Our medium term views

 Our key forecasts

  What we like right now

  The View

 The week that will be

The main data print ahead is today's US labor market report. Our economists are looking for an above consensus +200k number on payrolls (see Federal Reserve Watch 3 Sep 24) with solid details (see Labor Market Watch 4 Sep 24). Positioning into the report is long, and in steepeners (see US Rates Watch 2 Sep 24), especially after this week's data.

The 2s10s curve is some 20 bp steeper since our last weekly (see GRW 16 Aug 24). Dec FOMC OIS is 16 bp richer to price in a cumulative 111 bp of cuts in 2024. We therefore believe that rate markets may reverse some of the recent bull steepening even if the number is not quite as strong as our economists expect. We are buyers of the dip, but not chasing the market here (see Rates US and US Rates Viewpoint 4 Sep 24). Price action after JOLTS data suggests the technical consolidation phase in yields may already be breaking to the downside. If NFP comes in soft it could push yields even lower and the curve steeper (see Technicals and Technical Advantage 4 Sep 24).

Outside of today's labor market report (and subsequent speech by Governor Waller), we look ahead to next week's inflation print for August (see CPI Inflation Watch 5 Sep 24). Our economists expect headline CPI to rise by 0.2% m/m (0.15% unrounded) and core increased by 0.2% (0.21% unrounded), which we think would be neutral for markets.

For the Euro Area (EA), the focus is clearly on the ECB next week. We expect a second cut of 25 bp in the depo rate, with little guidance for the remainder of the year. Forecast revisions are likely to be minor but confirm the ECB's base case of a gradual return to target in inflation. The main question is to what extent the ECB acknowledges the downside surprises in activity data vs their forecasts, which could be taken dovishly by the market. We remain long EUR rates both outright and cross market (see Rates EU and European Rates Viewpoint 3 Sep 24). We also like selling payers in EUR 2y1y to fund payers in US (see US Rates Alpha 3 Sep 24).

It is also worth flagging that the details to the 2Q24 GDP print, published today, will allow us to update figures on savings rates and CAPEX for the EA, which have been painting a concerning long-term signal.

Finally, in the UK our economists are looking for another decline in the unemployment rate and a decent bounce in the monthly GDP print. We remain short in UK real yields both outright and cross market, in 2s5s steepeners and are now also paid the belly in 3s5s7s (see Rates UK and UK Rates Viewpoint 5 Sep 24).

The weeks that were

US data confirmed that the US economy is in a landing scenario - the main driver of our downside revision to 10y yield forecasts (Liquid Insight 5 Aug 24 and US Rates Watch 6 Aug 24). Between revisions to payroll data, manufacturing ISM and JOLTS, markets are again starting to worry about hard landing risks. We are at target on US 10y.

The BoC cut as expected and messaging confirms our bias of a cutting cycle. However, at this stage the CAD rate cycle looks more than perfectly priced. We are short CAD 3y1y OIS vs EUR (see European Rates Viewpoint 3 Sep 24).

  Rates - US

• Short term we are tactically underweight into NFP data; cuts are well priced
• Beyond Friday we favor dip buys; history = duration outperforms in cut cycles

NFPop quiz

Friday's employment data will set the tone for Fed cuts & the rates outlook. Rates clients suggest the key threshold is NFP headline around 125-150k & U3 4.2/4.3% (our own threshold assessment is here: Fork in the road). Rates clients say NFP >=150k with U3 4.2% will cement 25bps as starting cut size; clients say NFP <=125k with U3 4.3%+ means 50bps cut. The initial cut size will set the stage for remaining '24 meetings & Sept dot plot. Our economics expect NFP = 200k & U3 4.2%; if right, rates bear flatten.

Into the print we recommend clients be underweight the front end & in bear flatteners. This is a short-term tactical view but reflects very full Fed pricing, recent sharp rate decline, stretched positioning in longs & steepeners (see: weekly positioning report). We see larger rate reaction to strong vs weak employment data.

