to fade any rewidening but cautious on valuations and positioning. Roll the 10Yx20Y German forward swap spread widener from Mar24 to Jun24 contracts. Stay short vol via unhedged 6Mx10Y straddles; hold 1Yx1Y/1Yx10Y vol curve flattener across tails.
• UK: Hold tactical Aug24/Nov24 MPC OIS curve flatteners but close H4/U4 3M SONIA curve flatteners for a loss. Shift to a flattening bias on 1Y/1Yx1Y SONIA. Hold 5Yx5Y/15Yx15Y SONIA curve flatteners. Stay short the belly of the cash-and-duration neutral Jan32/Jan38/Oct41 gilt fly and enter tactical 10s/20s gilt curve steepeners.
• US: Data risks appear dual-sided, valuations are fair, and positioning is cleaner: stay neutral duration and look to trade extremes within recent ranges. Hold 2s/5s flatteners and 5s/30s steepeners as intermediates should outperform along the curve during the Fed on hold period and as the belly remains cheap. Despite recent Fedspeak on balance sheet composition, we do not think any changes are imminent. We review auction cyclical strategies and last month’s stripping activity. Front-end breakevens appear somewhat cheap, carry is favorable and the breakeven curve could flatten if CPI surprises to the upside: add longs in 1Y inflation swaps as a hedge to our long-end breakeven curve steepener exposure. Policy uncertainty remains elevated and we continue to favor well-hedged curve trades that isolate relative value and carry: initiate 3M fwd 3s/20s steepeners vs 15Mx3M/3Mx10Y flatteners. Lower yields and a WN CTD shift has made the Nov’53 basis appear wide: Sell the Nov ’53 WNM4 basis and buy a weighted amount of A+20bp payer swaptions. 10Y spreads appear narrow: Initiate 10Y spread wideners and weighted 5s/10s spread curve steepeners. Turn neutral on vol except in 30Y tails: Initiate shorts in 3Yx30Y swaptions.
• Japan: Swap spread continued to narrow across the curve. We explore how to interpret OIS pricing and our thoughts on swap spreads. Keep 10Y swap spread narrowers.
• Australia & New Zealand: Stay received August RBNZ OIS. Hold the received position in AUD 3s/5s/10s IRS. Enter a long in TCV Nov-2034 on asset swap.

Wake me up before spring ends

• DM rates market rallied this week mostly driven by central banks’ rhetoric suggesting patience on easing but clearly endorsing that easing will come in 2024, with only modest emphasis on the recent inflation upward surprises. Fed Chair Powell’s Congressional testimony repeated the remarks of the January post-FOMC press conference making it appropriate to start dialing back policy tightening sometime this year. ECB meeting was dovish in the forecasts with core inflation projections at 2.1% in 2025 and 2.0% in 2026, although the press conference was more balanced. President Lagarde stressed again that disinflation is a process and that more time is required to gain enough confidence, suggesting a clear and strong bias for June as the start of the easing cycle. Post ECB rhetoric on Friday endorsed broadly the message of June as date for the first cut, with hints that the cumulative easing will be larger than currently in the price, in line with our view that trough priced in are currently underestimating this upcoming easing cycle.

DM rates market rallied this week mostly driven by central banks’ rhetoric suggesting patience on easing but clearly endorsing that easing will come in 2024, with only modest emphasis on the recent inflation upward surprises ( Figure 1). The rally took place across jurisdictions with some underperformance at the short end of the curve in GBP rates.

The total amount of easing currently priced across jurisdictions is between 50bp and 75bp (with the exceptions of RBA and RBNZ across the central banks expected to ease) less than what was priced in late December ( Figure 2).

Figure 1: DM rates markets are still not pricing an aggressive easing cycle given current level of restrictiveness…

Cumulative change in OIS rate priced in the money market curve across jurisdictions by the end of 2025; bp

Source: J.P. Morgan.

Looking at the distribution of easing, we note that the amount now expected in the first half of 2024 is quite modest as a fraction of the overall move. Both Chair Powell and President Lagarde this week stressed that patience is still needed and imminent rate cuts are quite unlikely. The distribution of easing has shifted from early 2024 to 2H24 and 1H25 broadly in line with our expectation ( Figure 3). We continue to believe that the bulk of easing will be delivered between 2H24 and 1H25 but acknowledge the risk that the cycle could continue into 2H25 if modest macro volatility and slow soft landing dynamic still persist. We also note that across jurisdictions the Euro area is the only one where about half of the cumulative easing priced in is expected to take place by the end of this year, whereas in other jurisdictions most of the easing is expected to take place in 2025 ( Figure 4).

Figure 3: The distribution of easing has shifted from early 2024 to 2H24 and 1H25 broadly in line with our expectation

Cumulative change in OIS rate priced in the money market curve for mid-2024, end-2024, mid-2025, end-2025, mid-2026 and end-2026, bp

Source: J.P. Morgan.

Figure 4: Half of the cumulative easing for ECB is expected to take place this year whereas for other central banks most of the easing is expected next year

Cumulative easing priced in by the end of 2024 expressed as fraction of cumulative easing by the end of 2026; %

Source: J.P. Morgan.

During the past week, DM markets focused on the outcome of the ECB March meeting, which did not deliver any surprises on monetary policy decisions, as ECB kept policy rates and pace of QT unchanged, although it delivered dovish inflation projections. The ECB is now more confident about the ability to take inflation back to target over the medium term, but not sufficiently confident to start dialing back on the current restrictive stance. Indeed, Lagarde reiterated at the press conference that disinflation is a process that is yet to be completed.

The discussion on policy rate cuts is still felt premature at this stage and is unlikely in our view to come on the table before the new projections are published in June.  President Lagarde clearly indicated that more data, especially on the evolution of wages and profit margin, will be needed to gather additional confidence, and not enough data will be available by the April meeting. The dovish tilt in the press conference came via lowering core inflation projections in 2025 at 2.1% from 2.3% and in 2026 at 2% from 2.1% ( Figure 5). Core inflation at the end of the forecast horizon was just below the target in March 2022 (1.9%) when the ECB was still quite dismissing about chances of early hikes and has been above the target since the projections of June 2022, when core at the end of the forecast horizon was 2.3%.

ECB staff headline inflation projections at the March 2024, December, September and June 2023 meetings; %

Source: ECB, J.P. Morgan.

ECB staff core inflation projections at the March 2024, December, September and June 2023 meetings; %

Source: ECB, J.P. Morgan.

President Lagarde reiterated the emphasis on the three criteria to assess the monetary policy stance: a) inflation outlook, b) dynamic of underlying inflation, and c) strength of monetary policy transmission ( Figure 6).

a) On the inflation outlook, the revision to the forecast are now suggesting the achievement of the inflation target over the forecast horizon with a higher degree of confidence. Inflation is expected to continue its downward trend in the coming months. Further ahead, it is expected to decline to our target as labour costs moderate and the effects of past energy shocks, supply bottlenecks and the reopening of the economy after the pandemic fade.

b) On the evolution of underlying inflation, most measures have eased further; however domestic price pressures are still elevated, in part owing to robust wage growth and falling labour productivity. At the same time, there are signs that growth in wages is starting to moderate. In addition, profits are absorbing part of the rising labour costs, which reduces the inflationary effects. Measures of longer-term inflation expectations remain broadly stable, with most standing around 2%.

c) The transmission of monetary policy has been forceful and expected to continue. Financing conditions are restrictive and past interest rate increases continue to weigh on demand, which is helping push down inflation.

Figure 6: The three criteria to assess the monetary policy stance are moving in the right direction but it is too early to have full confidence that the disinflation process is completed

Criteria for Monetary Policy decisions and respective ECB assessment at the March Monetary Policy Meeting

Source: J.P. Morgan.

While the projections with inflation back to target over the forecast horizon delivered a dovish tilt, the message from the press conference was more balanced and calibrated. The notable exception in the disinflation dynamic is mainly coming from sticky domestic inflation, which is mostly driven by services and therefore more impacted by the evolution of wages.

President Lagarde also indicated that, for the ECB to start dialing back on the insofar delivered tightening, the actual inflation prints do not have to reach 2%, but they would have to be more confident of the evolution of the inflation outlook over the medium term, while being vigilant on the evolution of wages and profit margin. An increase in nominal wages in real income is one factor behind the ECB forecast of a recovery on the growth side; however for that increase not translating into second round-effects, the ECB needs to see wage increases moderating and, more importantly, some absorption from firms on lower margins, in line with current evidence.

While not speculating on details about market expectations, President Lagarde indicated that the “market is converging better” towards expectations in line with gradual removal of restrictive monetary policy stance. We continue to believe that the amount of easing priced between now and the end of 2024 is broadly fair whereas we continue to see more value in the market underpricing our expectation of the terminal rates in this upcoming easing cycle. This is true both in our baseline scenario of policy rate neutral around 2% or a recession scenario in which we believe policy rates could move into easing territory (between 1% and 1.5%).

The ECB did not mention anything about QT suggesting that the pace there is likely to be in auto-pilot, with no further PSPP reinvestment and partial PEPP starting from July. The ECB also indicated that the technical details of the revision of the operation framework will come out next week. Our bias is for a long-term balance sheet size that is larger than pre-QE period with a structural asset portfolio and full allotment long-term refinancing operations. These operations, available to banks by the end of the year, will be likely indexed to refi (with a narrower corridor between refi and depo rate) vs. high quality collateral and with maturity longer than current 7D and 3M (1Y or longer).

The message delivered by Lagarde last Thursday was echoed by several comments from other ECB Governing Council members on Friday ( Figure 7). In particular, on the start of easing cycle, ECB members were aligned in repeating that there is increasing confidence that they are getting close to the first cut. They indicated that there is high probability they will start cutting policy rates by spring, most likely in June, but they remain open minded and data-dependent, closely monitoring the progress on inflation particularly on wages and services inflation. Few speakers mentioned that, although not likely, a first cut in April could not be ruled out ex-ante but would likely depend on a sharp data deterioration. This seems to be broadly in line with our call for a first cut in June, with April kept open just as extreme option in a tail scenario.

Figure 7: The message delivered by Lagarde last Thursday was echoed by several comments from other ECB Governing Council members on Friday. The commentary endorses our view that, while market pricing for the first cut now seems to have corrected, more cuts are likely to be priced in for 2H24 and 1H25

Selected comments from ECB Members after the March Monetary Policy Meeting Press Conference

Source: Media sources, J.P. Morgan.

On the distribution of upcoming cuts, ECB speakers stressed that they will not necessarily be forced to cut at every meeting as they need to be gradual and gather confidence on the convergence of inflation down to target. We believe that a skip between meetings is more likely to happen at the beginning of the cycle as ECB will want to gather confidence on the inflation process, but less likely thereafter. On market pricing, Governing Council Members agreed that markets are now better aligning with ECB’s views on the start of the cutting cycle, although Kazaks also mentioned that, after they will begin to ease the policy stance, ECB is likely to cut more than what currently priced. The commentary from ECB Members therefore seems to endorse the message we have been highlighting in our last publications: while market pricing for the first cut now seems to have corrected, more cuts are likely to be priced in for 2H24 and 1H25.

Japan flows: Japanese investors turned even larger net buyers of foreign bonds with $17.5bn net purchases in January with flows being in line with those observed on average in 2023

The publication of the Balance of Payments data show that Japanese investors turned even larger net buyers of foreign bonds in January ($17.5bn net purchases) relative to December 2023. In January, Japanese investors continued to be net buyers of foreign debt especially in US ($23bn net purchases) with some marginal buying flows across other jurisdictions as well. In the Euro area, Japanese investors bought €0.7bn of debt securities in January: there were net buying flows in German government bonds (€2.1bn), but these purchases were offset by modest net selling in French and peripheral govies (net sales of €0.3bn in France and €0.7bn in Euro area periphery). January net purchase flows are broadly in line with the monthly flows observed on average in 2023 ( Figure 8).

Figure 8: In January 2024, Japanese investors continued to be net buyers of foreign debt especially in US ($23bn net purchases) with some marginal buying flows across other jurisdictions as well. January net purchase flows are broadly in line with the monthly flows observed on average in 2023

Japanese net purchases of foreign bonds and notes by region; bn of domestic currency* ($bn for Total) per month

Source: J.P. Morgan.

Figure 9: DM rates rallied over the week with underperformance of GBP rates at the short end

Changes in b/m yield since 1^st March 2024 by region, bp

Source: J.P.Morgan

Figure 10: While data releases in the Euro area were light this week with final GDP in line with expectations, US data broadly surprised to the downside but non-farm payrolls were higher than forecasted

Surprises vs. consensus expectations since 1^st March 2024

Source: Bloomberg Finance L.P.
Note: Jobless claims are reported as upside (downside) surprises if they are lower (higher) than consensus expectations. Inflation expectations are reported as upside (downside) surprises if they are higher (lower) than consensus.

Figure 11: Central bank policy rate expectations inferred from money market rates

Central bank policy rate expectations (1M forward OIS rates starting at month end) and J.P. Morgan policy rates forecasts*, %

Source: J.P. Morgan.
* Adjusted for the differential between policy rates and O/N OIS rates. For USD OIS, adjusted for the differential between the O/N OIS rate and the upper bound of the Fed funds target range.
Levels as of COB 7th Mar 2024.

Figure 12: Yield pick-up available via foreign bonds by country and investor currency, on a currency-hedged basis (3M rolling and maturity-matched hedges) and unhedged

Annualised yield pick-up* for euro-, yen-, US dollar- and sterling-based investors from foreign currency bonds vs. domestic bonds (German bonds for euro-based investors) of the same maturity, with no hedge, 3M rolling**and maturity-matched*** currency hedges; shaded regions for rolling and maturity-matched hedges show >1% yield pickup (darkest shading), between 0.5% and 1%, and between 0% and 0.5% (lightest shading), %

Source: J.P. Morgan. * Yield pick-up defined as foreign currency yield + hedge cost – domestic currency yield, using par govie curves (except Sweden, where we use benchmark bonds).** Cost of 3M rolling hedge defined as 3M FX cross-currency basis + domestic 3M swaprate (3s curve) – foreign-currency swaprate (3s curve).*** Cost of maturity-matched hedge defined as the maturity-matched FX cross-currency basis + domestic swaprate (3s curve) – foreign-currency swaprate (3s curve). Levels as at COB 7^th Mar 2024.

Figure 13: Yield pick-up available via foreign bonds by country and investor currency, on a currency-hedged basis (3M rolling and maturity-matched hedges) and unhedged (EUR countries and SSA )

Annualised yield pick-up* for euro-, yen-, US dollar- and sterling-based investors from different Euro-based bonds of the same maturity, with no hedge, 3M rolling** and maturity-matched*** currency hedges; shaded regions for rolling and maturity-matched hedges show >1% yield pick-up (darkest shading), between 0.5% and 1%, and between 0% and 0.5% (lightest shading), %

Source: J.P. Morgan.
* Yield pick-up defined as foreign currency yield + hedge cost – domestic currency yield, using par govie curves (except Sweden, where we use benchmark bonds).
**Cost of 3M rolling hedge defined as 3M FX cross-currency basis + domestic 3M OIS rate – foreign currency OIS rate.
*** Cost of maturity-matched hedge defined as the maturity-matched FX cross-currency basis + domestic OIS rate – foreign-currency OIS rate.
Levels as of COB 7th Mar 2024.

Figure 14: Current Global Rates trade recommendations

Changes are in bold and underlined

Source: J.P. Morgan.

Figure 15: 2024 trade performance summary

Hit rate* and average P/L** for all trades closed in 2024; bp of yield***

Source: J.P. Morgan.
* Hit rate defined as # positive trades / (total # trades - trades closed at flat).
** Avg. P/L across total # trades. Options trades are shown in bp of notional, except for money market trades which are shown in bp of yield and not included in avg P/L calculation.
*** Previously called CDS.
Note: For individual trade performance data, please see Global Fixed Income Markets Weekly: Trade Statistics, Fabio Bassi et al.

Stay OW duration with easing in sight

• DM yields rallied sharply over the week, with German yields outperforming in 2-10Y sector on dovish ECB delivery, mixed macro data and position technicals
• Hold strategic OW duration and steepening bias via long 5Y Germany and 10s/15s German steepener, respectively
• Keep a light intra-EMU portfolio with aim to re-scale into OW on any widening; keep OW 10Y EU vs. Germany as medium-term trade
• Periphery: neutral Italy; 10s/30s Italy credit curve too flat; hold long 10Y Greece vs. Italy and short 5-7Y Portugal vs. Spain
• Core: neutral; hold 10s/30s France steepener vs. Germany and 30s/50s France steepener
• Supply: around €21bn of conventionals via auction next week in Germany, Italy, Portugal and the Netherlands

DM yields rallied sharply over the week with German yields outperforming in 2-10Y sector ( Figure 16). The Euro rate move this week was by a combination of modestly dovish ECB delivery at March meeting, mixed macro data and position technicals, anecdotal evidence also suggests investors re-entering bullish exposures.

Figure 16: DM yields rallied sharply over the week with German yields outperforming in 2-10Y sector

Cumulative change of 10Y US, UK and Germany benchmark yield since 1 March 2024; bp

Source: J.P. Morgan.

As discussed in ECB: June is on (if the data don’t get in the way), G. Fuzesi, 7 Mar, the dovishness in the ECB delivery came mainly from the March staff projections where they revised core inflation lower more than our and consensus expectations. The communication at the press conference was balanced, as expected, with President Lagarde stressing the data-dependent approach with suggestions of June as the likely starting point of easing cycle.

We have been running a strategic OW duration stance via long 5Y Germany and yesterday’s ECB delivery goes in the right direction for this. The market has repriced lower the ECB’s cumulative easing expectation to around 2.25% from close to 2.50% in late February. We still find the current ECB terminal rate pricing again consistent with a soft-landing higher/restrictive-for-longer scenario and also not incorporating risks of a recession-driven cut cycle into easy territory. Our strategic OW duration stance has been based on expectation that the terminal pricing for the ECB should be closer to neutral 2% or modestly below neutral. Also, while we maintain a medium-term strategic bullish duration stance, we have been scaling the degree of OW exposure with tactical options based hedges (see European Derivatives section).

On German curve, we still believe it is too early to position for 2s/5s bull steepening as 2s/5s tend to bull steepen as we get within 1-2 months to start of the easing cycle (see here). Further out, we maintain a strategic steepening bias on the 5s/30s (10s/30s) curves. We hold 10s/15s German steepener as a low-beta proxy of our strategic steepening bias given attractive relative valuations ( Figure 17).

Figure 17: The 10s/15s German curve is trading more than 1bp too flat relative to 10s/30s curve

Bund Aug33/May38 curve regressed against Bund Aug33/Aug53 curve; past 3M; bp

Source: J.P. Morgan.

Intra-EMU spreads

Intra-EMU spreads tightened sharply over the past week with Italy being the relative outperformer. The strong tightening this week was, in our view, driven by a combination of strong performance of risky assets globally, modestly dovish ECB and position technicals, with anecdotal evidence suggesting liquidation flows of some UW positioning. The recent tightening has pushed peripheral spreads, especially Italian spreads, to levels last seen in early 2022, before the start of the ECB tightening cycle, and core spreads also close to the lower-end of past-12M ranges, but still in the top quartile of their longer-term ranges ( Figure 18).

Figure 18: Peripheral spreads are now trading at levels seen in early 2022, before the start ECB tightening cycle

10Y core and periphery weighted spread to Germany; bp

Source: J.P. Morgan.

We neutralized our OW intra-EMU stance in late February with the aim of re-scaling into OWs on any widening move, as we still retain a strategic medium-term constructive stance. In our view, the rising OW intra-EMU positioning, although far from pre-2022 highs ( Figure 19), and ongoing macro/monetary policy uncertainty, makes intra-EMU spreads increasingly susceptible to (temporary) correction over the near term, especially given their strong performance YTD. Also after the recent tightening, intra-EMU spreads, especially Italian spreads, are screening expensive relative to other Euro credit spreads ( Figure 20).