Waller at 11 AM Friday morning will guide the market on Sept cut size of 25 vs 50bps. We are looking for key words around cutting pace: "gradual" = 25bps, "accelerated" = 50bps. The Fed won't leave the market guessing for long about its Sept intentions. Next week CPI will need to materially surprise to shift Waller's guidance. Our economists expect continued inflation progress with a 0.2% on headline & core (see August US CPI).

Beyond Friday: bye summer & buy dips

Beyond tactical positioning around Friday's employment data, we still believe investors should look to buy any meaningful selloff in rates. We think 3y1y OIS levels above 3.1% (Fed's '26 median dot with unemployment at 4.1%) is an attractive buy.

After the summer holiday, we refresh our rate views (see Back to school cheat sheet). Our core rate views remain in-line with our mid-year update: keep buying dips, stay short spreads, vol surface steepener, & higher funding in 2H24 but stabilization / partial reversal in 1H25. We have also made several adjustments since our mid-year: we revised our rate forecasts lower, embraced our long held 5s30s nominal steeper, & now initiate a forward inflation steepener. Our cheat sheet further elaborates on these views.

History lesson: trading the cutting cycle

Clients have ask frequently about how to trade the cutting cycle. We use history as a guide. History says there is room for the duration rally to continue, though comparisons are challenging as this is the most well anticipated cutting cycle in history. This month's cut has been to some extent priced by the market since FF was at 3% (Exhibit 4).

We have argued in the past that investors want to buy duration around the last rate hike of the cycle (see: Liquid Insight). This guidance has proven true historically, but timing has been very important. Since the Fed's last hike in July '23, the 10y rate has rallied merely 10bps. Since the subsequent peak in 10y rates in October '23, rates have rallied 120bps. The change from peak puts this cycle's rally amongst some of the largest observed but buying three months before vs three months after the last hike made a large difference in this case.

While we continue to believe timing is important, history also tells us that there is more room for the rally to go (see: Trades for cutting cycle: a historical comparison). On average 10y rates have rallied 230bps from the timing of the first cut to the last cut of the cycle. We also historically see that most of the curve steepening is still ahead of us and occurs as cuts are delivered.

A major consideration this cycle is how much the market is already pricing for cuts (the carry hurdle for longs & steepeners). Exhibit 5 shows the historical moves across the curve in the 6 months after the Fed's first rate cut over cycles back to 1982. We also compare versus what forwards currently imply to discern which tenors/ steepener expressions are the most reliable trades. Three main takeaways stand out:

1. Most of the time rates rally more in the 6 months after the Fed's first cut than what the market currently prices in forwards.
2. The 5s30s steepener has a higher hit rate of historically beating what current forwards imply over the next 6 months vs the 2y10y.
3. The only expression examined below that has outperformed current forwards in all historical observations is being long 10y, though in the 2001 cycle barely so.

While we continue to believe that timing matters, history suggests that there is more room for duration to rally and steepeners to outperform. Using history as a guide, maintaining a long bias in duration and buying in dips are favorable strategies as the Fed begins its cutting cycle.

Bottom line: Friday's employment data will set the tone for Fed cuts & the rates outlook; Waller will make Fed intentions clear. Short term, we favor tactical underweights & flatteners into payrolls. Beyond Friday, we recommend clients buy rate dips. History says duration can outperform as the Fed launches cuts.

  Rates - EU

• We expect the ECB to cut the depo rate by 25bp; we prefer ECB trades further down the curve and like to be received on an outright and cross market basis
• The narrowing of the refi-depo corridor is expected to lead to €50-65bn of MRO/LTRO take-up; EGB spreads look relatively rich, ECB should not matter much

ECB preview: looking beyond September

Our economists expect the ECB to cut the depo rate by 25bp in September and no changes to guidance. Our base case is for the ECB to cut the depo rate by 25bp again in December 2024, and then to accelerate the 25bp cuts to one a meeting from March 2025 until the depo rate reaches 2.00% in July 2025. From then, we expect a further 25bp cut in March 2026 and June 2026, implying a terminal depo rate of 1.50%.

When compared with our forecasts, the market has priced in the possibility of an earlier acceleration of rate cuts by the ECB. In the case of October 2024, the market has priced in c. 40% probability of a 25bp cut, conditional on a 25bp cut being delivered in September. Our economists believe the bar for that to be delivered is too high. But they also believe the ECB is unlikely to clearly rule it out, which could be seen by the market as dovish. Since the July ECB meeting, market pricing of an October cut has been heavily influenced by euro area activity data, the US labour report, and Fed policy; where almost a full 25bp had been priced in for that meeting alone in early August. We expect these factors to continue driving market pricing of the October ECB meeting.

Our favourite ECB trades remain to be further out of the curve, where we have more conviction over the terminal depo rate and believe it will be lower than what is currently priced in. We like being received on an outright and cross market basis around the 2y1y point of the curve (see Rates-EU II for more details). Market pricing of a lower terminal may be supported by the expected small downward revisions to growth and inflation forecasts by ECB staff due in September.

The ECB will narrow its refi-depo corridor from 50bp to 15bp, effective from 18 September. This means the refi rate is expected to decrease from 4.25% to 3.65% in September. We believe the tighter corridor will support bank take-up of MRO and LTROs. The first MRO and LTRO that incorporates the tighter corridor will settle on 18 September and 25 September, respectively. We expect banks to take-up between €50bn and €65bn in MRO/LTRO in September 2024 (see Front-end EU for more details).

When it comes to Euro govie spreads this ECB meeting does not look set to be a major driver. The combination of a market that is Fed-led and an ECB that is not expected to deviate excessively from the current market narrative would indicate so. Country-specific headlines also do not seem to be moving this market significantly yet: the combined effect of headlines seeing Italy's Meloni committing to 3% targets within two years (soft market positive) and the choice to appoint Michel Barnier as French PM (seen as fiscally conservative vs alternatives) managed to move BTP-Bund and Bono-OAT spreads by 2bp and 1bp respectively. The changing market assessment around US economic deterioration and the risk market sensitivity to this story is likely to dominate EGB spread pricing over the ECB or idiosyncratic events in the short-term. Given what is priced (little), we have a spread widening bias.

  Rates - EU II

Value in longs: in 30y Bunds & new cross market positions

We have long held a structurally bullish view on EUR rates for end of 2024-2025, consistent with our economists' conviction that the ECB will have to cut rates to below 2% (their baseline being 1.5% by mid 2026). This is underpinned by very low inflation forecasts (headline revised to 1.4% on average in 2025 and 1.7% in 2026 - Exhibit 6), and below consensus growth (Global Economic Viewpoint, 2-Sep).

We have been tactical in our recommendations, closing our long Bund position as it hit target post Dec rally and closing our long Bund vs UST in May, as the theme of US-EUR divergence peaked. We cautioned against re-entering the UST-Bund widener in the near term, despite tighter levels than in 1Q, and expressed a neutral view on Bunds for the short term, with US rates / events in the driver seat. Indeed, this summer, the rally in 5-10y EUR rates can be fully explained by their historical beta to US rates. Still, we kept our received 2y1y €str recommendation (now 15m1y) as a long-term trade to express the view of a lower ECB terminal rate. The front-end is where EUR rates managed to outperform their 1y beta to US rates (Exhibit 7). We now take on board recent price action and recommend the following outright:

• Buying 30y Bunds, currently at 2.52%, targeting 2%, with stop at 2.82% (YTD high). The main risk is reduced duration demand from pension and insurers, but we see five reasons to justify building this position in September:

1. Macro: signs of fragilities are emerging, not just via the PMIs, which we tend to discount but also in the labour market for example. The ECB also appears more focused on growth risks. The next inflation prints should also start to show a significant gap vs ECB expectations, supporting a more dovish tone.
2. Supply: preparation for Sep supply started to put some pressure on duration / curve (in line with seasonality), but this year, 30y+ supply can be low (Exhibit 8).
3. positioning isn't as long as views (FX & Rates Sentiment Survey). There is therefore scope for supply to be better absorbed this year (see our discussion in Global Rates weekly, 13-Aug). At least, any dip in duration / curve steepening ahead of the supply could therefore be actively bought.
4. Pension and insurance hedging / duration buying can extend into Q4. The summer period tends to be quiet on that front. We expect demand to materialise around the primary market and with large pension funds de-risking as part of the transition to DC (Liquid insight, Jul 4th).
5. Entry levels: long-dated EUR rates have underperformed relative to their beta to US rates in the summer rally. It is where valuations are now most attractive relative to that and to our forecasts (Exhibit 9), as we expect 10s30s to grow more resilient to the steepening in 2s10s, extending the residual in Exhibit 10.

• Rolling the now 15m1y €str position (entered: 2.45%, current: 2.09%), ie closing it to enter a received 2y1y €str (current: 2.12%). The later benefits from a +2bp 6m roll as opposed to the -7bp 6m roll of the 15m1y. We target 1.7%, w/ stop at 2.35%. The main risks are stronger global data and higher than expected EZ core inflation.

Cross market, we believe EUR rates can outperform but selectively. We recommend:

• In swaps: receive 3y1y EUR vs paid in CAD (entry: 41bp, target: 80bp, stop: 15bp) The divergence In terms of cuts priced is most significant in 3y1y (Exhibit 11), with the market pricing in that the ECB would stop at above 2% in one year (vs our baseline of more cuts in '26), while the BoC would extend the cuts to c.2.6% (vs BofA baseline of 3%). The main risk to the trade is weak CAD/US data.
• In options: sell 1y2y ATM EUR payer (strike: 2.28%) to fully fund a 1y2y ATM+25bp USD payer (strike 3.4%) - target: 25bp, stop: -15bp. We see scope for the market to reprice US neutral higher to above our baseline of 3.25-3.5% if coming data reduce recession concerns. We favour options as downside surprises in US data would lead to outperformance of US rates. The risk to the trade is higher
  EZ inflation.

In inflation: receive 5y real rate in EUR vs UK (current: 38bp, target: -40, stop: 90bp). Our inflation strategist entered the trade to position for (1) larger UK short-dated issuance, (2) inflation risks limiting BoE cuts vs ECB, (3) UK's deteriorating IIP/GDP ratio vs the EZ (Liquid Insight 21-Aug). The risk is UK recession

  Rates - UK

• Four reasons investors might be reluctant to add UK duration longs in Autumn: (1) macro, (2) fiscal, (3) BoE QT & (4) seasonals. We stay short UK RYs vs. EUR.
• We enter a 3s5s7s Sonia fly (pay belly, receive wings) expecting banks' structural hedging flows to give way to mortgage paying flows.
• We sell UKT 4.5% 2028 Gilt to buy UKT 0.5% 2029 as a low cpn/high cpn trade: the theme might be past the peak, but there is still juice we think.

Tough Mudder

Investors hesitated to add to UK duration longs during and after the market's flight-to-quality episode in July (risk asset sell-off, weaker-than expected US data), despite sentiment seemingly turning more constructive (Exhibit 12 and our report, If in doubt, buy US duration, 9 August). While it is possible that UK rates "exposure" might catch up with "view" in September, we think it is more likely that investors remain cautious while navigating the UK (and global) macro obstacle course in the Autumn.