Figure 19: European investors current OW exposure is the highest since early-2022, but still well below the OW exposure witnessed during PEPP period

Net exposure of Euro area investors to peripherals vs. core; %

Source: J.P. Morgan European Client Survey, 7 March

Figure 20: Italian spreads are now screening expensive relative to other Euro credit spreads

10Y Italy-Germany spread regressed against EUR IG spreads to Germany; past 6M; bp

Source: J.P. Morgan.

Intra-EMU trading themes

We hold OW 10Y EU vs. Germany as a medium-term strategic OW exposure. On RV, we keep long 10Y Greece vs. Italy, short 5Y Portugal vs. Spain, 30s/50s France steepener and 10s/30s France steepener vs. Germany.

In Italy, we stay neutral on spreads at current level. Italian spreads have tightened along with broader risk assets with no clear Italy-specific driver and are now screening expensive relative to other credit. However, given lack of near-term domestic catalysts for widening and still supportive attractive risk-adjusted carry on spreads, we remain cautious in entering outright widening exposures. We continue to hold long 5Y CDS cash basis (buying 5Y Italy protection vs. buying 5Y Italy ASW) as a convex risk-off hedge. The Italian CDS basis directionality with credit spreads has broken down recently as the basis moved sideways at historical tight levels, despite broad credit tightening, highlighting the convex nature of the hedge ( Figure 21).

Figure 21: Italian CDS basis directionality with credit spreads has broken down recently as the basis moved sideways at historical tight levels despite broad credit tightening, highlighting the convex nature of the hedge

5Y Italy CDS – 5Y Italy ASW spread regressed against EUR IG spreads to Germany; past 12M; bp

Source: J.P. Morgan.

On Italian credit curve, the 3s/10s box flattened in the tightening move and is tracking its long-term relationship vs. spreads ( Figure 22). Further out, the 10s/30s credit curve remain around 3bp too flat vs level of spreads ( Figure 23).

Figure 22: The 3s/10s Italy-Germany box flattened in the tightening move and is tracking its long-term relationship vs. spreads

3s/10s Italy-Germany box regressed against 10Y Italy-Germany; par rates used; past 3Y; bp

Source: J.P. Morgan.

Figure 23: The 10s/30s Italian credit curve remains excessively flat

10s/30s Italy-Germany box regressed against 10Y Italy-Germany; par rates used; past 6M; bp

Source: J.P. Morgan.

Spain still remains one of our favorite picks for OW in intra-EMU space and will be looking to re-enter OW on any underperformance. We also hold shorts in 5Y Portugal vs. Spain on stretched valuations ( Figure 24), potential political uncertainty around Portuguese elections on 10 March and relatively higher OT supply pressures in 2024 vs. 2023. Portuguese bonds underperformed vs. Spain in 5-15Y sector over the past week.

Figure 24: Portuguese bonds trading around 20-30bp through Spain are quite dear: stay short 5-7Y Portugal vs. Spain

Portuguese bonds vs. interpolated Spanish z-spread curve; bp

Source: J.P. Morgan.

Figure 25: We see decent probability of Moody’s upgrading Greece to IG at their 15 March review

Rating statistics for Greece sovereign long term local currency debt;

Source: DBRS, Fitch, Moody’s, S&P, J.P. Morgan

We also stay OW 10Y Greece vs. Italy on limited GGB supply pressures over coming weeks and also as a potential risk-off hedge. Greek bonds underperformed vs. Italy during the risk-on tightening move over the past week. Also, on 15 March, Moody’s is to review Greece rating and we see decent probability of it upgrading Greece to IG ( Figure 25). Moody’s is the only major rating agency that still has Greece rated sub-IG.

Core spreads also tightened over the week and are now trading modestly tight on our fair value framework ( Figure 26). We remain neutral on core spreads over the near term and will be looking to re-enter OW exposure on any widening move.

Figure 26: Core spreads also tightened over the week and are now trading modestly tight on our fair value framework

10Y wtd. core spread vs. Germany regressed against 10Y wtd. peripheral spread vs. Germany (X1) and 10Y Germany yield (X2); past 4Y; bp

Source: J.P. Morgan.

In France, the intermediate sectors is now trading on the cheap side again on the credit curve ( Figure 27). Further out, the 10s/30s credit curve continues to screen excessively flat ( Figure 28) and we hold 10s/30s France steepener vs. Germany (OAT Nov33/May54 steepener vs. Bund Feb33/Aug54). We also hold 30s/50s France steepener, given excessively flat valuations and as a proxy for our strategic bullish duration and short volatility view.

Figure 27: The intermediate sectors on the French credit curve are now trading cheap again

Residual from regressing 50:50 3s/5s/10s France-Germany fly against 5Y France-Germany spread and 3s/10s France-Germany box*; par rates used; past 6M; bp

Source: J.P. Morgan.* 3s/5s/10s = 0.12*5Y + 0.50*3s/10s – 25.7; R-squared: 62%.

Figure 28: The 10s/30s French credit curve is trading excessively flat vs. level of 10Y spreads: hold 10s/30s France steepener vs. Germany

10s/30s France-Germany box regressed against 10Y France-Germany; par rates used; past 6M; bp

Source: J.P. Morgan.

Issuance

We expect around €21bn of conventional bond supply via auctions next week ( Figure 29), plus potential syndicates. Conventional auctions are scheduled in Germany, Italy, Portugal and the Netherlands.

Figure 29: We expect around €21bn of conventional bond supply next week

Euro area conventional bond issuance calendar until the end of March; official announcements and J.P. Morgan forecast; peripheral supply highlighted in grey, green bonds are marked in green; €bn

Source: Debt Management Offices, J.P. Morgan

DSL Jan30: fair

The 7Y benchmark is currently trading with a benchmark discount of around 2.0bp relative to surrounding DSLs, close to past 6M average 7Y benchmark discount.

Schatz Mar26: cheap

The 2Y benchmark is trading with a benchmark discount of around 4bp relative to surrounding bonds, close to upper-end of past 12M range and quite cheap in our view.

Bund Feb34: modestly dear

The 10Y benchmark outperformed relative to the off-the-run and is now trading with a benchmark roll (z-spread pick-up over off-the-run Bund Aug33) of around 1.8bp, close to the lower-end of the past 12M benchmark roll trading range. The 10Y German benchmark-roll has shown modest negative directionality to 5s/30s German swap spread curve over long history and the Bund Feb34 benchmark roll is dear based on that relationship ( Figure 30).

Figure 30: 10Y German benchmark-roll has shown modest negative directionality to 5s/30s German swap spread curve over long history and is now trading on the expensive side

10Y German benchmark roll (z-spread pick-up over off-the-run) regressed against 5s/30s Germany swap spread curve; past 5Y; bp

Source: J.P. Morgan.

Figure 31: The 3Y BTP benchmark is trading broadly fair and 7Y BTP benchmark is trading cheap in our view

BTPs Feb27 and Feb31 relative to interpolated BTP yield curve; bp

Source: J.P. Morgan.

OT Oct31: fair

The OT Oct31 is trading broadly fair on the OT curve based on par curve fair value framework. As discussed above, we find 5-7Y OT expensive vs. Bonos and hold UW 5Y Portugal vs. Spain.

OT Apr42: cheap

The OT Apr42 is trading modestly cheap on the OT curve based on par curve fair value framework.

Italy auctions

On Wednesday, we expect Italy to auction around €8bn of BTPs: the 3Y benchmark (BTP Feb27), the 7Y benchmark (BTP Feb31), and potentially a bond in the ultralong end. The auction details are to be announced on Friday evening, 8 March. The 3Y benchmark is trading at a discount of around 2.8bp relative to surrounding bonds, broadly fair, in our view ( Figure 31). Similarly, the 7Y benchmark is also trading at a discount of around 3.5bp relative to surrounding bonds, close to upper-end of 7Y benchmark discount range over the past 12M, on the cheap side, in our view.

Trade recommendations

German trades

• Keep long 5Y Germany
  Keep long €100mn Bobl Oct28 @ 2.28%. 3M carry: -6.0bp and 3M slide: -2.5bp. P&L since inception (21 July 2023): 15.2bp.
• Keep 10s/15s Germany steepener
  Keep long €50mn Bund Feb34 vs. short €41.2mn Bund May38 @ 15.4bp; 3M carry: -1.7bp and 3M slide: 0.7bp. P&L since inception (1 March 2024): 1.1bp.

Country selection & RV trades

• Keep long 10Y Greece vs. Italy
  Keep long €10.0 GGB Jun33 vs. short €10.2 BTP Nov33 @ -31.4bp. 3M carry: -2.0bp and 3M slide: 1.2bp. P&L since inception (03 Nov 2023): -17.8bp.
• Keep short 5Y Portugal vs. Spain
  Keep short €25mn OT Jun29 vs. long €24.5mn Bono May29 @ -27.7bp; 3M carry: 1.4bp and 3M slide : -0.2bp. P&L since inception (9 Feb 2024): 5.2bp.
• Keep long 5Y Italy CDS basis
  Keep long €25mn of 5Y Italy CDS vs. long €25mn BTP Feb29 vs. maturity matched swap @ 13.0bp; 3M carry: -1.2bp. P&L since inception (19 Jan 2024): 0.4bp
• Keep 30s/50s France steepener
  Keep long €25.0 OAT May54 vs. short €30.7 OAT May72 @ -51.0bp. 3M carry: -0.3bp and 3M slide: 1.2bp. P&L since inception (02 Feb 2024): -6.0bp
• Keep 10s/30s France steepener vs. Germany
  Keep short €50mn OAT May54 vs. long €104.5mn OAT Nov33 and long €43.2mn Bund Aug54 vs. short €104.2mn Bund Feb34 @ 31.5bp ; 3M carry: 0.2bp and 3M slide : 1.0bp. P&L since inception (23 Feb 2024): -0.5bp

SSA trades

• Keep long 10Y EU vs. Germany
  Keep long €25mn EU Jul34 vs. short €25.5 Bund Feb34 @ 56.2bp. 3M carry: 1.9bp and 3M slide: 0.7bp. P&L since inception (08 Feb 2024): 9.9bp.

*Unless specified, all trades are priced as of 1PM on non-payroll Fridays and 2PM on payroll Fridays.

Trades closed over the past 12 months

We are not there yet

• The ECB reiterated its data dependence stance while opening the door for a June cut based on expected evolution of underlying macro data. We expect the first 25bp cut in June followed by three more 25bp cuts in 2024 (September, October, and December)
• Keep receiving 1Yx1Y €STR, hold longs in Jun24 1Y Euribor midcurve call ladder, and stay in 1Yx5Y receiver spread vs OTM payer
• As bearish hedges, hold Apr24 1x2 put spread (96.375/96.25), 3s/10s conditional bear-flatteners, and 2s/10s flatteners via 1x2 A/A-20 YCSO CMS floor spread (3M expiry)
• Keep 2Yx2Y/10Yx10Y swap curve steepeners and hold conditional bull steepeners via 6M receivers
• Hold Jun24 Euribor calls versus SOFR
• Take profit in paying 2Yx1Y €STR/SOFR basis which has narrowed sharply on Yankee issuances
• Bais towards further tightening in FRA/OIS and 3s/6s on expected narrowing of the refi/depo corridor and term funding operations from the operational framework review to be announced next week
• Gradual decline in excess liquidity should also move €STR fixings relatively higher but only marginally at still high levels of excess liquidity before picking pace only when excess liquidity will approach around €1tn versus current level of €3.5tn
• Schatz €STR swap spread remains cheap vs. fundamental drivers in a long-term and short-term fair value model, but less out of line vs. other measure of risk given further narrowing in intra-EMU and FRA/OIS spreads and decline in rates volatility 
• Keep Jun24 Schatz €STR swap spread widener on valuations and as cheap risk off hedge targeting between 15bp and 20bp over the next quarter, with further support from the ECB revision of the operational framework biased to drive a modest richening of funding rates
• Neutral on Bund swap spreads: easier financial conditions and technical factor such as receiving from financial institutions support further narrowing; however, valuations are on the cheap side and narrowers are still quite subscribed across investors
• Roll the 10Yx20Y German forward swap spread widener from Mar24 to Jun24 contracts, via equi-notional Buxl widener and Bund narrower on valuations and as proxy for 10s/30s swap curve steepener
• Stay short 6Mx10Y unhedged straddle and hold 1Yx1Y/1Yx10Y volatility curve flatteners across tails

The overwhelming message from this week’s ECB meeting and press conference was that while sufficient progress has been made on the disinflationary trend and the ECB expects this trend to continue, they still want further confirmation of the same over the coming months before embarking on an easing cycle.  President Lagarde implicitly ruled out a April start by stating that only “a little” more information will be available by the April meeting but a “lot more” will be available by the June meeting (three additional PMI, CPI, and more wage settlements data).  Thus, while there is confidence on starting the easing cycle soon, the ECB clearly indicated that we are still not there yet.

President Lagarde acknowledged that market pricing will broadly converging towards ECB’s stance but refused to give any detail on potential timing or the pace of rate cuts once they begin the easing cycle.  Admittedly, the ECB has now begun discussing dialing back on the level of restrictiveness.  The relative small gap until the April meeting (five weeks, 11^th April) is unlikely to provide enough macro data for the ECB to act decisively then as well.  We continue to expect first 25bp cut to begin in June as the ECB will likely have enough confidence on the undergoing disinflationary trend, assuming data follows along expected lines. 

The €STR yields declined across the curve with an outperformance of reds and greens which was primarily fuelled by the US payroll data and anecdotal evidence of re-setting of some long duration positions earlier in the week ( Figure 32).

Figure 32: €STR forward yields declined across the curve with a decent outperformance of reds and greens over the week

Current (rhs, %) and change since 1^st March (lhs, bp) for various 1Y forward €STR yields;

Source: J.P. Morgan.

Figure 33: The €STR curve is now pricing around cumulative 27bp, 46bp, and around 72bp of cuts for April, June, and September meetings

ECB meeting steps for the €STR curve (rhs, bp) and current €STR curve (lhs, %);

Source: J.P. Morgan.

The €STR curve has broadly priced out any probability of a April cut (pricing around 4bp) and is currently pricing around 27bp, 46bp, and 72bp of cumulative cuts by the next three meetings ( Figure 33).  The downside surprise in the core inflation forecast for 2025 gives us confidence of a June cut although we acknowledge that market pricing is likely to also remain data driven over the coming days.  Over the near-term we wait for the US CPI next week after yields rallied post US payrolls which showed a large upside surprise in NFP but an equally surprising rise in unemployment rate and large downward revisions to previous two months data. 

The ECB delivery this week, or the lack of thereof, has done little to change our trading strategy, as discussed in the post ECB note yesterday.  We continue with our two-prong strategy.  At the core, we have a medium-term bullish duration bias targeting the reds/greens sector of the curve.  However, we acknowledge risk of domestic and global data surprising to the upside leading to further pushback of the easing cycle and therefore hold tactical bearish hedges.

As our medium-term duration view, we continue to receive 1Yx1Y (reds) €STR that we have been recommending over the past few weeks.  Admittedly, reds €STR have overall increased over the past few weeks but we retain conviction on this declining over the coming weeks.  Strong US data has resulted in relatively hawkish Fed rhetoric which has translated to higher Euro area yields as well ( Figure 34).  We believe that the current disinflationary trend in the Euro area will continue and against a backdrop of weak growth warrants a lower terminal rate over the medium-term than is currently priced on the €STR curve (around 2.25%).  Of course, it will require a catalyst for such a move and we expect continued decline in Euro area inflation and some weakening in US macro data would trigger such a move.  Anecdotal evidence suggests that market participants also expect yields to decline and are biased towards long duration positions although outright positioning remains rather low given the recent stop-outs of long duration positions. 

Figure 34: Stay received reds €STR and other variants of our bullish duration exposures expressed via options; €STR yields have been dragged higher by global dynamics

1Yx1Y €STR yield regressed against 1Yx1Y SOFR yield; past 2M; %

Source: J.P. Morgan.

Figure 35: Keep Jun25/Jun26 conditional bull steepeners expressed by Jun24 1Y and 2Y midcurve calls

Jun25/Jun26 Euribor yield spread regressed against Jun25 Euribor yields; past 2M; bp

Source: J.P. Morgan.

 We also retain our bullish duration exposure on the Euribor curve expressed via midcurve call ladders but rotated exposure from Mar24 1Y midcurve call ladders (97.5/97.75/97.875) into Jun24 1Y midcurve call ladders (97.5/97.75/98) with Jun25 Euribor futures at 97.47 (see Trade Recommendations).  Jun25 Euribor futures have to rally in excess of 65bp by option expiry for the trade to breach the lower (yield) breakeven levels.

Last week, we had highlighted that the money market curve is exhibiting negative directionality versus yields – a dynamic that we expect to continue and had thus recommended Jun25/Jun26 Euribor conditional bull steepeners implemented via 1Y and 2Y Euribor midcurve calls ( Figure 35).  The Jun25/Jun26 Euribor curve flattened around 5bp over the week.  The curve now appears too flat versus yields on a regression basis and we continue to hold this trade.  On an intuitive level, we would expect more easing to be front-loaded which should lead to a bull-steepening dynamic.

We also keep longs in 1Yx5Y receiver spreads (ATMF/ATMF-50bp) funded via selling OTM payers (ATMF+50bp).  Underlying 1Yx5Y swap yield is broadly overall unchanged over the past month, in line with the overall P&L ( Figure 36). 

Figure 36: Stay long 1Yx5Y receiver spreads funded via shorts in OTM payer

Cumulative change in premium (lhs, bp of notional) versus change in fixed dated 1Yx5Y swap yield (rhs, inverted axis, bp); since 2^nd Feb 2024

Source: J.P. Morgan.

Figure 37: We expect Jun24 Euribor futures to sell-off a few bps further over coming days if our view of no July rate cut is realized and/or market removes further easing from the curve; hold Apr24 Euribor put spread

Fair value assessment of Jun24 Euribor futures under various ECB outcomes for the next three meetings*; %

Source: J.P. Morgan.
Current Jun24 Euribor futures at 96.39.  We assume that Euribor/€STR spread remains constant at 8.5bp, in line with current levels.

As mentioned above, we also prefer to have tactical bearish hedges in our portfolio and retain these trades that we have been recommending over the past few weeks.  First, we keep April put spread (1x2) (96.375/96.25 versus Jun24 futures at 96.39).  While the cumulative easing priced by June and September meeting is broadly in line with our expectations, we still see upside to this trade in our Baseline scenario.  This mainly comes from the fact that the Jun24 futures encompass the June/September IMM period (17^th June 2024 to 16^th September 2024).  The June IMM date is after the ECB meeting date whereas the September IMM date is before the start of the September maintenance period (MP starts from 18^th September, ECB meeting on 12^th September).  Thus, any rate cut at the September meeting is unlikely to be reflected in the Jun24 Euribor futures.  In Figure 37, we project fair value of the Jun24 Euribor futures under various ECB scenarios and see attractive risk-reward to holding Apr24 put spreads.

Second, we keep 3s/10s conditional bear-flattener that we had recommended last week.  With non-trivial risk of central banks pushing back the start of the easing cycle, conditional flatteners offer good hedges to our otherwise bullish duration view.  The swap curve has exhibited a strong negative directionality versus yield ( Figure 38).  The 3s/10s curve has flattened around 2bp with 3Y rallying around 10bp making the curve too flat versus yields.  Nevertheless, we believe that the trade offers attractive protection in a sell-off.

Figure 38: Keep 3s/10s conditional bear flattener

3s/10s swap curve regressed against 3Y swap yield; past 3M; bp

Source: J.P. Morgan.

Figure 39: Keep 3Mx(2s/10s) YCSO A/A-20bp 1x2 CMS floor spread as a positive carry bearish hedge

3Mx(2s/10s) curve spread; since 1 Jan 2022; bp

Source: J.P. Morgan.