In July/August, the GBP market underperformed its YTD beta-to-US-rates sharply, perhaps signalling a reassessment of the "quality" aspect of UK rates relative to elsewhere (Exhibit 13). We think some of this has been reflected as a term premium rebuild (see our report, 10y underperforms terminal rate pricing, 13 August), which we have long been arguing was necessary in the UK, given its macro fragilities (see our report, The art of falling apart, 24 August 2022).

We see several reasons for this reluctance to add UK duration longs persisting:

1. Macro: Our economists highlight rising risks that labour market easing over the past few months is stalling; they also expect the recent services inflation weakness to reverse in the next print, confirming the need for a slow easing cycle from the BoE. For 2024, we expect one more cut in November (see our note, Back to School: Stronger growth, upside inflation risks, cautious BoE, 3 September).
2. Fiscal: Risks to Gilt supply from the DMO and BoE combined, post the September QT decision and the Budget, seem skewed to the upside (see our report, Gilt supply in 4Q24: modestly lighter but upside risks beyond Oct, 3 September). Risks stemming from the Budget relate to (1) the Chancellor's plans and commitments and (2) the widely-discussed possibility of changes to the debt stock definition used in the fiscal rules. Gilt supply might also be raised by (3) an upward Remit revision in October because of a higher-than-projected CGNCR in the first four months of the fiscal year.
3. BoE QT: The BoE's decision on Gilt stock reduction for the next year and its Gilt sales schedule for 4Q24 should arrive alongside the September MPC minutes. Our working assumption is that the BoE will vote to reduce Gilts held in the APF by £100bn over the next twelve months, implying £13bn of "active" Gilt sales. But the BoE's confidence in the progress of QT, expressed in August's Monetary Policy Report, hints at upside risks to our £100bn assumption.
4. Seasonality: Seasonals are not to be blindly followed (and indeed they did not fully pan out this July/August), but a seasonal Gilt yield rebound in September appears to be more frequent vs. Bunds (74% of the time) than USTs (Exhibit 3).

 All this could lead to the 10y Gilt-Bund spread breaking above the recent trading range (c. 170-175bp), we think, providing a tailwind for our real yield spread widener view (see our report, Pay 5y UK real rates vs. euro, 21 August and Inflation - UK section). We are hesitant to recommend UK-US cross-market trades at this point given our US team's warning on stretched market positioning at the front-end (see our note, A risk to UST market's favorite trade, 29 August).

Banks' flows shifting from structural receiving to mortgage paying

It is notable that Sonia underperformance in July/August, relative to its prior US beta, was the most pronounced in the 5y area (the underperformance was also very clear from 1y forward gap curve moves, Exhibit 15). We expect this underperformance to continue as banks' structural hedging flows (receiving fixed) continue to fade in Autumn, giving way for potentially more mortgage hedging (paying fixed) ahead:

• 1H24 retail bank results confirmed that the size of structural hedges should, on balance, be trimmed over the course of this year, after "deposit churn" effects. We think it is likely that most of the hedging activity for this year is complete (without a big change in prevailing yield levels).
• Meanwhile, significant recent cuts in residential mortgage rates points to a pickup in mortgage hedging ahead, other things being equal. According to BoE data, the average two- and five-year fixed mortgage rates peaked in May 2024 (Exhibit 5). Pre-emptive mortgage rate cuts were seen in the weeks leading up to the Bank rate cut in August. Net mortgage borrowing rose £2.8bn in July, marking its highest point since November 2022 and net mortgage approvals for house purchase rose from 60.6k in June to 62k in July, the highest point since September 2022.

The shift in Sonia flows from receiving to paying could lend support to our existing Sonia 2s5s curve disinversion theme, in our view (Sell SFIM9 vs. SFIM6 Sonia, 14 June). Trade entered at -19.5bp with a target of 10bp and stop of -35bp. Current: -1.0bp. Risk to the trade is inflation turning out materially stickier than expected.

To better position for possible shifts in bank flows ahead we recommend a Sonia 3s5s7s fly trade (pay belly, receive wings), which correlates with 2s5s but offers less negative carry than the steepener (-1.4bp/3m by our estimates). The fly also appears slightly rich relative to 2s5s over a 2-year history. We enter the trade at -12bp with a target of 10bp and a stop at -21bp. Risk to the trade is UK government increasing Stamp Duty tax at the Budget, slowing housing market activity.

Low coupon/high coupon: past peak performance, but there might still be juice

In late 2022, we wrote about low coupon Gilts' tax advantages for retail, advantages that would become more valuable relative to, say, deposits, as rates rose (Will retail discover Gilts? A little bit complicated maybe, but not too taxing, 10 November). Retail investors are taxed on income only, making lower coupon Gilts more tax efficient for taxed investors holding them directly (rather than in a fund). An investor paying 40% tax buying a 1/8% coupon issue would pay only 5bp of income tax per annum, as we understand it (we emphasise that we are not tax experts and only discuss this as a potential influence on the market).

Lately, we have been asked whether the tax treatment of Gilts for retail might change at the Budget. It obviously could, but any new system would necessarily be more complicated, we think, making change unlikely (we would say) when direct retail ownership is so tiny (according to official statistics).

We would attribute the recent richening of the low coupon issue in RV to increased scarcity of 1/8% coupon issues with only two nominal Gilt issues remaining (Low coupon/high coupon Gilt spread trade hits target, 5 August). Evidence of a surge in retail appetite for holding Gilts directly is harder to come by (bar anecdotal evidence). Official data reports little evidence of retail appetite for holding Gilts directly. 1Q24 data showed households holding a trivial £3.1bn Gilts directly.

Factors that will influence the low coupon/high coupon Gilt spread trade ahead:

7. Less low coupon Gilt supply from the BoE. It was noteworthy that BoE sales of short-dated Gilts in July (two £800mn ops, 15 July and 29 July) saw £1.1bn of the operation size (in market value terms) being purchases of UKT 0.125% 2028, with most of other Gilts acquired in that operation also low coupon issues. In August, the one short-dated Gilt sale saw £0.63bn allocated to UKT 0.125% 2028. If we are right that BoE sticks with £100bn QT for the year beginning October - implying £900m short Gilt sales per quarter by our estimates - then that potential source of low coupon Gilts will slow to a trickle.
8. Potential increase in appeal following the Budget. Retail demand for tax-efficient Gilts might be influenced by the Budget outcome, with a Chancellor seeking revenue raising opportunities. Any curbing of tax advantages on pensions, ISAs, etc., might increase the relative appeal of low coupon Gilts.

Having closed our previous low coupon/high coupon Gilt spread trade in August, we now enter a new, more tactical, trade for the Autumn, noting that:

• UKT 0.5% 2029 still trades relatively cheap compared to other low-coupon issues while offering a relatively high after-tax yield and free-float (Exhibit 17, Exhibit 18).
• UKT 4.5% 2028 is at risk of reduced demand ahead of the DMO's launch of the new March 2028 conventional Gilt in mid-November.

We recommend selling UKT 4.5% 2028 Gilt to buy UKT 0.5% 2029. We will monitor the trade on a z-spread basis, entering the trade at -8bp, setting a target of
-20bp, with a stop-loss at +4bp. Risk to the trade is a change in the tax treatment of Gilts for retail.

 Rates - JP

• Fed rate-cutting cycles have not always triggered a sustained decline in JGB yields.
• Paying the 1y1y OIS would be one potential near-term investment strategy.

Fed's rate cut does not always trigger decline in JGB yields

The market sees the Fed as likely to begin cutting rates at its 17-18 September FOMC. However, we note that past Fed rate-cutting cycles have not always led to a sustained decline in JGB yields. Even if the Fed begins cutting in September, we doubt this would mark the start of a downtrend in JGB yields.