Third, with the ECB expected to remain on hold for the next three months, we continue to hold attractive carry structures that offer some protection in a bearish move.  To this effect, we hold 3Mx(2s/10s) A/A-20bp 1x2 CMS floor spread that we had recommended two weeks ago.  Flatteners benefit from attractive carry; 3M carry on 2s/10s curve is around 15bp.  The CMS curve has flattened around 6bp since inception and the speed of this flattening has resulted in only small overall P&L in the trade.  The lower breakeven for the trade is around -70bp which we ascribe low probability to be breached.  It will require significant repricing of policy rates higher for the curve to flatten to below -70bp over the coming few weeks ( Figure 39). 

At the long-end, we find the 10s/30s curve, for instance, to be trading too flat versus fundamental drivers and we remain biased towards long-end steepeners ( Figure 40).  As we have discussed before, in addition to structural bullish duration bias being supportive of steepeners, other factors also support steepeners over the medium term.  For instance, receiving flows from European banks in the intermediate sector to unwind some of the paying hedges set up during the hiking cycle and/or protect the net interest income as rates decline should support steepeners.  Similarly, Dutch PF regulatory changes and already high level of hedging of interest rate risk from these PFs should limit the extent of receiving flows at the ultra-long end and not be a deterrent to our steepening views. Finally, technical factors such as unwinding of 10s/30s steepeners have reduced curve exposure contributing partly to current valuations, and making a steepening more likely in a reds driven rally.

Figure 40: The 10s/30s curve appears too flat versus fundamental drivers.  We have a steepening bias over the medium term

10s/30s swap curve regressed against 1) 3Yx1Y swap rate (X₁, quadratic fit), 2) ECB balance sheet as a % of GDP (X₂), 3) Feds UST 10Y ACM term premium (X₃), 4) VIX (X₄), 5) ECB Jan22 meeting dummy (X₅), and 6) 10s/30s USD swap curve (X₆); since 1 Jan 2017; bp

Source: J.P. Morgan.

Figure 41: Hold 2Yx2Y/10Yx10Y swap curve steepeners (proxy to 10s/30s) and 10s/30s conditional bull steepeners

2Yx2Y/10Yx10Y swap curve regressed against 3M forward 10s/30s swap curve; past 3M; bp

Source: J.P. Morgan.

Overall, as a medium-term bullish proxy, we keep steepening exposure at the long-end of the curve via 10s/30s conditional bull-steepeners (6M receivers) and 2Yx2Y/10Yx10Y swap curve steepener, which is a carry-efficient proxy for 10s/30s but offers better carry characteristics ( Figure 41).

On a cross market basis, we have a strong conviction on expected outperformance of Euro yield versus USD, given relative difference in macro fundamentals, despite the recent underperformance over the past few days.  We continue to stay long Jun24 Euribor calls versus SOFR that we had recommended few weeks ago.

Take tactical profit in €STR/SOFR paid position

In our 1Q24 Cross currency basis outlook, we had recommended paying 2Yx1Y €STR/SOFR basis.  The rationale was to benefit from expected relative monetary policy dynamics (outperformance of SOFR yield versus €STR and a larger reduction in Fed’s balance sheet versus ECB), positive carry from paying front-end basis, and expected increase in Yankee issuance.  The cross-market rate differential has exhibited decent volatility with 2Yx1Y €STR/SOFR differential broadly unchanged since our publication; the basis has broadly followed the contours of the evolution of this cross-border differential but with limited empirical relationship on a YTD basis ( Figure 42).  Balance sheet dynamics is a slow moving factor.  That leaves cross-border issuance being the main driver for the basis over the short-term.  Indeed, a decent increase in Yankee issuance has been the primary driver of the recent narrowing of the €STR/SOFR basis ( Figure 43).  To recap, Yankee issuances are issuances by non-US institutions in USD and when swapped back in domestic currency leads to narrower basis.  Anecdotally, KFW and EIB issuances in USD over the recent past have been the primary driver of the basis.  We take tactical profit in paying 2Yx1Y €STR/SOFR basis as a potential reduction in these cross border flows as we approach the quarter-end, in line with historical seasonality, could lead to some near-term retracement ( Figure 44) (see Trade Recommendations).

Figure 42: The €STR/SOFR basis has narrowed recently while exhibiting limited relationship with cross market yield differential

2Yx1Y €STR/SOFR basis (lhs) versus 2Yx1Y €STR/SOFR yield differential (rhs; inverted); since 1 Jan 2024; bp

Source: J.P. Morgan.

Figure 43: A sharp increase in Yankee issuance has been the main driver of the narrowing of the basis over the past few days

2W MA of Yankee and Reverse Yankee issuances (lhs, €bn/day) versus 2Yx1Y €STR/SOFR basis (rhs, bp); since 1 Feb 2024;

Source: J.P. Morgan and Dealogic.

Figure 44: Take profit in paying 2Yx1Y €STR/SOFR basis as historically Yankee issuance tends to decline going into March-end which could pull the basis wider over the short-term

Average 2W MA of Yankee issuance around February end; past 20Y; €bn/day

Source: J.P. Morgan and Dealogic.

Basis update

Recent media articles have suggested the possibility of the ECB reducing the Refi/Depo corridor (currently at 50bp) further.  ECB commentary has suggested that they are likely to retain full-allotment at its lending operation.  As the TLTRO-III borrowings come to an end over the coming months, we expect some of these borrowings to shift to the LTROs.  A tighter corridor would reduce the financial penalty for banks that want to borrow from the ECB via the LTRO operations, which have minimal uptake currently.  Additionally, a tight corridor would ascertain that overnight funding rates (€STR, for example) stay close to the depo rates even as excess liquidity is declining over the coming years.

We highlight that on these Long Term Refinancing Operations (LTROs), the ECB will likely introduce further refinancing operations at refi rate with potentially narrower corridor, in our view as part of the revision of their operational framework expected to be announced next week.  In our view, the term of these operations will make them attractive for the banking sector especially if no stigma will be attached.  Overall, a reduction in the corridor and availability of relatively cheaper term funding from the ECB at the refi-rate would likely have a direct impact on banks borrowing rate in capital markets and potentially compress FRA/OIS (€STR/Euribor) and other bases (3s/6s, for example). Indeed, front-end FRA/OIS rates have tightened over the last few days ( Figure 45). 

Current and 2W ago FRA/OIS levels and changes; bp

Source: J.P.Morgan.

Current and 2W ago 3s/6s basis; levels and changes; bp

Source: J.P.Morgan.

 Additionally, upcoming ECB policy actions (balance sheet shrinkage via slowdown in PEPP reinvestments and TLTRO-III repayments) will likely have a modest upward impact on €STR fixings.  We analyse below the impact on €STR fixings coming from the channel of decline in excess liquidity.

During the period of increase in excess liquidity in the pre-QE era, we modelled EONIA-refi spread as a % of the Refi/Depo corridor as a function of excess liquidity.  As seen in Figure 46, the EONIA/depo spread (as a % of the corridor) remained broadly stable for excess liquidity above say €500bn.  On absolute levels, EONIA fixings averaged around 4-5bp above the depo rate (or about 10% of the initially 50bp corridor), especially in the NIRP world ( Figure 47).  This translates to €STR/depo spread of around 3-4bp (€STR being below depo rate).  

Figure 46: Historically, excess liquidity has to decline below €500bn for it to have a significant impact on front-end rates. While this may appear too low in the current context, we do not expect any significant impact on €STR (and other front-end rates and spread) from €0.5-1tn decline in excess liquidity

(EONIA-refi) spread as a % of the (Refi-DFR) corridor regressed against excess liquidity;1 Jan 2010 – 31 Dec 2014; %

Source: J.P. Morgan and ECB.

Figure 47: EONIA fixings have been tethered to the depo rate in the high excess liquidity world

Depo rate (lhs, %), Refi/Depo corridor and EONIA/Depo spread (rhs, bp); since 1 Jan 2010;

Source: J.P. Morgan and ECB.

€STR/DFR spread versus 1M MA of excess liquidity; since 1 Jan 2020; bp

Source: J.P.Morgan and ECB.

€STR/DFR spread versus 1M MA of excess liquidity; since 1 Nov 2022; bp

Source: J.P.Morgan and ECB.

 More recently, the relationship has been mixed.  To analyse the impact of decline in excess liquidity on front-end rates, say €STR/depo spread which is currently around 9.5bp, we look at the relationship of the €STR/depo spread versus excess liquidity in Figure 48 and analyze this into three segments using recent data points.  From Jan2020-Dec2021, excess liquidity was increasing at a rapid pace and the €STR/depo spread continued to increase with a beta of around 1.3bp per €1tn increase in excess liquidity.  This spread widened further around 1.5-2bp during the course of 2022 even though excess liquidity remained broadly unchanged and this is potentially due to almost but not full transmission of ECB rate hikes to overnight funding markets during the first half of the hiking cycle (black dots in Figure 48).

Since the ECB modified the TLTRO-III conditions and allowed earlier repayment of TLTRO-III borrowings in October 2022 (effective from late November 2022), excess liquidity has declined around €1tn.  However, €STR fixings continues to remain broadly stable with a very low beta to decline in liquidity (around 0.4bp increase in €STR fixings relative to depo rate per €1tn decline in excess liquidity) (blue dots in Figure 48).  In the process of declining excess liquidity, we would anticipate the recent low beta of the spread versus excess liquidity to persist and potential increase in the beta (or commensurately a decline in the €STR/depo spread) only when excess liquidity comes closer to €1-1.5tn, although with limited empirical evidence on this even at these high levels of excess liquidity.

In an exaggerated case where the beta of Jan2020-Dec2021 resumes (1.3bp per 1tn decline in excess), €STR fixings will likely fix around 4-5bp below depo rate only when excess liquidity drops to around €0.5-1tn level.  In summary, we expect the €STR/depo spread to narrow with declining excess liquidity but in a non-linear fashion with a relatively larger decline only as excess liquidity approaches the €1-1.5tn mark versus current level of around €3.5tn, a dynamic that is still quite few years away given current pace of QT and ECB virtually ruling out outright sale. 

Overall, we believe that medium-term risks are biased towards tighter €STR/Euribor (E/E) spreads and is a challenge to our previous bias of a steeper E/E curve.  We have a bias towards narrower E/E basis especially in the 2H24 sector which also offers attractive roll-down.

Swap Spreads

Since our last weekly publication swap spreads were mixed across the curve, with a broad steepening of the swap spread curve and a notable underperformance of swap spreads vs. 6s, itself anecdotally driven by broad unwinding of basis wideners mostly in the belly of the curve, resulting in tighter FRA/OIS and 3s/6s ( Figure 49) (Figure 49 below).  Flows were driven by rolling activity with light swapped issuance receiving activity, broadly in line with seasonality.  The directionality was mixed across the curve and also different between €STR and 6s swap spread due to the repricing of FRA/OIS across the EUR curve mentioned above ( Figure 50) (Figure 50 below). We attribute the narrowing in a rally up to the 10Y sector to broad cheapening of swap spreads on the back of easier financial conditions, whereas increased demand of duration at the long end of the curve has been a catalyst for the Buxl outperformance over the week.

Figure 49: Swap spreads were mixed across the curve, with a broad steepening of the swap spread curve and a notable underperformance of swap spreads vs. 6s

Current level of Jun24 Schatz, Bobl, Bund, and Buxl maturity matched 6s and OIS (€STR) swap spreads and change since 1^st March; bp

Source: J.P. Morgan.

Figure 50: The directionality was mixed across the curve and also different between €STR and 6s swap spread due to the decent repricing of FRA/OIS across the EUR curve

Ratio between changes since 1^st March of Jun24 Schatz, Bobl and Bund €STR and 6s swap spreads and changes in respective invoice yields; bp

Source: J.P. Morgan.

On a longer horizon we note that the directionality of swap spreads to yield levels has been broadly negative across the curve with more pronounced negative beta at the long end of the curve ( Figure 51) (Figure 51 below).  Looking ahead we believe that the negative directionality at the long end of the curve could remain, with our bullish duration outlook broadly supporting wider swap spreads at the long end of the curve.  On the rest of the curve we continue to believe that eventually easier financial conditions and lower ECB terminal rates should support narrower swap spreads in the belly of the curve, although the bulk of the cyclical narrowing seems to have already taken place vs. our valuations.  On a long term analysis, we find that German swap spreads continue to trade close to their multi-year narrowest levels especially at the short and very long end of the curve, which are the sectors we favour for widening positions ( Figure 52) (Figure 52 below).

Figure 51: Directionality has been broadly negative across the curve with more pronounced negative beta at the long end of the curve

Beta and r-squared of 2Y, 5Y, 10Y, 15Y and 30Y German b/m, Schatz, Bobl, Bund and Buxl swap spreads against respective benchmark and invoice yields since late December 2023 and between early September 2023 and 27 December 2023; %

Source: J.P. Morgan.

Figure 52: German swap spreads continue to trade close to their multi-year narrowest levels especially at the short and very long end of the curve

Current levels and statistics since early 2015 of Schatz, Bobl, Bund and Buxl swap spreads versus €STR; bp (except for Z-score unitless)

Source: J.P. Morgan.

At the short end of the curve, we note that the broad cheapening trend of Schatz €STR swap spread paused and partly reversed over the past few trading sessions, where we anecdotally have seen some dip buying across speculative investors.  The valuation framework is consistent with our previous analysis.  Schatz €STR swap spread is trading particularly cheap in our long-term fair value model vs. funding rates, intra-EMU spreads and short dated volatility (residual at -13bp) but the level of residual remains still negative (residual at -8bp) once volatility is removed as regressor, suggesting that current valuations do not carry any risk premium ( Figure 53) (Figure 53 below).  Similar considerations on still cheap valuations but less extreme than recent lows can be seen by looking at our short-term fair value model ( Figure 54) (Figure 54 below).

Schatz invoice OIS swap spread (vs. €STR rolling 1st of delivery month) and model levels implied by regression against: 1) 1M MA of adjusted* RFR Germany repo specialness (X1); 2) 10Y BTP/Bund and 3Y OAT/Bund since the US Presidential election (X2 and X3); 3) 3Mx2Y EUR swaption volatility (X4); since 1 Jan 2015; bp

Source: BrokerTec, MTS trading, and J.P.Morgan.
* Excluding year-end effect from 23 December 2016 to 6 January 2017 and the year-end and quarter-end effect over the last business days of the quarters. Repo Funds Rate Germany (RFR Germany) is a daily euro repo index calculated from trades executed on the BrokerTec and MTS electronic platforms. All eligible repo trades are centrally cleared and RFR Germany is calculated and published by ICAP Information Services. RFR Germany is calculated with repo trades that use German sovereign government bonds. RFR Index source: Broker Tec and MTS trading.

Schatz invoice OIS swap spread (vs. €STR rolling 1st of delivery month) and model levels implied by regression against: 1) 1M MA of adjusted* RFR Germany repo specialness (X1); 2) 10Y BTP/Bund and 3Y OAT/Bund since the US Presidential election (X2 and X3); since 1 Jan 2015; bp

Source: BrokerTec, MTS trading, and J.P.Morgan.
* Excluding year-end effect from 23 December 2016 to 6 January 2017 and the year-end and quarter-end effect over the last business days of the quarters. Repo Funds Rate Germany (RFR Germany) is a daily euro repo index calculated from trades executed on the BrokerTec and MTS electronic platforms. All eligible repo trades are centrally cleared and RFR Germany is calculated and published by ICAP Information Services. RFR Germany is calculated with repo trades that use German sovereign government bonds. RFR Index source: Broker Tec and MTS trading.

Figure 54: Similar considerations on still cheap valuations but less extreme than recent lows can be seen by looking at our short-term fair value model

Schatz invoice OIS swap spread (vs. €STR rolling 1st of delivery month) and model levels implied by regression against: 1) 10Y BTP/Bund swap spread vs 6s (X1), 2) 1M MA of adjusted* RFR Germany repo specialness (X2); since 1 Jan 2021; bp

Source: BrokerTec, MTS trading, and J.P.Morgan.
* Excluding year-end effect from 23 December 2016 to 6 January 2017 and the year-end and quarter-end effect over the last business days of the quarters. Repo Funds Rate Germany (RFR Germany) is a daily euro repo index calculated from trades executed on the BrokerTec and MTS electronic platforms. All eligible repo trades are centrally cleared and RFR Germany is calculated and published by ICAP Information Services. RFR Germany is calculated with repo trades that use German sovereign government bonds. RFR Index source: Broker Tec and MTS trading.

Over the past few weeks, we also have run a short-term regression of Schatz swap spreads vs. other risk-off measures across the Euro area fixed income markets and we note that Schatz swap spread was screaming cheap vs. other risk off indicators such as short dated volatility, intra-EMU and FRA/OIS spreads.  Over the past week the roll effect and a modest rebound in Schatz €STR swap spread has taken place as volatility has softened and while intra-EMU and FRA/OIS spreads narrowed.  Therefore, on short term regression Schatz €STR swap spread is no longer too cheap vs. other risk-off measures ( Figure 55) (Figure 55 below).  We attribute this more to the idiosyncratic moves in intra-EMU and FRA/OIS rather than an assessment of the broad valuation.

Next week the ECB will most likely provide details about the revision of the operation framework, with some indication about the evolution of their structural asset portfolio and long term refinancing operations.  Our bias is that both the structural portfolio and the collateral needed for long term refinancing operations will at the margin support a richening of funding rates which should eventually be supportive of wider swap spreads ( Figure 56) (Figure 56 below).  Additionally, while not our baseline we note that any change to the remuneration of either banks reserve or government deposits, should support some demand in short dated paper either outright or in repo, further supporting the richening of funding rates.

Combining our valuation analysis and this operation framework review we continue to find value in outright widening exposure at the short end of the curve and continue to hold Jun24 Schatz €STR swap spread widener targeting a level between 15bp and 20bp over the next quarter.

Figure 55: On a short term regression Schatz €STR swap spread is no longer too cheap vs. other risk-off measures

Statistics of regression of Schatz/OIS swap spread vs 3Mx10Y volatility, 10Y BTP/Bund, 10Y OAT/Bund, and 1Yx1Y FRA/OIS; past 3M

Source: J.P. Morgan.

Figure 56: We expect the revision of the ECB operational framework to be supportive of richer funding rates

Key components of the ECB operational framework review with expected outcome and implications for funding rates;

Source: J.P. Morgan.

 The broad narrowing of swap spread in the belly of the German curve has been a consensus trade of 2024 which has delivered in line with expectations.  For a while we have discussed that in addition to the increase in collateral availability and easier financial conditions an additional factor to support swap spreads narrowers in the belly of the curve has come from financial institutions unwinding paying flows on the swap curve implemented into the hiking cycle.  After the move, we find that Bund swap spread is now trading marginally too narrow in our revised fair value model, where we use the stock effect of QE by looking at the ECB balance sheet as percentage of GDP ( Figure 57) (Figure 57 below).  Similar considerations on valuations are obtained using a short term model vs. short term rate (as measure of financial conditions) and the flow of ECB purchases (currently negative given QT) ( Figure 58) (Figure 58 below).

We stay neutral on Bund swap spreads.  Over the medium term we still believe that  easier financial conditions and technical factor such as receiving from financial institutions support further narrowing however we remain caution as valuations are now on the cheap side and narrowers still quite subscribed across investors.

Figure 57: Bund swap spread is now trading marginally too narrow in our revised fair value model, where we use the stock effect of QE by looking at the ECB balance sheet as percentage of GDP…

Front Bund invoice swap spreads vs. 6s regressed against 1) 10Y German benchmark yield (X1), 2) 10Y weighted peripheral spread (X2), 3) a dummy variable to include the PEPP announcements in March 2020 (X3), 4) 1M OIS rate 2Y forward (X4), 5) ECB balance sheet as % of GDP (X5), 6) 1M moving average of fixed rate swapped issuance (X6); since 1 Jan 2019; bp

Source: ECB, Dealogic, and J.P.Morgan.

Figure 58: …with similar conclusion in a short term model the flow of ECB purchases

Front Bund invoice swap spreads vs. 6s regressed against 1) 1M moving average of the flow of ECB purchases (X1), 2) 1M OIS rate 2Y forward (X2); since 1 Jan 2021; bp

Source: J.P. Morgan.

Further out the curve we have been fading the cheapening of Buxl via holding outright widener in the 10Yx20Y German forward swap spread, implemented via Mar24 Bund narrower and Mar24 Buxl widener.  The forward swap spread cheapened significantly over the period driven by lack of duration demand and anecdotal unwinding of 10s/30s steepeners.  We roll widening exposure in 10Yx20Y German forward swap spread from Mar24 to Jun24 contracts on valuations and as proxy for steeper 10s/30s swap curve ( Figure 59) (Figure 59 below) (see Trade Recommendations).

Figure 59: Roll widening exposure in 10Yx20Y German forward swap spread from Mar24 to Jun24 contracts on valuations and as proxy for steeper 10s/30s swap curve

Forward swap spread between Bund and Buxl regressed against 10s/30s EUR swap curve; past 3M; bp

Source: J.P. Morgan.

Volatility

Implied volatility continued its decline over the past week across the curve with a decent underperformance of the front-end of the curve and a steepening of the volatility curve across tails ( Figure 60) (Figure 60 below).  1M delivered volatility when measured on a close-to-close basis has collapsed over the past few days and is now significantly below current level of implieds. 

Current and 1W ago levels (lhs) versus 1W change in various 3M expiry options implied volatility; bp/day

Source: J.P. Morgan.

Current implied vol versus 1M delivered vol of underlying swaps; bp/day

Source: J.P. Morgan.

Figure 61: Front-end implieds are at YTD lows whereas 10Y implieds are at the lows of the past couple of years

3Mx2Y and 3Mx10Y implied volatility; since 1 Jan 2021; bp/day

Source: J.P. Morgan.

 Front-end implieds are now at their YTD lows although but still well above the lows seen in October 2023 when the ECB was expected to stay on hold for a few months; aggressive repricing of central bank expectations has subsequently pushed implieds higher.  Long-end implieds, on the other hand, are now at the lows of the past couple of years ( Figure 61) (Figure 61 below). 

Looking ahead, evolution of ECB easing expectations will remain the primary driver of volatility across the surface and especially at the front-end.  To this effect, ECB delivery from yesterday is supportive of some further decline in implieds.  To recap, President Lagarde implicitly put the bar for a April cut to be very high given lack of significant additional data by that meeting while bringing a June cut into center-stage.  Indeed, ECB commentary since the meeting has been consistent with this message and various board members are implicitly endorsing the start of the easing cycle in June ( Figure 62) (Figure 62 below). Of course, the path of future cuts, after the first cut, will also be a dominant driver for volatility going forward. 

Figure 62: Commentary from ECB board members post the meeting is laying grounds for a June start to the easing cycle.  The interim on-hold period is supportive of lower volatility although we remain wary of outright short gamma positions

Selected comments from ECB members

Source: J.P. Morgan.

For now, we keep our overall bearish gamma bias driven by an expected on-hold stance from the ECB, and even a measured start of the easing cycle.  The sharp decline in delivered volatility offers attractive gamma carry to outright short gamma positions. 

Nevertheless, we express our bearish view only selectively and focus this at the 10Y+ sector via unhedged straddles that we have been recommending over the past few weeks.  While outright short gamma positions are attractive, especially at the front-end, they are fraught with risk.  At the front-end, current levels of implieds, which are close to the recent lows, could attract interest from investors to buy volatility either outright or potentially via bullish structures.  Indeed, 3Mx2Y implieds below 6bp/day could appear optically attractive and have rebounded higher from these levels in the past.  Additionally, ECB’s data dependence nature makes front-end yields vulnerable to surprises in macro data points with a larger sensitivity.  To this effect, the upcoming US CPI could impart some volatility over the short-term.  Thus, with jump risk remaining high, we remain wary of outright short gamma positions at the front end despite an expected on-hold ECB.

We prefer to keep short volatility positions at the long-end via unhedged straddles in 6Mx10Y which we have been recommending over the past few weeks.  Fixed dated/fixed strike implieds have declined around 0.4bp/day since inception which accounts for around 40% of the total P&L of around 70bp of notional since inception ( Figure 63) (Figure 63 below).  Underlying swap yield has hovered around the strike (strike of 2.54% versus current level of 2.50%) leading to large theta gains in the trade.  We expect this dynamic to continue over the near term and stay in the trade.   

Figure 63: Stay short 6Mx10Y unhedged straddle as yields are expected to remain in a range over the coming weeks

Evolution of fixed dated/fixed strike implied volatility (lhs, bp/day) and straddle premium (rhs, bp of notional); since 2^nd Feb 2024;

Source: J.P. Morgan.

Figure 64: Keep 1Yx1Y/1Yx10Y volatility curve flatteners across tails

(1Yx10Y – 1Yx1Y) implied volatility spread versus 1M delivered volatility spread (1Yx10Y – 1Yx1Y); since 1 Jan 2022; bp/day

Source: J.P. Morgan.

We also continue to hold 1Yx1Y/1Yx10Y volatility curve flatteners across tails.  In our view, high jump risk and uncertainty around evolution of ECB policy rate expectations even after the first cut will likely keep volatility at the front-end supported.  The implied volatility spread has exhibited some volatility but is overall marginally higher since inception on 26^th Jan 2024) – a move against the trade.  However, 1Yx1Y swaps have delivered around 1bp/day more than 1Yx10Y swaps versus a difference of 0.5bp/day priced in the options market.  We expect the relative outperformance of the front-end to continue and hold the trade ( Figure 64) (Figure 64 above).

Trade Recommendations

Swap Curve

• Take profit in paying 2Yx1Y €STR/SOFR basis
  Close paying €100mn 2Yx1Y (swap start date 15 Jan 2026, swap maturity date 15 Jan 2027) €STR/SOFR basis at -15bp vs entry level at -18.5bp; P&L since inception (12^th Jan 2024): +3.5bp;
• Keep receiving 1Yx1Y €STR swap
  Keep receiving 1Yx1Y (swap start date 9 Jan 2025, swap end date 9 Jan 2026) €STR swap at 2.52% vs entry level at 2.15%; P&L since inception (5^th Jan 2024): -37bp;
• Keep long Mar24 1Y Euribor midcurve call ladder
  Keep 1000 97.5/97.75/97.875 Mar24 Euribor 1Y midcurve call ladder at 1.5cents vs entry level at 5 cents with Mar25 Euribor futures at 97.29 vs entry level at 97.45; P&L since inception (9^th Feb 2024): -3.5cents;
• Keep long Jun24 1Y Euribor midcurve call ladder
  Keep 1000 97.5/97.75/98 Jun24 Euribor 1Y midcurve call ladder at an entry level of 3 cents with Jun25 Euribor futures at 97.48;
• Keep long Apr24 96.375/96.25 1x2 put spread
  Keep long 1000 Apr24 96.375 Euribor put versus short 2000 96.25 put at 3.3cents vs entry level at 3cents (Jun24 Futures at at 96.35 vs entry level at 96.38); P&L since inception (23^rd Feb 2024): 0.3cents;
• Keep Jun25/Jun26 conditional bull steepener via Jun24 midcurve calls
  Keep long 1000 97.50 Jun24 1Y Euribor midcurve calls (Jun25 Euribor futures at 97.47) versus short 1000 97.625 Jun24 2Y Euribor midcurve calls (Jun26 Euribor futures at 97.71) to remain in a bull steepener at a forward spread of -24bp vs entry level at -12.5bp; P&L since inception (1^st Mar 2024): -4cents;
• Keep Mar25/Mar26 conditional bull steepener via Mar24 midcurve calls
  Stay long 1000 97.75 Mar24 1Y Euribor midcurve calls (Mar25 Euribor futures at 97.29) versus short 1000 97.875 Mar24 2Y Euribor midcurve calls (Mar26 Euribor futures at 97.70) to remain in a bull steepener at a forward spread of -42bp vs entry level at -12.5bp; P&L since inception (5^th Jan 2024): -1.5cents;
• Keep 10s/30s conditional bull steepener via 6M receiver
  Keep long €100mn 2.47% 6Mx10Y receiver (ATMF strike 2.52%, notification date 14 June 2024, maturity date 18 June 2034) versus short €41.8mn 2.35% 6Mx30Y receivers (ATMF strike 2.35%, notification date 14 June 2024, maturity date 18 June 2054) to remain in a 10s/30s bull steepener at forward and spot spreads of -28bp and -29bp vs entry level at -12bp, respectively; P&L since inception (14^th Dec 2023): -44bp of notional;
• Keep 2Yx2Y/10Yx10Y swap curve steepener
  Keep receiving €350mn 2Yx2Y swap (swap start date 13 Feb 2026, swap end date 14 Feb 2028) versus paying €100mn 10Yx10Y swap (swap start date 13 Feb 2034, swap end date 15 Feb 2044) to remain in a 2Yx2Y/10Yx10Y swap curve steepener at 12bp vs entry level at 23bp; P&L since inception (9^th Feb 2024): -11bp;
• Keep 3s/10s conditional bear flattener via 3M payers
  Keep long €314.7mn 3Mx3Y 2.98% payer (ATMF at 2.86%, notification date 3 June 2024, maturity date 5 June 2027) versus short €100mn 3Mx10Y 2.75% payer (ATMF at 2.69%, notification date 3 June 2024, maturity date 5 June 2034) to remain in a 3s/10s bear flattener at forward and spot spreads of -19bp and -30bp, respectively, vs. entry level at -23bp; P&L since inception (1^st Mar 2024): 6bp;
• Keep long 3Mx(2s/10s) CMS 1x2 floor spread(A/A-20)
  Keep long €1bn notional of 3Mx(2s/10s) -30.1bp CMS floor (maturity date 28 May 2024) versus short €2bn notional of 3Mx(2s/10s) -50.1bp CMS floor (maturity date 28 May 2024) at 4.3bp vs entry level at 3.5bp; P&L since inception (23^rd Feb 2024): 1bp of yield;
• Keep 1Yx5Y ATMF/ATMF-50 receiver spread vs selling ATMF+49.4 OTM payer
  Stay long €100mn 2.37% 1Yx5Y receiver versus selling €100mn 1.87% 1Yx5Y receiver versus short €100mn 2.865% 1Yx5Y payer (ATMF strike 2.37%, notification date 3 Feb 2025, maturity date 5 Feb 2030) to remain in a A/A-50bp receiver spread vs selling A+49.4bp payer at a forward rate of 2.44% vs entry level at 2.37%; P&L since inception (2^nd Feb 2024): -5bp of notional;
• Keep long Jun24 ATM Euribor calls versus SOFR OTM calls
  Keep long 1000 96.5 Jun24 Euribor call (Jun24 Euribor futures at 96.47) versus selling 1080 95.25 Jun24 SOFR calls (Jun24 SOFR futures at 95.125) at a forward spread of 140bp vs entry level of 125bp; P&L since inception (9^th Feb 2024): -1cent;

Swap spread

• Keep Jun24 Schatz €STR swap spread widener
  Keep long 1000 Jun24 Schatz futures versus paying €114.8mn CTD maturity matched €STR swaps (swap start date 6 June 2024, maturity date 19 Mar 2026) to remain in a swap spread widener at 10.7bp vs entry level at 11.1bp; P&L since inception (23^rd Feb 2024): -0.4bp of yield;
• Rotate into Jun24 from Mar24 10Yx20Y swap spread widener via Bund and Buxl futures
  Buy1000 Jun24 Buxl futures versus paying €152.3mn CTD maturity matched swaps (262% risk, swap start date 6 June 2024, swap maturity date 15 Aug 2048) (Buxl widener) versus short 1000 Jun24 Bund futures versus receiving €135.4mn CTD maturity matched swaps (swap start date 6 June 2024, swap maturity date 15 Feb 2033) (100% risk in Bund narrower) to enter in a 10Yx20Y forward swap spread widener at -35.5bp;
  Keep long 1000 Mar24 Buxl futures versus paying €157mn CTD maturity matched swaps (258% risk, swap start date 7 Mar 2024, swap maturity date 15 Aug 2048) (Buxl widener) versus short 1000 Mar24 Bund futures versus receiving €138mn CTD maturity matched swaps (swap start date 7 Mar 2024, swap maturity date 15 Feb 2033) (100% risk in Bund narrower) to remain in a 10Yx20Y forward swap spread widener at -33bp vs entry level at -10bp (spread defined as 2.58*30Y – 10Y); P&L since inception (5^th Jan 2024): -23bp;

Volatility

• Keep short 6Mx10Y unhedged straddle
  Keep short €100mn 2.54% 6Mx10Y straddle (notification date 2 Aug 2024, maturity date 19 Aug 2034) at 5.16bp/day vs entry level at 5.54bp/day; P&L since inception (2^nd Feb 2024): 64bp of notional;
• Keep 1Yx1Y/1Yx10Y volatility curve flatteners across tails
  Keep long €100mn 1Yx1Y 2.34% straddles (notification date 27 Jan 2025, maturity date 29 Jan 2026) versus short €12.4mn 1Yx10Y 2.59% straddles (notification date 27 Jan 2025, maturity date 29 Jan 2035) to remain in a volatility curve flattener at -0.55bp/day vs entry at -0.5bp/day. This trade was structured to be gamma-neutral at inception and requires active delta hedging; P&L since entry (26^th Jan 2024): 1.9bp of notional;

* P/L for open trades calculated using Thursday’s closes. New and closing trades are priced as of 1PM on non-payroll Fridays and 2PM on payroll Fridays.

Trades closed over the past 12 months

Source: J.P. Morgan.

Trades closed over the past 12 months

Source: J.P. Morgan.

Still waiting patiently on the sidelines

• Gilt yields rallied on the week, driven mainly by US data and Powell commentary and ECB rhetoric suggesting easing will occur this year though a degree of patience is still needed
• The Chancellor’s budget delivered no major surprises as the main taxation and revenue raising policies had been flagged in the media in advance. The net fiscal boost to growth is a small positive, particularly for 2024, at around 0.2-0.3%, which then eventually turns into a drag in future years. This is unlikely to have any material impact on BoE policy decisions over the coming months, but we note that the growth profile does look slightly more positive than before
• We keep our Aug24/Nov24 PC OIS curve flattener with around 28bp of easing priced and given the 65bp of cuts priced by the end of this year is less than our forecast for 75bp of BoE easing over 2H24. Given the lack of BoE urgency, sticky wage growth and potential upside growth risks, we close our H4/U4 3M SONIA flattener for a loss
• We have a flattening bias on 1Y/1Yx1Y SONIA as a way of expressing a medium-term bullish bias, but wait until we have seen next week’s US CPI and UK labour market data before entering
• Following the Chancellor’s Budget on Wednesday, the DMO announced the financing remit for FY24/25 with an increase in planned gilt sales vs. current fiscal year of just under £30bn to £265bn. Gilt issuance net of redemptions and QT will likely remain high, at over £150bn for FY24/25
• 10s/30s gilt curve looks modestly too flat vs. the level of yields and adjusting for the 10s/30s UST curve, but the negative directionality of 10s/30s gilt curve vs. 1Yx1Y SONIA has weakened with the curve exhibiting stronger positive directionality with 5Yx5Y gilt yields
• We enter tactical 10s/20s gilt curve steepeners given RV considerations and upcoming 20Y supply on 19 March (4T43)
• In SONIA space, we keep our 5Yx5Y/15Yx15Y SONIA curve flattener as a low-beta bearish duration proxy, although we note that the RV dislocation has mostly corrected
• Gilt RV: we stay short the belly of the cash-and-duration neutral Jan32/Jan38/Oct41 gilt fly (risk weights 40%/-100%/60) as a mean-reversion trade. This fly is non-directional vs. both the level of outright yields and vs. the slope of the Jan32/Oct41 curve. For investors looking for a bearish duration RV proxy, we note that selling gilt Oct31 vs. gilt Dec30 on an equational basis looks attractive, given the relative richness of the implied forward yield vs. 5Yx5Y yields

Gilt yields rallied on the week driven mainly by US data, Powell commentary and ECB rhetoric suggesting easing will occur this year, but a degree of patience is still needed with the UK curve flattening as 30Y yields outperformed ( Figure 65).

In terms of our portfolio, we stay cautious on outright duration as, absent a slide into recession i,t will likely take time to assess the degree of inflation stickiness. We hold tactical Aug24/Nov24 MPC OIS curve flatteners, but close H4/U4 3M SONIA curve flatteners for a loss. We shift to a flattening bias on 1Y/1Yx1Y SONIA as a way of expressing a medium-term bullish bias, but do not look to enter until after we have seen next week’s UK labour market report and US CPI data. We hold 5Yx5Y/15Yx15Y SONIA curve flatteners as a low-beta bearish duration hedge to our portfolio. In RV space, we stay short the belly of the cash-and-duration neutral Jan32/Jan38/Oct41 gilt fly and enter tactical 10s/20s gilt curve steepeners.

Figure 65: Gilt yields rallied over the past week with the 2s/10s curve flattening

Change in benchmark gilt yields since 1 Mar; bp

Source: J.P. Morgan.

Figure 66: The DMP survey shows a slight fall in one-year ahead wage expectations in February, but the 3M average is still above 5%

DMP expected wage growth over next 12M, single month reading and 3M average; %

Source: BoE DMP

The Chancellor’s budget delivered no major surprises, as the main taxation and revenue raising policies had been flagged in the media in advance. Fiscal headroom was smaller than expected, with the total net fiscal easing announced worth about £5bn (~0.1%/GDP). National Insurance Contributions were cut by 2% for employees (to 8%) and self-employed (to 6%) with some offset from revenue raising measures. The net fiscal boost to growth is a small positive, particularly for 2024 at around 0.2-0.3%, which then eventually turns into a drag in future years as tax raising measures only really kick-in from 2025. This is unlikely to have any material impact on BoE policy decisions over the coming months, but we note that the growth profile does look slightly more positive than before. Over the longer run, we still expect fiscal tightening post the election, given medium-term growth forecasts remain overly optimistic.

Domestic data was relatively light this past week with housing data showing a modest up-tick in house price growth and the BoE DMP survey showing one-year expected wage growth still above 5% on a 3M average basis in February ( Figure 66), which is well above the BoE’s average earnings forecast for the end of next year. Next week sees the January labour market report, which will likely show the labour market is still very tight, but interpreting the ONS labour market data at the moment is difficult, given the sampling issues, and the survey won’t move to the new data collection methodology until September.

Figure 67: We keep tactical Aug24/Nov24 MPC OIS curve flatteners

MPC OIS steps and cumulative tightening priced (RHS); bp

Source: J.P. Morgan.

Figure 68: The SONIA curve prices more easing over 2025 vs. 2024, a different dynamic compared to ECB rate cut expectations

Expected change in policy rates priced into meeting dates curves for ECB, Fed and BoE over 2024 and 2025; bp

Source: J.P. Morgan.

MPC OIS for 2024 are little changed over the week, with almost 25bp of easing priced for August and a cumulative 65bp of easing priced by the end of this year. We keep our Aug24/Nov24 PC OIS curve flattener with around 28bp of easing priced and given the 65bp of cuts priced by the end of this year is less than our forecast for 75bp of BoE easing over 2H24 ( Figure 67).

Further out SONIA rates have rallied and the money market curve has bull flattened with over 140bp of BoE easing now priced by the end of 2025 although implied Bank Rate priced at 3.75% looks too high vs. our 3% end of 2025 Bank Rate forecast . The SONIA curve now prices more easing over 2025 than 2024 compared to the roughly equal amount of Fed easing priced for 2024 and 2025 and the larger ECB easing priced for 2024 vs. 2025 ( Figure 68). This reshuffling of BoE rate cut expectations likely reflects the sticky and still elevated backdrop level of wage growth in the UK alongside a slightly more positive growth backdrop. We see some upside risks to growth but we would need to see evidence of more concrete strength in the output data over a sustained period for this upside growth risk to start to impact our view on the timing of the first BoE cut in August.

The H4/U4 3M SONIA curve is little changed on the week and has been hovering around its current levels for the past few weeks and, given the lack of BoE urgency, sticky wage growth and potential upside growth risks, we see limited potential for this curve to flatten and we close our H4/U4 3M SONIA flattener for a loss. In the event of a further rally in yields, expect a continuation of the recent dynamic of 2025 SONIA rates outperforming 2024 SONIA rates rather than flattening H4/U4 SONIA curve. For example, the U4/U5 3M SONIA curve has flattened as yields have rallied over the past week. This curve is highly positively correlated with the level of yields, but we note that, on a regression basis, U4/U5 3M SONIA curve looks too flat vs. the level of U5 SONIA ( Figure 69), whereas 1Y/1Yx1Y SONIA curve looks fair value vs. the level of 1Yx1Y SOFR. We have a flattening bias on the 1Y/1Yx1Y SONIA curve as a way of expressing a medium-term bullish bias, but do not look to position until after we have seen next week’s US CPI print and UK labour market report.  

Figure 69: U4/U5 SONIA curve remains highly directional with yields, but looks too flat on a regression basis

U4/U5 3M SONIA curve regressed against U5 3M SONIA yield; past 6M; bp

Source: J.P. Morgan.

Figure 70: 1Yx1Y SONIA continues to exhibit strong positive directionality vs. 1Yx1Y SOFR. We expect this to continue until we see more visibility on the timing of the start of Fed easing

1Yx1Y SONIA regressed against 1Yx1Y SOFR; past 6M; %

Source: J.P. Morgan.

Cross market, 1Yx1Y SONIA continues to exhibit strong directionality with 1Yx1Y SOFR ( Figure 70), as has been the case for the past 6M as both the BoE and Fed have been firmly on hold, and looks fair value on a regression basis. We expect this strong correlation to persist until the start of Fed easing approaches. Next week sees the February US CPI data and we expect a 0.3% m/m print, in line with consensus, but we see a greater level of uncertainty around next week’s release and acknowledge the risk that a second consecutive upside surprise could push US real yields higher, at least temporarily. We stay neutral outright duration.

£265bn gilt issuance next fiscal year

Following the Chancellor’s Budget on Wednesday, the DMO announced the financing remit for FY24/25 with an increase in planned gilt sales vs. current fiscal year of just under £30bn to £265bn and no change to net T-bill issuance for debt management purposes. Given this, we expect gilt issuance net of redemptions and ongoing BoE QT sales to be slightly above £150bn for FY24/25, the highest level historically ( Figure 71). Looking further ahead, based on estimates for borrowing and CGNCR for FY25/26, we estimate gross gilt sales will remain elevated at around £260bn and that gilt sales net of redemptions and QT will only modestly decline to around £135bn for FY25/26.

Figure 71: Gilt issuance net of redemptions and QT will likely remain high at over £150bn for FY24/25

Total gilt issuance gross and net of redemptions and QE/QT, fiscal year; £bn

Source: DMO, J.P Morgan, BoE

Figure 72: For FY24/25, long-end gilt issuance will fall and medium gilt issuance will increase to 31% given the DMO plans to conduct a syndication in the medium sector

FY23/24 and FY24/25 gilt issuance by maturity bucket*; £bn and % of total issuance

Source: DMO, J.P Morgan*We also provide an estimate of supply by bucket with the unallocated redistributed across medium, long and linker maturity buckets

The total gilt sales announced for FY24/25 were close to median market expectations, but were around £20bn higher than our forecasts, reflecting a slightly larger deficit and no net contribution from T-bills for next fiscal year. The issuance split across the curve ( Figure 72) was broadly in line with expectations, with a reduction in the proportion of long maturity issuance, but a slightly larger-than-expected increase in the proportion of medium gilt issuance, given the DMO announced a planned syndication in the medium bucket. The proportion of index-linked gilts was reduced by just over 1% to 10.9%, but we would expect some of the £10bn in currently unallocated issuance to eventually be issued in index-linked gilts, potentially taking the proportion of linker issuance closer to 11.5%. This is probably a bit larger than consensus market expectations.

Next fiscal year, the DMO plans to conduct 4 conventional, one medium and three long, and 3 index linked gilt. We pencil in conventional syndications in May, June, October and next January and linker syndications in July, November and next February. Figure 73 shows our projections for total conventional gilt supply in cash and 10Y equivalent terms until the end of FY24/25 (i.e. 1Q25) and we expect total supply net of QT will remain elevated in 2Q24 before an expected dip over 3Q24. Figure 74 shows our expected monthly issuance in 10Y equivalent terms, split by maturity sector with May, October and next January expected to be the largest issuance months.

Figure 73: Total conventional gilt supply net of QT will remain elevated in 2Q24 before an expected dip in 3Q24

Total conventional gilt supply net of BoE QT sales by quarter, in cash terms and 10Y duration equivalents; £bn

Source: DMO, J.P. Morgan, BoE

Figure 74: We expect monthly issuance in 10Y equivalent terms to be close to £40bn in May, October and next January

Monthly conventional gilt supply net of BoE QT sales split by maturity sector in 10Y duration equivalents; £bn

Source: DMO, J.P. Morgan, BoE

Curve and RV

Long-end yields have outperformed over the past week with 30Y yields rallying almost 15bp and they are now 6bp below their 3M averages. The 10s/30s gilt curve has flattened around 3bp over the week with 10s/30s SONIA curve flattening a couple of bp with the 10s/30s gilt curve at its flattest level since late November and the 10s/30s SONIA curve at its flattest level since late December. From a valuation standpoint, the 10s/30s gilt curve looks a few bp to flat vs. the level of yields and adjusting for the 10s/30s UST curve on a 6M regression basis ( Figure 75). However, we note that, over the past month, the 10s/30s gilt curve has exhibited weaker negative directionality with the level of 1Yx1Y SONIA and has become more positively correlated with intermediate forward yields ( Figure 76). We think a bull-flattening dynamic of 10s/30s is temporary and reflects a combination of potential wash-out of steepening positions, some modest ‘relief’ flattening post a budget that did not contain any problematically large fiscal easing as well as anecdotal end-user demand.

Figure 75: 10s/30s gilt curve looks modestly too flat vs. the level of yields and adjusting for the 10s/30s UST curve…

Residual from regressing 10s/30s gilt curve *vs. 1Yx1Y SONIA, 30Y gilt yield and 10s/30s UST curve; past 6M; bp

Source: J.P. Morgan.*10s/30s gilt = -0.25 * 1Yx1Y SONIA + 0.3 * 30Y gilt + 0.03 * 10s/30s UST + 0.25. R-squ: 92%, std. error: 8bp

Figure 76: …but the negative directionality of 10s/30s gilt curve vs. 1Yx1Y SONIA has weakened with the curve exhibiting stronger positive directionality with 5Yx5Y gilt yields

10s/30s par gilt curve regressed against: 1) 1Yx1Y SONIA, 2) 5Yx5Y par gilt yield, in separate single factor regressions; past 1M; bp

Source: J.P. Morgan.

On a micro basis, we note that the 10s/20s gilt curve looks too modestly too flat vs. the 20s/30s gilt curve ( Figure 77) and, given upcoming 20Y supply (4T43 tap on 19 March), alongside the general small RV flatness of 10s/30s, we enter tactical 10s/20s gilt curve steepeners (gilts Jan33/Jan44). However, this is very much a tactical view. In SONIA space, we keep our 5Yx5Y/15Yx15Y SONIA curve flattener as a low-beta bearish duration proxy, although we note that the RV dislocation has mostly corrected with the curve close to fair value vs. the level of 1Yx1Y SONIA ( Figure 78).

Figure 77: The 10s/20s gilt curve looks modestly too flat vs. 20s/30s gilt curve; we enter tactical 10s/20s gilt curve steepeners

10s/20s gilt curve (Jan33/Jan44) regressed against 20s/30s gilt curve (Jan44/Oct53); past 6M; bp

Source: J.P. Morgan.

Figure 78: We keep 5Yx5Y/15Yx15Y SONIA curve flatteners as a low-beta bearish duration hedge

5Yx5Y/15Yx15Y SONIA curve regressed against 1Yx1Y SONIA; past 6M; bp

Source: J.P. Morgan.

In RV space, we stay short the belly of the cash-and-duration neutral Jan32/Jan38/Oct41 gilt fly (risk weights 40%/-100%/60%, cash weights £27mn/-£47mn/£20mn). We also note that the implied forward yield between gilts Dec30 and Oct31 is at its richest levels since last December and looks 20bp rich on a regression basis vs. 5Yx5Y par gilt yields ( Figure 79). Given this, for investors looking to express a bearish duration view selling gilt Oct31 vs.Dec30 on an equational basis looks attractive.

Figure 79: For investors looking for a bearish duration RV proxy selling gilt Oct31 vs. Dec30 on an equational basis looks attractive given the relative richness of the implied forward yield vs. 5Yx5Y yields

Implied forward yield between gilts 4T30 and 4s31 regressed against 5Yx5Y par gilt yield; since 21 Feb; %

Source: J.P. Morgan.

Trade recommendations

New Trades

• Enter tactical 10s/20s gilt curve steepener (Jan33 vs Jan44 gilts)
  Buy £50mn Jan33 gilt and sell £30.5mn Jan44 gilt @47.5bp. 3M C+S is -2.6bp.

Closed Trades

• Close H4/U4 3M SONIA flattener
  Unwind short 250 H4 3M SONIA futures and long 250 U4 3M SONIA futures @-44bp. P/L since inception is -31bp (recommended 19 Jan)

Existing Trades

• Keep Aug24/Nov24 MPC OIS curve flatteners
  Pay £150mn Aug24 MPC OIS and receive £150mn Nov24 MPC OIS @-29.7bp. P/L since inception is +1.7bp (recommended  23 Feb)
• Keep 5Yx5Y/15Yx15Y SONIA curve flatteners
  Pay £100mn 5Yx5Y SONIA and receive £58.1mn 15Yx15Y SONIA@10.2bp. 3M slide is -2.7bp. P/L since inception is 0bp (recommended 1 Mar)
• Keep short belly Jan32/Jan38/Oct41 gilt fly (cash-and-duration neutral)
  Buy £34mn Jan32 gilt, sell £50.4mn Jan38 gilt and buy £32.9mn Oct41 gilt (risk weights 40%/-100%/60%, cash weights £27mn/-£47mn/£20mn) @6.24bp. 3M C+S is -0.6. P/L since inception is -1.5bp (recommended 1 Mar)
• Hold long Jun24 1Y SONIA midcurve call ladderBuy 1000 96.3/96.7/96.9 Jun24 SONIA 1Y midcurve call ladder at 3.3 cent with Jun25 SONIA futures at 96.1; P/L since inception is +2.3 cents (recommended 26 Jan)
• Keep short 1Yx1Y UK RPI
  Keep shorts in 1Yx1Y UK RPI @400bp. P/L since inception is -15bp (recommended 21 Feb)

*All existing trades priced as of 2:00pm UK 8 Mar 2024.

Trades closed over the past 12 months

In the long run we are all dead (and the Fed owns more T-bills)

• February employment picture raises the probability of a soft landing, but the pace of labor market rebalancing is a bit slower than we had expected. Rising growth expectations are supportive of higher yields
• Yields are in the middle of their recent ranges and valuations appear fair in the context of the market’s Fed policy, inflation, and growth expectations, indicating two-sided risks to yields...
• ...while duration and curve positioning is cleaner now as well, also indicating two-sided risk over the near term. We are neutral on duration and recommend using extremes in ranges to add or reduce duration
• Nevertheless, Fed on hold cyclicals suggest intermediates have room to outperform on the curve, and the belly is flagging as cheap: maintain 2s/5s flatteners and 5s/30s steepeners
• Powell indicated he supports the Fed’s Treasury holdings shortening in duration over time, but over the “longer term” and this is a 2025 event in our minds, should it come to fruition
• We review auction cyclical trading strategies. There’s limited evidence of long-end curve steepening ahead of long-end supply in recent months, but butterfly trades have generally been more consistently profitable. The profitability of auction cyclical trades has been small in recent months, particularly in the context of still-high delivered volatility
• Stripping activity totaled $5.6bn in February, slightly below recent averages. Demand remained concentrated in the long end. Overall, given the high pension funding status, we would expect stripping activity to remain consistently firm in 2024

Market views

Treasury yields declined 5-11bp over the week, with intermediates outperforming, similar to what we have observed in other Fed on hold periods. Three factors drove this rally. First, Chair Powell’s semiannual Congressional testimony did not differ greatly from his comments at the January FOMC press conference, indicating the strength of January labor market and inflation data hadn’t materially changed his view on the economy (see No change in Powell’s policy message, Michael Feroli, 3/6/24).  Second, February data are consistent with disinflationary trends, albeit at a slower pace than we had expected: nonfarm payrolls rose 275k in February (consensus: 200k), but this was accompanied by 167k in downward revisions to the prior two months. Moreover, the unemployment rate rose 0.2%-pts to 3.9% (consensus: 3.7%) and average hourly earnings rose a soft 0.1% (consensus: 0.2%, see also Strong job growth with a few blemishes, Michael Feroli, 3/8/24). Finally, non-US monetary policy developments were dovish: the ECB stayed on hold, but changes in the staff forecasts were more dovish than we had expected, mainly because core inflation for 2025 was cut by 0.2%-pts to 2.1%oya, rather than by just 0.1%-pt (see ECB: June is on (if the data don’t get in the way), Greg Fuzesi, 3/7/24). With these moves, Treasuries have retreated squarely into the middle of the ranges they’ve held the past three months, and range volatility has drifted to its lowest levels since early-2022, before the Fed embarked on its torrid tightening campaign ( Figure 80 & Figure 81).

Figure 80: Intermediate yields have declined to their lowest levels in a month, and have retraced back to the middle of their 3-month range

5-year Treasury yields with 3-month moving average; %

Source: J.P. Morgan

Figure 81: Range volatility has collapsed to two-year lows

5-year Treasury rolling 3-month range volatility; bp

Source: J.P. Morgan

Overall, today’s data seemed to alleviate some of the market’s concerns that the economy could be accelerating and likely reinforce the soft landing message, justifying a Fed that eases by midyear: OIS forwards are now pricing greater than 80% probability of a 25bp cut by the June FOMC meeting and more than 90bp of eases this year ( Figure 82). However even with this data, the pace of job growth makes it challenging for us to think our forecasted 50k average payroll growth in 2Q24 will be realized. Accordingly, we have revised our 2Q24 real GDP forecast from 0.5% to 1.5%, and our first half projections now look firmly above trend. In aggregate, we’ve now raised our year-ahead growth forecasts by 0.8%-pts over the last 2 months, which helps to anchor Treasury yields at higher levels all else equal ( Figure 83).

Figure 82: This week’s data increase confidence the Fed could ease by midyear, albeit at a more pedestrian pace...

Expected change in Fed funds effective rate by FOMC meeting implied by OIS forwards, 3/1/24 vs. 3/8/24; bp

Source: J.P. Morgan

Figure 83: ...as the serial upward revisions to growth expectations continue

US GDP Forecast Revision Index;

Source: J.P. Morgan

Looking ahead, there’s significant event risk on the horizon, with CPI due Tuesday and retail sales on Thursday. We look for a 0.28% rise in core CPI, a significant moderation from last month’s hot print, with a softening in OER offset by a moderation in goods deflation. Meanwhile, we also forecast control group retail sales rose 0.6%-pt in February (consensus: 0.4%), fully reversing last month’s weakness (see Economics). Combining these factors, the data next week have mixed implications for the Fed outlook, and do nothing to suggest Treasury yields will dislodge from the ranges they have held the last three months. Moreover, valuations give us little advantage either: 10-year Treasuries are fairly valued in the context of the market’s medium-term Fed policy, inflation, and growth expectations. Accordingly, fundamentals and valuations indicate Treasury yields are likely to remain rangebound over the near term.

Meanwhile, position technicals are largely neutral and do not suggest the risks to yields are one sided in the coming weeks. At the beginning of February, our macro hedge fund beta to 10-year yields indicated that levered positioning for lower yields was more than 2 standard deviations above the prior 1-year average, and that actively-managed bond funds were also leaning somewhat long duration ( Figure 84). Walk this data forward to today, and macro funds and asset managers have reduced these longs, and positioning is now modestly shorter than where it’s been, on average, over the last year. Interestingly this trend extends to our other indicators, though the CFTC’s weekly Commitment of Traders data shows that speculative positioning in interest rate futures is more than two standard deviations above it’s medium-term average. At face value, this could indicate that hedge fund positioning is still long, but we think this move in futures positioning reflects relative value funds paring back on Treasury cash/futures basis longs, which we can also observe via the ongoing buildup in dealer inventory in 3- to 6- and 7- to 11-year Treasuries.

Figure 84: Macro hedge funds and bond funds have pared back their duration overweights over the last five weeks...

Current value of various position indicators* versus 2/1/24; 1-year z-score

Source: CFTC, Bloomberg Finance L.P, SG, HFR, J.P. Morgan* JPM Client Survey refers to a 4-week moving average of our Treasury Client Survey Index; (Longs+Neutrals)/(Shorts+Neutrals), see Survey Says: Using the Treasury Client Survey to predict rates moves, 7/21/23 for more details. CFTC refers to the non-commercial net longs in UST and ED futures contracts reported by the CFTC. CTA beta is the four-week partial beta of SG CTA Index to 10-year UST yields. Real money beta is the eight-week partial beta of excess returns of the 20 largest actively managed US core bond funds to 10-year UST yields. Macro HF beta is the six-week partial beta of HFRX Macro/CTA Index to 10-year UST yields

Figure 85: ...And bond funds have substantially reduced their curve steepener exposure

Partial beta with respect to 5s/30s Treasury curve in our active bond fund model*;

Source: J.P. Morgan, Bloomberg Finance L.P.
* Model is a 3-month regression of daily excess returns on the bond fund index versus daily changes in 1) 10-year US Treasury yields, 2) 5s/30s Treasury curve, 3) JULI spread to Treasury, 4) MBS Production coupon Treasury OAS, and 5) 3Mx10Y swaption volatility

This is evident in other indicators: Figure 85 shows actively-managed core bond fund exposure to a steeper 5s/30s curve peaked in early-January, but has been falling since, and this beta has declined to its lowest level since the early fall, when rising term premium was at the forefront of market focus. More broadly speaking, the reduction in duration and curve exposure is reflective of reduced risk appetite in an uncertain world: as we’ve highlighted in recent weeks, the share of neutrals in our weekly Treasury Client Survey has been rising. Figure 86 shows that the share has declined modestly from its historic peak in mid-February, but remains in the 98th percentile of observations over the last quarter century.

Figure 86: The share of neutrals in our weekly survey sit in the 98^th percentile over the last 25 years

Share of neutrals in J.P. Morgan Treasury Client Survey; %

Source: J.P. Morgan

In sum, when we look holistically at the Treasury market, the data flow have mixed implications for yields over the coming week, valuations are fair, and position technicals are neutral. Hence, we have little conviction on duration here and recommend using the extremes in ranges to either add or reduce duration. That being said, with the Fed on a path to ease by mid-year, we continue to recommend trades which tend to outperform in the periods between the final hike of a cycle and the first ease: on average the 5-year sector outperforms along the curve (see Treasuries 2024 Outlook, 11/21/23). Furthermore, most front-end curve pairs remain steep and long end pairs remain flat after controlling for Fed policy, and inflation expectations, as well as the Fed’s ownership share of the Treasury market, reinforcing the value in the 5-year sector along the curve (see Treasuries, US Fixed Income Markets Weekly, 3/1/24). Thus, we continue to recommend 5s/30s steepeners, as a longer-term expression of our Fed view, while also holding 2s/5s flatteners to add carry, knowing this portion of the curve has a tendency to steepen only once the first Fed ease is nearly imminent.

In the long run we are all dead (and the Fed owns more T-bills)

Governor Waller took markets by surprise last week, raising two “longer-term issues” on the Fed’s balance sheet strategy. First, to no surprise, he said he would like to see the Fed’s MBS holdings fall to zero over time. Second, and this is a newer development, he “would like to see a shift in Treasury holdings toward a larger share of shorter-dated Treasury securities.” As we discussed last week, we were sympathetic to this, given the Fed’s Treasury holdings are substantially longer in duration than the overall Treasury market. However, we also argued that Fed leadership supported this notion in 2019 as QT came to a conclusion, but still chose to reinvest MBS paydowns into Treasuries proportionally across the curve (See Treasuries, US Fixed Income Markets Weekly, 3/1/24). This led us to believe the bar was still high for the Fed to focus reinvestments in T-bills, even though there’s sound logic in this thesis.

Chair Powell was asked about the balance sheet in his Q&A with the Senate Banking Committee yesterday, and his comments were strikingly similar to Waller’s. Powell argued “Over time, we’d love not to own a lot of MBS and I could see a case for shortening the maturity but it’s not something that would happen quickly...That’s sort of a longer-term aspiration.” The similarity of these comments indicates the risk of the Fed shortening the duration of its Treasury holdings is higher than previously appreciated, but still not our modal view, considering the Fed’s modus operandi 5 years ago. However, we do not think any change is imminent: both speakers framed such a move as being longer-term in nature. Moreover, we do not expect the Fed to announce it will slow the pace of QT until June, and envision the process running through year-end. Thus, the Fed will not begin actively reinvesting its MBS paydowns to keep its balance sheet stable until early-2025. Further, to the extent that this is worth approximately $180bn in secondary demand for Treasuries annually from the Fed, and alongside ~$100bn in annual demand for T-bills to keep reserves ample, it would take years to reweight the SOMA in line with the overall Treasury market duration. Thus, while we look forward to hearing more on this topic in the minutes of the March FOMC meeting, it’s very difficult to trade this potential policy decision while it’s unlikely to be implemented until next year.

Treasury auction cyclicals update

Since August, Treasury has been making increases to coupon auction sizes to contend with large federal budget deficits, as well as the ongoing SOMA runoff. While Treasury slowed the pace of increases in longer-duration auctions last November, the intra-month peak in duration supply remains concentrated around the middle of the second week of the month, corresponding with mid-month supply and the 30-year bond auction in particular ( Figure 87). Notably, 47% of the monthly Treasury duration supply remains focused in the second week of the month, unchaged from two years ago, while the remainder is distributed over the balance of the month.

Figure 87: Duration supply peaks in the second week of the month, with 10- and 30-year auctions

Duration supply by business day of the month, average between 2022-2023; $bn of 10-year Treasury equivalents

Source: US Treasury, J.P. Morgan

Given the growth in long-duration Treasury issuance since last summer, we believe it is appropriate to revisit auction cyclicals and evaluate how they have evolved in recent months. Figure 88 summarizes the performance of various duration- and beta-weighted curve and butterfly trades over the 5- and 10-days prior to and following 30-year auctions, averaged over the last 6 months. Given the concentration in duration supply discussed earlier, it would make intuitive sense to see a larger concession into the concentration of long-duration auctions, culminating with the 30-year bond. Accordingly, we would expect long-end curve steepeners and belly-richening butterflies to be profitable leading into the 30-year auction, while flatteners and belly-cheapening butterflies to be profitable post-auction.

However, it is notable that there’s limited evidence of long-end curve steepening in the period 1- to 2-weeks before the 30-year bond auction, though the performance looks somewhat better when beta-weighting these trades to take out the directionality with respect to rate levels. Interestingly, butterfly trades have generally been more consistently profitable than curve trades during this period. Further, we find that beta-weighted flies with the belly anchored in the 5- or 7-year point exhibit some of the most consistent results, richening ahead of and cheapening following the 30-year auction. Overall, it is notable that the profitability of these auction cyclical trades has been small in recent months, particularly in the context of still-high delivered volatility in the Treasury market.

At a more granular level, we find that beta-weighted belly-cheapening 3s/5s/30s butterflies initiated on the day of the 30-year auction and unwound 10 days later have been profitable in five out of the last six months, cheapening by 1.9bp on average ( Figure 89). A closer inspection shows this cheapening was disproportionately concentrated in the 3-day period following the auction, when this fly cheapened by an average 1.2bp. Thus, for investors looking at auction cyclicals, we recommend initiating beta-weighted 3s/5s/30s belly-cheapening butterflies following the 30-year auction and unwinding 5 to 10 business days later.

Figure 89: 5s have tended to cheapen on the 3s/5s/30s fly in the two weeks following the 30-year auction

Cumulative changes in beta-weighted* 3s/5s/30s butterflies in the 10 days following 30-year auctions; bp

Source: J.P. Morgan
*Weights are derived from 3-month level and curve betas, by regressing the 50:50 3s/5s/30s butterfly spread on the 5-year leg and the 3s/30s curve

STRIPS update

Earlier this week, Treasury released the MSPD for February, which showed P-STRIPS outstanding rose $5.6bn over the month, slightly below recent averages ( Figure 90) Along the curve, demand remained concentrated in the long end, with $6.6bn in stripping activity in the 28- to 30-year sector, $0.5bn below the previous three-month average ( Figure 91). Indeed, the most stripped securities during the month were the current and the old 30-year bond, at $1.4bn each ( Figure 92). Overall, the pace of stripping demand remains robust in the first two months of the year. Looking ahead, an environment in which DB pension funded ratios remain consistently above 100% supports steady demand at the long end.

Figure 90: P-STRIPS outstanding rose $5.6bn in February, near recent averages...

Monthly change in P-STRIPS outstanding with prior 3-month moving average; $bn

Source:  US Treasury, J.P. Morgan

Figure 91: ...and demand remained concentrated in the long end

Monthly change in P-STRIPS outstanding by sector in February 2024 vs. prior 3-month average; $bn

Source:  US Treasury, J.P. Morgan

Trade recommendations

• Maintain 5s/30s steepeners

-   Stay long 100% risk, or $112mn notional of T 4.875% Oct-28s
-   100% risk, or $29.9mn notional of T 4.75% Nov-53s
- (US Treasury Market Daily, 11/22/23: P/L since inception: -0.6bp)

• Maintain 75%/6% weighted 5s/10s/30s belly-cheapening butterflies

-   Stay long 75% risk, or $43mn notional of T 4.625% Sep-28s
- Stay short 100% risk, or $33.3mn notional of T 3.875% Aug-33s
- Stay long 6% risk, or $1mn notional of T 4.125% Aug-53s
- (US Fixed Income Markets Weekly, 9/29/23: P/L since inception: -8.2bp)

• Maintain 2s/5s flatteners

-  Stay short 100% risk, or $135mn notional of T 4.875% Nov-25s
-  Stay long 100% risk, or $56.5mn notional of T 4.375% Nov-28s
- (US Fixed Income Markets Weekly, 12/8/23: P/L since inception: 5.0bp)

• Unwound 5-year duration longs

-   Unwind long 100% risk, or $113mn notional of T 4% Jan-29s
-(US Fixed Income Markets Weekly, 2/9/24. Unwound in US Treasury Market Daily 3/7/24. P/L: 3.3bp)

• Maintain 20s/ old 30s flatteners

-  Sell 100% risk, or $19mn notional of T 4.75% Nov-43s
-  Buy 100% risk, or $14.6mn notional of T 4.75% Nov-53s
- (US Treasury Market Daily, 2/15/24: P/L since inception: 0.1bp)

• Maintain 100:95 weighted 4% Feb-28 / 4% Feb-30 steepeners

-   Stay short 95% risk, or $46mn notional of T 4% Feb-30s
- Stay long 100% risk, or $69.6mn notional of T 4% Feb-28s
-(US Fixed Income Markets Weekly, 2/23/24: P/L since inception: 0.4bp)

The sun is the same, in a relative way, but vol is lower

• Central banks across both sides of the pond are looking for more data to corroborate disinflationary trends, which is helping to preserve policy uncertainty at elevated levels. At the same time, other factors have become less significant and policy expectations are paramount in terms of driving yields. This has caused yields to be volatile on a day-to-day basis, but also mean-reverting and ultimately rangebound over longer time intervals
• In mean reverting markets, volatility is in the eye of the delta-hedger. There is significant volatility for a daily (or even more frequent) delta hedger to monetize, while there is significant carry to be earned by an infrequent delta hedger who encounters far less volatility. Historically, implieds get pressured lower in such periods, representing a significant headwind for long volatility positions …
• … we turn neutral on volatility in most sectors, except on 30-year tails where we are comfortable recommending a less-frequently delta hedged short vol position - sell 3Yx30Y swaption volatility
• Rangebound yields continue to lead us to favor carry trades, but we prefer well hedged trades that improve the carry-to-risk ratio …
• … initiate 3M forward 3s/20s swap curve steepeners versus 15Mx3M / 3Mx10Y swap curve flatteners
• Lower yields have caused the WNM4 CTD to shift to the high coupon Nov 53s, which has much lower wildcard option potential. But the CTD basis remains wide net of carry, making selling the basis attractive as a bullish trade …
• … we recommend selling the 4.75% Nov 53 WNM4 factor weighted basis, and buying a weighted amount of A+20bp payer swaptions, to monetize the richness of the basis
• We remain neutral on swap spreads in most sectors, but now favor widening exposure in the 10Y sector where spreads appear narrow to fair value. Outright spread wideners, or weighted 5s/10s swap spread curve steepeners are attractive ways to express such a view

The sun is the same, in a relative way, but vol is lower

Monetary policy was the main focus on both sides of the Atlantic this week. On the central bank front, this week brought forth Chair Powell's testimony to Congress in the US, and the ECB meeting in Europe. Chair Powell largely reiterated his comments after the January FOMC meeting, suggesting that more data would help increase confidence in disinflationary trends but also that such confidence was likely not far away. This appears to point to June as the likely timing for a rate cut from the Fed. The ECB's message was similar - after holding rates steady on Thursday, the ECB's signaling suggests to our economists that a June cut is on the table if data doesn’t get in the way (see ECB: June is on (if the data don’t get in the way), G. Fuzesi, 3/7/2024). Data came in on the slightly softer side this week such as the ISM data, although our economists think that some key details actually firmed up (see US: Services ISM declines, but some key details firm, D. Silver, 3/5/2024). This softness also extended to Friday's employment report - although Feb Nonfarm Payrolls surprised to the upside at 275K, this was more than offset by downward revisions to the previous print. In addition, the unemployment rate ticked up to 3.9% and Average Hourly Earnings was softer as well. All in all, our economists are of the view that the employment report is indicative of a labor market that is coming back into balance, and is supportive of a June timing for the first ease (see United States, M. Feroli, 3/8/2024).

Thanks to this week's softness in the data, yields are lower on the week across the curve, with the belly outperforming slightly as seen in Figure 1. But the broader picture remains one of data dependency and policy uncertainty. As we have been highlighting in recent weeks, options markets (which give us distributional information, as opposed to forward rates which merely tell us averages) are implying distributions for future policy rates that can be thought of as a weighted combination of very different scenarios. As recently as mid-week, the implied distribution derived from the prices of various calls and puts on SFRZ4 could be decomposed into a blend of scenarios ranging from 5 cuts by year end to one more hike by year end (Figure 2). Moreover, the weights on these scenarios were not too dissimilar, suggesting that policy uncertainty remains considerably high. This is the case even after Friday's jobs report - as seen in the exhibit, the current implied probability density function for SFRZ4 continues to combine scenarios that include 5 cuts, 2 cuts and 1 hike for the most part. This suggests that the policy outlook remains highly uncertain, and the cone of future policy paths is wide enough to span additional hikes as well as moderate cuts by year end. Thus, markets will likely continue to remain highly data dependent, especially on data related to labor markets and inflation.

Figure 1: Yields are lower on the week, with the belly outperforming

Selected statistics for UST yields, dates indicated; %

Source: J.P. Morgan.

Figure 2: Policy uncertainty remains considerably high even after Friday’s employment report, with options markets assigning significant weights to scenarios involving 5 cuts, 2 cuts as well as 1 hike by year end

Weights for conditional Normal distributions* that can be combined to create a composite distribution that recovers prices of SFRZ4 calls and puts at various strikes**, scenarios with largest weights, the weight differential between the two scenarios with the largest weights, and 12/31/2024x1M OIS rate (%), as of the dates shown***

Source: J.P. Morgan., CME
*1 hike scenario corresponds to mean of 5.5%, unchanged corresponds to a mean of 5.25%, 1 cut corresponds to mean of 5%, 2 cuts corresponds to mean of 4.75%, 3 cuts corresponds to mean of 4.5%, 5 cuts corresponds to mean of 4% and 9 cuts corresponds to mean of 3%
**Implied distribution is assumed to take the form of a weighted sum of many individual Gaussian density functions with various different means and standard deviations. The weights are solved for by calibrating to the prices of 3M SOFR ATM and OTM calls and puts, while also recovering the underlying futures price*** 3/8 values represent live levels as of 9:30 am ET

Not only is policy uncertainty high, but the relative importance of policy expectations as a driver of yields across the curve is as high as it has been in recent years. One way to see this is by examining the percentage of total variation in yields that is accounted for by the first principal component. We have previously described an approach to inferring this on a forward-looking basis - i.e., we combine information from the swaptions and YCSO markets to construct implied principal components (see i-PCA: Implied Principal Component Analysis). Among other things, this methodology allows us to examine the percentage of total variance in yields that is expected to be explained by the first implied principal component. As seen in Figure 3, this is now back to highs last seen in the early stages of the hiking cycle. In other words, policy expectations remain by far the most significant driver of yields across the curve as highlighted by the dominant nature of the first implied principal component.

The combination of these two facts - the elevated significance of policy expectations and the elevated level of policy uncertainty - is leading to some interesting consequences. As one might expect, in a world where policy expectations are paramount but policy uncertainty is also elevated and persistent at the same time, yields would likely exhibit choppy behavior day-to-day but remain ultimately rangebound. This is highlighted in spectacular fashion in Figure 1, where we can see that 1M and 3M changes in yields have been remarkably small in magnitude. In other words, this creates conditions where yields are volatile over shorter sampling intervals, but mean reverting and less volatile over longer intervals. This has implications for volatility, since long as well as short gamma positions can be simultaneously attractive to different investors with different risk tolerances and delta-hedging frequencies (see Options discussion below). But rangebound yields also imply that carry trades should remain attractive, although the volatility over shorter time intervals means that such trades will need to be well hedged and properly constructed in order to better isolate carry exposure relative to macro curve volatility. This remains a key theme for us in trading the swap yield curve.

We have previously (even somewhat repeatedly) highlighted the attractiveness of yield curve box trades that mitigate macro curve risk while isolating carry and relative value. One such trade that is currently attractive is to initiate 3M forward 3s/20s swap curve steepeners, paired with 85% of the risk in a SFRM5 / 3Mx10Y curve flattener. As seen in Figure 4, the 3s/20s curve has been well correlated to the 15Mx3M / 3Mx10Y curve, with a beta of ~85%. Moreover, the 3M slide on this weighted box trade is ~5bp, while the 2Y standard deviation of 3M changes in the weighted box spread is about 20bp (annualized), making the ratio of annualized carry to annualized risk ~1. Therefore, we recommend this trade (see Trade recommendations).

Figure 3: Policy expectations are by far the most significant factor driving yields across the curve, as highlighted by the dominant nature of the first implied principal component

Percent variation explained by the first implied principal component*, 3/1/2022 - 3/7/2024

Source: J.P. Morgan.
*Calculated as the ratio of the eigenvalue corresponding to the first implied principal component to the sum of all eigenvalues, for a detailed description of our implied PCA framework, see i-PCA: Implied Principal Component Analysis

Figure 4: The 3s/20s curve has been well correlated to the 15Mx3M / 3Mx10Y curve, with a beta of ~0.85

The 3M forward 3s/20s swap yield curve (%), regressed against the 15Mx3M / 3Mx10Y forward swap yield curve (%); 3/1/2023 - 3/7/2024

Source: J.P. Morgan.

Treasury Futures Basis

On a relative value basis, we now recommend selling the Nov '53 basis into WNM4, versus buying payer swaptions. Thanks to the recent decline in yields, the CTD into June ultra-long bond futures has shifted to the 4.75% Nov '53s, which, despite its longer maturity, has a lower factor-adjusted bpv causing it to become CTD in a rally (see Soft landings, TouchdoWNs, and Safety in the End Zone, 2/9/2024). Further CTD shifts are unlikely in a rally, while the CTD can shift towards the 2.25% Aug 49s in a selloff.

One interesting consequence of this CTD shift to the Nov '53s is that the wildcard option value is much lower. Historically, for several years now, the WN sector has been characterized by CTD bonds with low coupons and conversion factors, which has caused wildcard option values to be significant. But thanks to the relatively high conversion factor of the Nov '53s, wildcard option value is likely much lower so long as the Nov '53s remain CTD. However, the net basis of the CTD is at ~15/32nds, which we estimate is significantly wide relative to traditional CTD switch optionality. In other words, markets are continuing to price in significant wildcard optionality in the WNM4 contract.

This basis likely remains wide because of the possibility of the CTD shifting to a lower coupon, thereby resurrecting wildcard option value. But that is only likely in a selloff, meaning that selling the WNM4 Nov '53 basis is especially attractive as a bullish trade. Thus, it is desirable to consider packages that combine selling the Nov '53 WNM4 factor-weighted basis with a suitable hedge in a selloff. 

We estimate that a $100mn (bond notional) long factor-weighted basis position in the Nov '53 basis can be replicated by using ~$11mn notional in a A+20bp payer swaption (5/30/2024 expiry, swap start 6/3/2024, swap maturity 11/15/2053) . As seen in Figures 5 and 6, the replication returns are projected to track long basis returns fairly closely with a beta of 1, but with the replication strategy outperforming (as evidenced by the negative intercept in Figure 6). Therefore, we recommend selling the 4.75% Nov 2053 WNM4 factor weighted basis, and buying 6/3/2024x11/15/2053 A+20bp payer swaptions to take advantage of the likely cheapening of basis in a rally while also mitigating exposure in a selloff (see Trade recommendations).

Figure 5: A long WNM4 basis exposure is projected to be well replicated using A+20bp payer swaptions ...

Estimated P/L on the 4.75% Nov ‘53 WNM4 basis*, versus projected P/L on a replication strategy using A+20 5/30/2024 x 11/15/2053 payer swaptions (swap start 6/3/2024); 32nds

Source: J.P. Morgan.
* Calculated as projected basis net of carry at 6/3/2024 horizon minus current net basis (as of 3/6/2024 COB).
** Replication assumes $11mn of A+20bp 5/30/2024 x 11/15/2053 payer swaptions (swap start 6/3/2024)

Figure 6: … with a projected beta of one, and with an intercept that points to outperformance of the replication

4.75% Nov ‘53 estimated basis P/L versus projected P/L from the replication strategy*, ticks , 3/6/2024; 32nds

Source: J.P. Morgan.
*Replication strategy P/L and basis P/L as described in the previous Figure

Swap spreads

Swap spreads were relatively flat on the week, with spreads modestly narrower in the front end and modestly wider in 7s and 20s (Figure 7).

Figure 7: Swap spreads were mostly flat on the week

Selected statistics for maturity matched swap spreads, 3/1 - 3/8; bp

Source: J.P. Morgan.

Looking ahead, the past month has seen unusually heavy high grade corporate issuance (Figure 8), but this could slow going forward (our high grade credit strategists expect $130bn in March, versus nearly twice that pace in the past month). Since some fraction of such debt issuance tends to be swapped to floating, a slowing in the pace of corporate supply should lessen narrowing pressure on swap spreads. But this will likely be felt further out the curve this time. Thanks to the fact that swap spreads in the 2- to 5-year sector have widened for somewhat idiosyncratic reasons (because of a correction in the 5-year sector, and Fed-speak related to shorter maturity reinvestments by the Fed in the 2-year sector), spreads look fair or even wide in the front end, but significantly narrow to fair value in the 10-year sector (Figure 9).

Figure 8: The past month has seen unusually heavy high grade corporate debt issuance

Monthly high-grade issuance, 3/1/2018 - 3/7/2024; $bn

Source: J.P. Morgan.

Figure 9: Swap spreads in the 10-year sector appear most narrow relative to fair value

Maturity match swap spreads versus fair value for select sectors* (left, bp) and residual** from FV model (right, bp)

Source: J.P. Morgan.
* For a detailed description of our 2Y swap spread Fair Value model, see Governer Vol-ler moves the market, 3/1/2024. A detailed discussion of our Fair Value model for 5Y, 10Y, and 30Y swap spreads can be found in our Interest Rate Derivatives 2024 Outlook.
** Residual is defined as actual value minus fair value

On an outright basis, this leaves us neutral on 5- and 30-year swap spreads (where spreads appear fair) and in the 2-year sector (where spreads appear wide to fair value, leaving us waiting for better entry levels into wideners ahead of upcoming negative T-bill issuance in 2Q24). But we do recommend positioning for wider swap spreads in the 10-year sector, where valuations are significantly attractive and recommend paying in 4.5% Nov 2033 maturity matched swap spreads (see Trade recommendations). Another attractive way to initiate exposure to wider spreads in the 10-year sector is via 5s/10s swap spread curve steepeners. As seen in Figure 10, the 5s/10s maturity matched swap spread curve has generally been well correlated to the level of spreads with a beta of ~40%, but now appears too flat. Therefore, we recommend initiating exposure to 5s/10s 100:60 weighted swap spread curve steepeners (see Trade recommendations).

Figure 10: The 5s/10s swap spread curve has been well correlated to the level of spreads but now appears too flat

The 5s/10s swap spread curve (bp), regressed against 10Y maturity matched swap spreads (bp); 3/1/2023 - 3/7/2024

Source: J.P. Morgan.

Options

We started the week expecting high volatility around multiple key events across both sides of the pond. However, with the key messaging from Chair Powell and ECB broadly similar, and in line with previous messaging, and Payrolls showing some signs of softening, realized volatility was indeed elevated but in a manner that leaves yields ultimately rangebound as we noted earlier. History tells us that rangebound yield environments draw in gamma selling from investors who have the flexibility to delta hedge less frequently, which tends to pressure implied volatility lower. Such has been the case this week - implieds are lower across the grid led by the upper left (Figure 11), which has been painful for our long gamma positions. 

Figure 11: Implieds are lower across the surface with shorter expiries underperforming longer expiries

Selected statistics for various SOFR swaptions structures, 3/1 - 3/8; bp/day

Source: J.P. Morgan.

Looking ahead, we turn neutral on volatility in most sectors, with the exception of 30-year tails, where we turn bearish. One, we are cognizant of the fact that much of the near term event risk is now behind us (and CPI almost behind us). This means that next set of data releases that could cause significant shift in the weights in implied distributions (and thus cause forwards to move significantly) is likely not imminent. Thus, we could be heading into a period of lower realized volatility, even if it is just a brief one.

Two, policy uncertainty has yet another implication as we noted earlier in this piece. Although yields have been choppy and volatile on a day-to-day basis, they have also been rangebound over longer horizons as a clear policy path has yet to emerge. As a result, realized volatility has been much lower when sampled less frequently than daily - for example the standard deviation of rolling 1-day changes in 10-year swap yields is around 7 bp/day, but the standard deviation of rolling 1-month changes is a mere 3 bp/day currently (Figure 12). This is because without emergent policy clarity yields have simply tended to mean-revert rather than trend.

Indeed, we can define a measure of mean reversion simply by examining how measured realized volatility declines with sampling interval size. Specifically, we define our mean reversion metric as the slope of a linear fit of realized volatility (measured using 1D, 1W, 2W and 1M sampling intervals) versus the length of that sampling interval (1, 5, 10 and 20 business days respectively). As Figure 13 shows, this mean-reversion index has increased to near its 9-month peak levels. This means that while an active delta hedger can monetize a lot of realized volatility, there is also carry to be earned from selling volatility and delta hedging less frequently. This level of mean-reversion has likely acted as a headwind for long volatility positions by pressuring implied volatility lower. Moreover, such increased mean reversion appears here to stay, until policy clarity emerges down the road.

Figure 12: A tree falling in the forest only makes a sound if you look - realized volatility has been elevated for a daily delta-hedger who samples yields every day, but is muted for the less frequent delta hedger who samples rates less frequently …

Realized volatility on 6Mx10Y swap yields over a rolling 3-monthperiod, calculated using 1-day changes and using 20-business day changes, past 1 year; bp/day

Source: J.P. Morgan.

Figure 13: … which is of course indicative of elevated mean reversion

Index of mean reversion, calculated each day as -1 times the slope of a fit of realized volatility based on 4 different sampling intervals*, versus the length of those intervals, past 9 months

Source: J.P. Morgan.
*The four intervals used are 1D, 1W, 2W and 1M. On each day, rolling 3-mmonths of data are used to calculate each of the 4 realized volatilities, which are then regressed against interval length (1, 5, 10 and 20 respectively) to calculate the slope of the fit.

One sector where we are comfortable in recommending a less-frequently delta-hedged short swaption straddle position is in 30-year tails. As Figure 14 shows, implied volatility appears cheap to fair value in most tails, with the exception of 30-year tails where it appears rich to fair value. Thus, in addition to the anticipation of continuing mean reversion, the case for selling volatility in 30-year tails benefits additionally from rich valuations, especially in intermediate expiries. Therefore, we recommend selling 3Y expiry swaptions on 30-year tails, delta hedged less frequently than daily (see Trade recommendations)

Figure 14: Implieds continue to appear mostly cheap, with the exception of 30-year tails

Residual from our fair value model for implied volatility*, 3/7/2024 ; bp/day

Source: J.P. Morgan.
* Fair value for implied volatility is calculated as per the model detailed in Figure 36 of our Interest Rate Derivatives 2024 Outlook

Trading Recommendations

• Initiate 3M forward 3s/20s swap curve steepeners, paired with 85% of the risk in a SFRM5 / 3Mx10Y curve flattener
  Yield curve box trades that mitigate macro curve risk while isolating carry and relative value is an attractive way to trade the yield curve. The 3s/20s curve has been well correlated to the 15Mx3M / 3Mx10Y curve, with a beta of ~85%. The 3M slide on this weighted box trade is ~5bp, while the 2Y standard deviation of 3M changes in the weighted box spread is about 20bp (annualized), making the ratio of annualized carry to annualized risk attractive at about 1.
  -Receive-fixed in $100mn notional of a 06/08/24x3Y SOFR swap at a yield of 3.911% (PVBP: $287.7/bp per mn notional). Pay-fixed in $20.8mn notional of a 06/08/24x20Y SOFR swap at a yield of 3.671% (PVBP: $1386.5/bp per mn notional).
  -Sell 978 contracts of the SFRM5 at 96.205. Receive-fixed in $29.9mn notional of a 06/08/24x10Y SOFR swap at a yield of 3.665% (PVBP: $817.7/bp per mn notional).
  -This trade uses risk weights of 1.0/-1.0/-0.85/0.85 on the 3Mx3Y/3Mx20Y/SFRM5/3Mx10Y swaps respectively.
  
• Initiate 10Y swap spread wideners using the Nov ‘33 issue
  Swap spreads in the 10Y sector are significantly narrow to fair value and we recommend initiating swap spread wideners. Moreover, the off-the-run Nov ‘33 issue slides wider over time.
  -Pay fixed in 4.5% Nov 15 2033 maturity matched SOFR swap spreads. Buy $100mn notional of the 4.5% Nov 15 2033 (yield: 4.098%, PVBP: $807.4/bp per mn notional), and pay fixed in $100.3mn notional of a maturity matched SOFR swap (coupon: 3.718%, PVBP: $805.3/bp per mn notional) at a swap spread of -38.0bp.
• Initiate 5s/10s off-the-run swap spread curve steepeners (100:60 weighted)
  On top of attractive valuations in the 10Y sector, the 5s/10s swap spread curve has been well correlated to the level of spreads but now appears too flat. Moreover, we use the off-the-run issues for both the 5Y leg using the Jan ‘29 issue as well as the 10Y leg using the Nov ‘33 issue, each of which slide favorably over time.
  -Receive fixed in 4% Jan 31 2029 maturity matched SOFR swap spreads. Sell $100mn notional of the 4% Jan 31 2029 (yield: 4.076%, PVBP: $438.6/bp per mn notional), and receive fixed in $97.6mn notional of a maturity matched SOFR swap (coupon: 3.832%, PVBP: $449.5/bp per mn notional) at a swap spread of -24.4bp.
  -Pay fixed in 4.5% Nov 15 2033 maturity matched SOFR swap spreads. Buy $32.6mn notional of the 4.5% Nov 15 2033 (yield: 4.098%, PVBP: $807.4/bp per mn notional), and pay fixed in $32.7mn notional of a maturity matched SOFR swap (coupon: 3.715%, PVBP: $805.3/bp per mn notional) at a swap spread of -38.3bp.
  
• Sell the 4.75% Nov 2053 WNM4 basis, versus buying payer swaptions
  Lower yields have caused the CTD to switch into the high coupon Nov 53s, which has a much higher factor and lower wildcard option value. But the basis continues to price in significant wildcard optionality, which is unlikely to materialize except in a selloff that causes the CTD to shift back to the Aug 49s, which could happen if yields rise. This makes it attractive to sell the June WN CTD basis, while hedging the risk of higher yields and its nonlinear impact on the basis.
  -Sell $50mn notional 4.75% Nov 53 WNM4 factor weighted basis at 6.25/32nds (sell $50mn notion of the bond versus buying 414 WNM4, at a gross factor weighted basis of 6.25/32nds).
  -Buy $6mn notional payer swaptions (notification 5/30/2024, swap start 6/3/2024, maturity 11/15/2053, ATMF 3.505%, strike 3.705%, premium 157.5bp of notional). This premium is equivalent to about 6/32nds in terms of the basis. This hedge ratio is larger than our estimate in the analysis, in order to account for other sources or risk such as swap spread movements or curve moves.
  -Term out the repo to 6/3/2024 at 5.33%. This results in -8.5/32nds of carry, and a net basis of 14.75/32nds.
  
• Initiate outright shorts in 3Yx30Y swaption implied volatility, but delta hedge monthly or if rates move by over 25bp in either direction since the last delta hedge
  Event risk has come to pass and yields have been mostly rangebound over longer horizons, which is beneficial to sell volatility and delta hedge less frequently in order to earn carry. Moreover, implied volatility appears cheap to fair value in most tails, with the exception of 30-year tails where it appears rich to fair value.
  -Sell $100mn notional 3Yx30Y ATMF swaption straddles. (Notification date: 2027-03-08, swap tenor: 30Y, ATMF: 3.358%, strike: 3.358%, spot premium: 1963.1bp per notional, forward premium: 2217.8bp per notional, bpvol at inception: 5.47bp/day). This trade assumes active delta hedging, once every 21 business days, or if rates move by over 25bp in either direction since the last delta hedge
• Unwind longs in 6Mx10Y swaption straddles
  We turn neutral on gamma and recommend unwinding this trade at a loss of 6.6abp. For original trade write up, see Fixed Income Markets Weekly 2024-03-01.
• Unwind exposure to rising term premium by selling the belly of a 35/65 weighted 3M forward 5s/10s/15s butterfly
  This trade has reached our self imposed horizon of 3 months for trades of this nature and we recommend unwinding this trade at a loss of 1.5bp. For original trade write up, see Fixed Income Markets Weekly 2023-12-08.
• Maintain SFRM5 / Blues flatteners paired with a 110% risk weighted 3M forward 2s/10s steepener
  P/L on this trade is currently 1.1bp. For original trade write up, see Fixed Income Markets Weekly 2024-03-01.
• Maintain longs in the belly of a U5/M6/H7 SOFR Futures butterfly (-0.37:1:-0.63 risk weighted)
  P/L on this trade is currently -0.1bp. For original trade write up, see Fixed Income Markets Weekly 2024-03-01.
• Maintain Z5/U6 SOFR futures flatteners paired with H6/Z6 SOFR futures steepeners (0.85:1 risk weighted)
  P/L on this trade is currently 0.4bp. For original trade write up, see Fixed Income Markets Weekly 2024-03-01.
  
• Continue to sell 2Yx30Y swaption volatility versus buying 50% of the vega risk in 2Yx2Y swaption volatility , and pay fixed in 2Yx10Y swaps to neutralize the bullish bias in this trade
  P/L on this trade is currently -0.8abp. For original trade write up, see Fixed Income Markets Weekly 2024-02-23.
  
• Maintain 20s/30s 1.33:1 risk weighted maturity matched spread curve steepeners hedged with a 30% risk weighted 20s/30s steepener, but use an equi-notional blend of the Nov 53s and Aug 53s to create a synthetic approximate par bond in the 30Y leg
  P/L on this trade is currently -2.2bp. For original trade write up, see Fixed Income Markets Weekly 2024-02-23.
  
• Maintain 3M forward 2s/3s swap curve flatteners hedged with a 14% risk weighted long in the M4 3M SOFR futures
  P/L on this trade is currently 3bp. For original trade write up, see Fixed Income Markets Weekly 2024-02-23.
• Continue to receive fixed in the belly of a 0.625/1.0/0.375 weighted 3M forward 2s/7s/20s swap butterfly, with an additional 15% risk weighted long in June 2024 3M SOFR futures
  P/L on this trade is currently 3.7bp. For original trade write up, see Fixed Income Markets Weekly 2024-02-23.
• Maintain conditional exposure to a flatter 1s/5s swap yield curve in a selloff using 3M expiry payer swaptions
  P/L on this trade is currently 1.9bp. For original trade write up, see Fixed Income Markets Weekly 2024-02-23.
• Maintain 2Y forward 2s/5s swap curve steepeners paired with 40% risk in 3M forward 2s/5s flatteners
  P/L on this trade is currently -2.3bp. For original trade write up, see Fixed Income Markets Weekly 2024-01-26.
  
• Continue to overweight 6Mx2Y swaption straddles versus a theta-neutral amount of 6Mx5Y swaption straddles
  P/L on this trade is currently -3.4abp. For original trade write up, see Fixed Income Markets Weekly 2024-01-19.
  
• Maintain exposure to long curve volatility by buying 6Mx2Y and 6Mx10Y straddles (41:60 vega weighted) versus selling 6Mx5Y straddles
  P/L on this trade is currently -2abp. For original trade write up, see Fixed Income Markets Weekly 2023-12-08.
  

Closed trades over the past 12 months

P/L reported in bp of yield for swap spread, yield curve and misc. trades, and in annualized bp of volatility for option trades, unless otherwise specified

Note: trades reflect Thursday COB levels, and unwinds reflect Friday COB levels

Swap spreads are still in the spotlight

• This week the decoupling between JGBs and swaps continued to be the main theme in the JPY rates universe. While swaps were mostly flat on the week, the JGB curve bearishly steepened as a result of a weaker-than-expected 30Y auction.
• A barrage of BoJ headlines hit the wires this week, and the market appears to have priced in a higher likelihood of a March NIRP exit. That said, it had relatively limited impact on short- to long-term JGB yields.
• As we approach the March Monetary Policy Meeting (MPM), a wide range of participants reached out to us on how to interpret the market pricing. We explore how we interpret it and what kind of assumptions we make.
• On the swap spread front, we envisage that we have come to a point where investors have done thorough analyses on the money market (including post-NIRP GC/OIS) and are shifting to a wait-and-see mode on that front. In that sense, front-end swap spreads may have to wait for additional catalysts before it finds new equilibria. For now, we are neutral on 2Y swap spreads. In the intermediate sector, whether the MoF reopens JS165 (5Y) at the auction next week is something to look out for from a standpoint of its impact on SC. From the perspective of positioning, generally speaking, it appears more balanced compared to few weeks ago, but JGB futures remain rich vs. swaps both on the absolute spread level and on the swap spread curve, leaving room for further repricing once the rollover is behind us. Said differently, we see a decent likelihood of longer-end swap spreads becoming tighter vs. shorter-end swap spreads. What the market may be potentially discounting is the BoJ’s balance sheet policy, in our view.
• Swap spreads are likely to remain choppy as investors continue to adjust their views to a new normal. That said, there appears to be stronger interest to express swap spread tightening views. As a result, for now, we recommend keeping 10Y swap spread narrowers, but will revisit the trade next week.
• On the topic of the money market, the BoJ released the February call market data on Thursday. In the uncollateralized call market, regional banks remained dominant borrowers of cash. In the T-bill space, the 6-month and 12-month T-bill yields continue to trade above 0%.
• MoF’s flow/BoJ’s balance of payment data: In February, Japanese investors net bought JPY1.81tn of foreign bonds, driven by trust banks (+0.98tn). Trust banks net sold JPY1.16tn of foreign equities, which indicate that pension funds' rebalancing likely contributed to the bulk of these flows.

Market view

This week we had a couple of important data releases. First, the February Tokyo CPI report released on Tuesday came in hot, and the January prints were revised upward: (Headline) 2.6% YoY vs. 2.5% expected; (Core; ex. fresh food) 2.5% vs. 2.5%; and (BoJ core; ex. fresh food and energy) 3.1% vs. 3.1%. Our economics team notes that the global core measure (ex. food and energy) came in at 2.5% YoY, and other core measures also rose sharply, reflecting several technical factors. In addition, the January wage report released on Thursday was solid. Labor cash earnings rose 2.0% YoY vs. 1.2% expected, and underlying details continue to look favorable as constant sample data (which remove sampling volatility) have now risen at or above 2.0% YoY for four consecutive months. Our economists point out that this week’s generally strong set of data is good news from the BoJ's perspective. Market attention now shifts to Shunto (i.e., the spring wage negotiations), and they note that initial indications suggest a headline result approaching the 6% handle and the rise in base wages of around the 3% mark.

On the BoJ front, a barrage of headlines hit the wires this week. Of note, Jiji press (one of the major Japanese media outlets) reported on Wednesday that at least one of the board members will say it is appropriate to end NIRP at the March MPM. Bloomberg reported that BoJ officials are getting more confident over the strength of wage growth on the same day, followed by another Bloomberg report on Thursday, which noted that some government officials support near-term NIRP removal. On the communication front, Board Member Nakagawa gave a speech on Thursday, stating that Japan is making progress toward achieving the 2% inflation target, and once the BoJ concludes that the goal has been attained, it decides on whether to tweak its unconventional policy. Governor Ueda also spoke in parliament, reiterating the same message he and Deputy Governor Uchida have given lately. Against the backdrop of all these headlines, the market appears to have priced in a higher likelihood of a March NIRP exit. That said, it had relatively limited impact on short- to long-term JGB yields ( Figure 94). As discussed later, the market is still in the dark about and appears to be looking for specifics of NIRP exit (i.e., tiering; timing of rate hike effectiveness).

Figure 94: Decoupling continues as JGBs keep underperforming swaps

Change in rates from March 1 to March 8 (bp)

Source: J.P. Morgan.

Figure 95: Market appears to have priced in a higher likelihood of a March NIRP exit after a barrage of BoJ headlines

BoJ policy rate outlook based on OIS market expectations (TONAR) (bp)

Source: J.P. Morgan. Note: We think market expectations should be seen in fairly broad terms as it is unclear whether market participants are expecting rate hikes to be effective immediately or from the next reserve maintenance period (see here).

As we approach the March MPM, a wide range of participants reached out to us on how to interpret the market pricing ( Figure 95). To recap, market participants are currently pricing MPM swap contracts while dealing with three unknowns: (1) the timing of a NIRP exit announcement (e.g., March, April or after April); (2) the timing of a rate hike becoming effective (immediate or from the next Reserve Maintenance Period, RMP), and (3) the choice of tiering or TONAR fixings. Given the number of uncertain variables, one has to make assumptions when interpreting the market pricing.

For example, if we assume the market is pricing in a 100% likelihood of a March NIRP exit announcement and it is expecting TONAR to settle at 9bp after NIRP removal, then the current March BoJ swap level (4.1bp) implies a 30% or so probability of NIRP removal becoming effective immediately and approximately 70% probability of from the next RMP ( Figure 96). Alternatively, we can slice it somewhat differently: if we assume that the market is pricing a 50% chance of immediate effectiveness of NIRP removal and 50% of the next RMP scenario and assume TONAR settles at 9bp, then the current March BoJ swap level implies an 80% or so probability of NIRP exit announcement at the March MPM and 20% probability of hold.

We think it is safe to assume TONAR fixings of 9bp (or implicitly a one-tier structure) given that is the level at which the June MPM contract currently trades (Note: June is the only time the next business day of the MPM and the next RMP fall on the same day). Market participants' views on the timing of rate hike effectiveness appear to be more or less equally split, and if we are correct, that leaves the market pricing of a March NIRP removal probability well above 50%.

Figure 96: When dealing with BoJ policy rate expectations, one needs to make assumptions as there are three unknowns: (1) the timing of a NIRP exit announcement; (2) the timing of a rate hike becoming effective, and (3) the choice of tiering, which affects TONAR fixings

Theoretical OIS pricing under different scenarios (bp) and implied probabilities under two scenarios (%): Assume TONAR stays at -0.01% before the rate hike becomes effective/under no hike. For hike scenarios, we present OIS pricing at (1) different TONAR fixings (5bp, 7bp, 9bp) and (2) different timings at which the hike becomes effective (immediately/from the next Reserve Maintenance Period, RMP) (as of March 8)

Source: J.P. Morgan estimates. Note 1: We use the following start/end date: (Mar BoJ) 21 Mar. - 26 Apr. Note 2: Under “immediate”, we assume that TONAR reprices higher from the next business day of announcement.

Away from the market pricing, at the center of the stage was continued tightening of swap spreads ( Figure 97). The first leg of the tightening which started in mid-February seems to be driven by the recognition that GC may cheapen vs. TONAR after NIRP removal as we explored in our plumbing report, and consequently, swap spreads repriced tighter led by the front end. However, the tightening did not stop there, triggering a second wave of fast-paced repricing, and this time, we have seen the tightening across the curve. The second leg of the move appears to be more influenced by adjustments to positioning. Even as swap spreads tightened, JGB futures remained rich vs. swaps partly because we are in the middle of rollover season, but on Wednesday, it finally repriced. At the back end, it appears that foreign investors led the swap spread tightening to a certain degree as LCH-JSCC clearing spreads became tighter/more negative in lockstep ( Figure 98).

Figure 97: Swap spreads continued to tighten this week, likely led by positioning

Change in swap spreads from March 1 to March 8 (bp)

Source: Bloomberg Finance L.P., J.P. Morgan.

Figure 98: LCH-JSCC clearing spreads became more negative

LCH-JSCC clearing spreads (bp)

Source: J.P. Morgan, Bloomberg Finance L.P.

We envisage that we have come to a point where investors have done thorough analyses on the money market (including post-NIRP GC/OIS) and are shifting to a wait-and-see mode on that front. In that sense, front-end swap spreads may have to wait for additional catalysts before it finds new equilibria. For now, we are neutral on 2Y swap spreads. In the intermediate sector, whether the MoF reopens JS165 (5Y) at the auction next week is something to look out for from a standpoint of its impact on SC. From the perspective of positioning, generally speaking, it appears more balanced compared to few weeks ago, but JGB futures remain rich vs. swaps both on the absolute spread level and on the swap spread curve, leaving room for further repricing once the rollover is behind us. Our estimate of positioning indicates that foreigners are still net long futures, and our Delta One team's CTA signal is indicating short ( Figure 99). Said differently, we see a decent likelihood of longer-end swap spreads becoming tighter vs. shorter-end swap spreads.

What the market may be potentially discounting is the BoJ’s balance sheet policy, in our view. We think there is a decent likelihood of further Rinban reductions taking place from April along with the MoF’s issuance cuts, as we assessed here. The BoJ purchases JGBs at the pace of JPY5.9tn/month (gross), and given that about JPY70tn of JGBs are expected to be redeemed in the next 12 months, the annual net purchase pace currently stands near zero, and further Rinban reductions seem highly likely bring that number to sub-zero. Said differently, the BoJ is likely to start passively returning collateral to private hands, which should pressure funding rates to a certain degree. Furthermore, we analyzed QT and its implications in depth here, and our view is that the front end of the curve should enjoy strong stock effects for some time as the BoJ holds around 70% of the total outstanding; however, the back end will likely bear the brunt, although the degree of impact would depend on the BoJ’s reinvestment approach. We are attentive to how the BoJ's balance sheet policy will filter through to swap spreads pricing as well.

Swap spreads are likely to remain choppy as investors continue to adjust their views to a new normal. That said, there appears to be stronger interest to express swap spread tightening views. As a result, for now, we recommend keeping 10Y swap spread narrowers, but will revisit the trade next week.

In the cash space, the 10Y auction on Tuesday was digested with ease even as some market participants had expected a weak auction result given that the pre-auction 10Y yield was around 0.7%. In contrast, the 30Y auction on Thursday was weaker than market expectations. Some investors had expected the auction to be digested with ease as it was the final 30Y auction in FY2023 and as higher equity prices should be supportive of rebalancing flows. As a result, the JGB curve steepened, and the broad curve such as 5s/30s stands at post-July MPM levels.

Switching to the JGB futures rollover, JBH4/JBM4 has been calm and steady, closing at 60 sen as of March 8 ( Figure 100). We had expected JBH4/JBM4 to follow the recent roll pattern and that as we go into the peak of the rollover activity (from around LTD-5 to around LTD-3), short rollover should take the lead, pushing up the calendar spread higher. The calendar spread followed the recent roll patterns to a certain degree and moved higher from the local low of 57 sen, but it only moved in a very tight range. It appears that rollover supply-demand dynamics was more balanced this time.

Figure 99: Our estimate of positioning indicates that foreigners are still net long futures

JGB futures' estimated net positions: investment trusts and trust banks (left) and foreigners (right) (latest data point as of March 1)

Source: JPX, J.P. Morgan estimates. Note: We arbitrarily set January 2019 as the base point.

Figure 100: Calendar rollover has been calm and steady

JBH4/JBM4 rollover over the past 14 business days (yen; contracts)

Source: JPX, J.P. Morgan. Note: 1 contract is worth JPY 100mn.

On the topic of the money market, the BoJ released the February call market data on Thursday. In the uncollateralized call market, regional banks remained dominant borrowers of cash ( Figure 101). They have been likely trying to use up the macro add-on balance by borrowing cash in the call market, considering that when the BoJ introduced NIRP, it tied the basic balance to the amount of reserves each bank parked at the BoJ between the January and December 2015 RMP. While the market consensus on tiering has shifted from a one-tier structure to two-tier, it appears that regionals are still willing to borrow at TONAR very close to 0%. In the collateralized call market, investment trusts remained major cash lenders with the BoJ's monthly data pointing to around JPY2.6tn of cash lent by them on average. As we noted here, the collateralized space warrants close attention after NIRP removal.

In the T-bill space, we had 6-month and 3-month T-bill auctions on Thursday and Friday, respectively. For the former, the average and cut-off yields came in at 0.0259% and 0.0418%, respectively, moving above 0% for the first time since 2014, while for the latter, the respective yields came in at -0.0606% and -0.0409%. The 6-month and 12-month T-bill yields continue to trade above 0% as of today's close ( Figure 102). Both of these instruments may not be preferable for some money market investors such as money reserve funds (MRFs) due to WAL/WAM requirements, but we will closely monitor incoming holdings data for any sign of their return to the T-bill universe.

Figure 101: In the uncollateralized call market, regional banks remained dominant borrowers of cash

Borrowing and lending by selected players in the uncollateralized call market (JPY tn; average balance)

Source: BoJ, J.P. Morgan.

Figure 102: T-bills with remaining maturity longer than six months have positive yields

T-bill yields by issue (%) (as of March 8)

Source: J.P. Morgan.

MoF’s flow/BoJ’s balance of payment data (Feb/Jan)

The MoF released the monthly portfolio flow data for February, and the BoJ released the balance of payment data for January (see here for the full report) ( Figure 103).

In February, Japanese investors net bought JPY1.81tn of foreign bonds, driven by trust banks (+0.98tn). Trust banks net sold JPY1.16tn of foreign equities, which indicate that pension funds' rebalancing likely contributed to the bulk of these flows. Banks net purchased JPY0.54tn, decelerating from the January net buying amount of JPY1.99tn. The MoF's weekly transactions data indicate that Japanese investors net purchased JPY1.5tn of intermediate/long-term bonds and JPY0.3tn of short-term bonds during the week ending February 9. However, they turned net sellers in the second half of February. Japanese investors may have interpreted the strong employment report as a one-off/seasonal factor-driven matter, but a series of strong data including CPI and PPI likely weighed on their sentiment. As a result, they may have decided to stay on the sidelines to assess whether the January data points are one-off/seasonal or not.

Turning to the regional breakdown data for January, Japanese investors net bought JPY2.67tn of foreign bonds. US bonds saw an inflow of JPY3.28tn. EU bonds in aggregate saw an outflow of JPY0.05tn, but the breakdown was mixed: French and Italian bonds saw outflows of JPY0.12tn and JPY0.05tn, respectively, but German bonds saw a large inflow of JPY0,33tn.

Figure 103: Japanese investors net bought JPY1.81tn of foreign bonds led by trust banks (+0.98tn)

Flow of foreign bonds (short-term + intermediate/long-term) Investment (JPY tn)

Source: BoJ, MoF, J.P. Morgan.

Trade Recommendations

• 10Y swap spread narrower: Receive 10Y swap (JSCC) @ 0.858% vs. Short 10Y JB373 (maturity: 20 December 2033) @ 0.665%. P/L since inception is +7.3bp (recommended 19 January 2024)

Data-driven

• Australia’s 4Q GDP data were expectedly soft, though with better than expected details on income and saving which bode well for 1H24
• ...still, AUD fixed income has been driven mostly by the headline, not the nuances...
• ...and the combination of softer local data with higher for longer in the US drives outperformance and flattening
• Maintain a tactical bias to flattening through the belly (received AUD 3s/5s/10s IRS)
• With Fed cuts and curve steepening in cash looking a less imminent, the EFP curve shape should be relatively stable
• This shines a light on spread product, particularly previous underformers on ASW
• Open a long in TCV Nov-2034 given significant cheapness vs other issuers, seemingly driven by fiscal news which appears priced near-term
• Bank securities demand should remain stable given more evidence this week that mortgage growth will remain subdued
• NZD rates have held on to the post-RBNZ rally...
• ...the 5s/10s curve has re-flattened, though the set-up for performance on longer ASW and received outright in the sub-2Y bucket remains favourable
• Trade portfolio: Stay received August ’24 RBNZ OIS, hold NZGB Apr-2033s on ASW and enter a long in TCV Nov-34 on ASW.

Strategy Update: AUD flattening as policy traction accumulates

This week’s 4Q GDP data in Australian was soft as expected, consolidating the sense that the RBA’s fast policy transmission is restraining demand ( Figure 104). We have expected 2H23 to mark the trough for sequential growth, and there were some positive details in the report regarding households’ spending capacity (saving rate higher), which complements growing support from conclusion of rate hikes, falling inflation, and approaching tax cuts into 2024. Productivity outcomes were also substantially better in the back end of last year, suggesting the supply-side will not be constraining the RBA’s policy choices.

Still, in an environment of heightened uncertainty, market participants have placed significant weight on the simple headline data, which have been mostly softer of late. AUD fixed income is rallying, which alongside pruning of Fed cuts has pushed the 3s/10s curve to the flattest levels this year. As discussed in recent weeks, evidence of policy traction in the local data will make it difficult for the market to price the ‘higher r*’ narrative in Australia, even if it is has been flirting with this concept in the US. As such, we see the prospects for underperformance in AUD as now much more conditional on a significant US rally/steepening, which hasn’t appeared imminent.

Figure 104: Australian GDP growth has slowed to 1.5%oya in annual terms, though we see the sequential pace picking up in 1H24

%oya

Source: ABS, J.P. Morgan.

Figure 105: Realized RBA policy has been more hawkish (or more explicitly, less dovish) by comparison to pricing, which is consistent with a flatter curve

LHS: Overnight cash less 3M-ago pricing (Bp); ACGB 3s/10s futs curve (%-pt, inverted)

Source: J.P. Morgan.

The AUD 3s/10s curve also looks relatively fair on a standalone basis given the RBA has on net been less dovish compared to market expectations, proxied by the realized cash rate less 3M-ago pricing ( Figure 105). Given this we have been more tactically inclined to trades with a flattening bias through the belly (receive belly AUD 3/5/10s IRS) and which fade cuts to an extent.

Spread product is also worth another examination given that a flatter bond curve should interrupt EFP flattening, and allow a reprieve for some previous underperformers on asset swap. We entered 2024 with the view that the 3s/10s EFP box would trend narrower, then invert on late cycle dynamics through the year. That remains our view, though we have executed this tactically given catalysts have been only intermittent. In the past, the main driver of substantial flattening and inversion of the AUD swap spread curve has been steepening in cash, as rate cuts became imminent (more so, if credit spreads are also widening, Figure 106). Neither appear in play yet, and swap spreads are close to fair on current inputs ( Figure 107).

Figure 106: Compared to 3Y EFP, 10Y EFP is more sensitive to ACGB curve shape and less sensitive to credit spreads

EFP model betas to selected inputs

Source: J.P. Morgan. Coefficients for 100bp increase in 3s/10s futures curve, 10bp widening in AUD 5Y corp iTraxx. Models also use AUD swap curve RMSE, OIS slope from overnight to next futures expiry (indexed by days to expiry), 1Y XCCY basis and USD swap spreads. Daily data since 2010.

Figure 107: The 3s/10s EFP box is slightly flat to fair value on current inputs

AUD 3s/10s EFP box, Bp

Source: J.P. Morgan. Model as described previously.

Recently, the moderate risk of bond steepening/credit widening has appeared smaller than the roll effect into the next futures contract, where paying the box may appear more optically attractive (i.e. assuming a subset of investors treat EFP as a spot instrument from a range perspective, Figure 108). We have been defensive on semis this year, mostly reflecting cyclical risks to EFP shape mentioned above. If the latter is to be range-bound for now, prior underperformance of some names relative to swap becomes more attractive to fade.

Figure 108: 3s/10s EFP is implied steeper out of the M4 futures contract

EFP roll optics (using back contract less front), Bp

Source: J.P. Morgan.

Figure 109: TCV paper has widened significantly on ASW, and the noisier/flatter relationship to 10Y EFP vs other issuers suggests fiscal concerns have been a driver

X axis: 10Y EFP (Bp); Y axis: TCV Nov-34 spread to 10Y ACGB (dark blue), QTC Jul-34 spread to 10Y ACGB (light blue), Bp

Source: J.P. Morgan.

Longer TCV paper in particular has cheapened to other large semi issuers (NSWTC and QTC). The underperformance seemingly reflects idiosyncratic fiscal factors, given the TCV-ACGB 10Y spread has maintained a noisier and flatter relationship to 10Y EFP, than has say QTC ( Figure 109). Fiscal bad news appears to have been cleansed near-term with 10Y TCV paper back near range wides to swap. Domestic bank demand for semi paper should also remain relatively firm if rate cuts are not imminent, and on recent data, including home lending commitment flows out today, bank securities demand is unlikely to be crowded out by an acceleration in mortgage activity. Further, if banks have any inclination to resist the balance sheet contraction that will occur as TFF matures, it will result in greater securities demand (funded by market sources). Enter a long in TCV Nov-34 on asset swap.

Figure 110: From late last year the NZ long-end appeared to receive real money inflows, as the 10Y spread to US narrowed in a fashion not explained by front-end spreads, and was negatively correlated to NZD FX

T Stats of inputs in rolling 3M regression of NZGB-UST 10Y spread

Source: J.P. Morgan. Rolling 3M models including all of the above variables, on daily data.

Figure 111: After the RBNZ’s non-hike, the front-end parallel rallied and 5s/10s steepened, but the latter was quickly undone

NZD IRS curve spreads, Bp

Source: J.P. Morgan.

Shifting to New Zealand, in the wake of the RBNZ’s on-hold decision, front-end rates have held onto the post-meeting rally, but have only undone around half the sell-off from the last month or so. Until fears of a renewed hiking cycle re-emerged, the NZD long-end seemed to have been garnering real money inflows, as longer-end asset swaps were richening, and the NZGB-UST 10Y spread had been less-driven by (compressing) front-end spreads, and more associated with FX appreciation (the negative T stat in Figure 110).

Owning the NZ long end has indeed made sense on the view that, under an accumulation of weak data, either the RBNZ would pivot dovish or at least, delayed cuts would weigh on term yields. The policy ‘scare’ has however undone the FX dimension of this story lately and suggests perceived risks of a larger policy mistake – which would hurt growth and so the exchange rate – may have risen. The initial steepening of NZD 5s/10s as the RBNZ definitively rejected a hike – which had been priced at 30% probability – has been quite quickly undone ( Figure 111). This again suggests the market is implying the risk that the RBNZ will act too late to stop a more disorderly deterioration in the economy. All that considered, the path to cycle turns in NZ is typically not straightforward and the policy shift tends to be abrupt when it does come. Stay received August ’24 RBNZ OIS, hold 2s/5s IRS steepeners on box to AUD and hold NZGB Apr-2033s on ASW.

Summary of main views

Trade Recommendations

• Own TCV Nov-2034 on asset swap
  Long 25K DV01 TCV Nov-2034 on ASW (3M) at +54.8bp.
• Receive August RBNZ OIS
  Receive 25K DV01 Aug ’24 RBNZ OIS at 5.35%. P/L since inception (Feb 29): +0.8bp
• Receive the belly of the AUD 3s/5s/10s IRS fly
  Receive 50K DV01 AUD 5Y IRS at 4.139%, pay 25K DV01 AUD 3Y IRS at 4.122%, and pay 25K DV01 AUD 10Y IRS at 4.414%, a spread of -28.8bp. P/L since inception (Feb 22): -0.3bp
• Pay May RBA OIS vs receiving April RBNZ OIS (*)
  Pay 25K DV01 May ’24 RBA OIS at 4.275%, receive 25K DV01 Apr ’24 RBNZ OIS at 5.445%, a spread of -116.5bp. P/L since inception: -6.8bp.
• Own NZGB Apr-2033s on ASW
  Long 25K DV01 NZGB Apr-2033 on asset swap, at a yield spread of +5.1bp.
• Pay the belly of the AUD 1s/2s/3s IRS fly
  Pay 50K DV01 AUD 2Y IRS (Q/Q) at 3.981%, receive 25K DV01 AUD 1Y IRS (Q/Q) at 4.180%, receive 25K DV01 AUD 3Y IRS at 3.892%, a curve spread of -11.0bp. P/L since inception (14 Dec): +0.7bp
• Long 2025 ACGBi BEI
  Long 25K DV01 2025 ACGBi breakeven at 2.90%. P/L since inception (6 Dec): -24.8bp
• Hold a NZD-AUD 2s/5s IRS box steepener
  Pay 25K DV01 of NZD 2s/5s IRS at -77.75bp, receive 25K DV01 of AUD 2s/5s IRS at -14.8bp, a spread of -62.9bp. P/L since inception (22 Jun): +0.1bp

Note: One star (*): high-conviction trade and no star: moderate-conviction trade. Closed trades marked as at submission, existing trades marked as of the close of business (Sydney) on 7 March 2024.

Forecasts

Source: J.P. Morgan. Current values are from Bloomberg Finance L.P.

Trades closed last 12M

Source: J.P. Morgan. Note: Past performance is not indicative of future results.

* Levels as of 2.30pm London time.

Note: In billions of local currency except for Japan which is in ¥trillions
* Marketable bonds include: conventionals, linkers, floaters zero coupons and international bonds.
** Maturities in all currencies and jurisdictions and include secured, unsecured and securitised issuance, including MTNs but excluding short-term (maturity of less than one year) and self-funded deals (deals where there is only one bookrunner and it is also the issuer). The data also include any government guaranteed issuance by the banks but no direct issuance by government or government sponsored institutions.
*** Austria, Belgium, Finland, France, Germany and Netherlands.
**** Greece, Ireland, Italy, Portugal, Spain.
Source: J. P. Morgan, Dealogic, Bloomberg

Selective list as of 08 Mar 2024. Forecasts are m/m, nsa, unless stated & times are local. Telephone your J.P. Morgan representative for an update/more details.
Highlighted data are scheduled for release on or after the date shown. Times shown are local.