Start of past Fed rate-cutting cycles and JGB yields

A comparison of the Fed funds rate and JGB yields since FY00 shows that the start of a Fed rate-cutting cycle does not automatically drive down JGB yields (Exhibit 19). We looked at 2yr/10yr JGB yields for the 60 days before and after the Fed's first cut in the past five cycles. While yields fell rapidly in 2001, we think this mainly reflected the slowing US real economy combined with the start of monetary easing by the BoJ. JGB yields also declined during the Fed's 2019 rate-cutting cycle, but we think this reflected growing uncertainty about US-China trade friction.

However, JGB yields rose when the Fed cut rates in 2020. This coincided with the COVID-19 pandemic, and we think the move reflected investors reducing bond positions or selling to generate cash in response to rising financial market volatility and falling liquidity.

Upside risk for JGB yields

To sum up, we think the start of Fed rate cuts would only cause JGB yields to begin falling if this is accompanied by a marked slowdown in the US (and in turn global) economy. However, the rate-cutting cycle that we expect the Fed to kick off at the September FOMC will not be its response to a recession, but rather a move to normalize the undue tightness created by the current policy rate (for details, see Federal Reserve Watch: No Jay-walking at Jackson Hole 2024-08-23). At this point, the start of rate cuts at the September FOMC does not contradict our forecasts for an uptrend in JGB yields.

We see further upside for JGB yields given (1) the expected increase in net supply caused by the BoJ scaling back JGB purchases, (2) high valuations relative to fundamentals, and (3) the potential for additional BoJ rate hikes. July monthly wage data showed another strong result, and the Japan's real policy rate is still deeply negative, which suggests that the BoJ could "adjust the degree of monetary easing" further. We expect JGB yields to gradually rise through year-end as successive auctions result in larger tails. Our end-CY24 10yr JGB yield estimate is 1.25%.

Two scenarios for September FOMC

JGB yields across the curve are now lower than they were in late July. If weak US employment report (scheduled for release on 6 September) makes a 50bp Fed rate cut at this month's FOMC more likely, we think paying the 1y1y OIS when it falls below 0.5% (from around 0.53% now) would be an attractive investment strategy. Given that the BoJ is expected to raise its policy rate by 25bp, we think a decline in 1y1y OIS to below 0.5% would be an undershoot.

Our US economists' base case for the 19 September FOMC is for an initial 25bp rate cut. Given the Japan's LDP leadership election in September and the US presidential election in November, another option would be to target carry returns if the market does not price in changes in BoJ policy in the near term. 10yr JGB #375 appears to offer better carry and roll than the off-the-run 10yr issue. We therefore think holding #375 until the 18-19 December MPM would be an option.

 Rates - CA

•  CAD rates rallied, curve steepened post BoC meeting. Markets expect 25bp cut at Oct meeting, with low risk of 50bp priced

BoC cuts policy rate by 25bps

The Bank of Canada (BoC) cut its policy rate target by 25bp on Sept 4, as expected, to put the rate at 4.25% (E. 4.25%, BofA 4.25%). This is the third consecutive 25bp cut. Additionally, the BoC continues to normalize its balance sheet.

Our economist expects another 25bp cut at the Oct meeting on the back of a continued slowdown in core inflation and weakening labor market. The BoC will have two CPI prints and two labor market reports prior to its next meeting. Our house call is for the BoC to deliver 25bp cuts at every meeting, reaching 3.75% by YE24 & 3.0% terminal by Apr '25.

CAD rates rally, curve steepens

The BoC meeting led CAD rates to rally, with the front end leading the curve steeper. Governor Macklem's guidance of "reasonable to expect further cuts" gave no definitive expectations on the pace or size of cuts, but markets are pricing a 30bp cut in Oct, which we interpret as a 20% chance of a 50bp move. We think markets are justified in pricing a small likelihood of this outcome given that realized growth in June and July missed BoC's more optimistic forecasts. Following Macklem's comments, the hurdle for a 50bp cut still appears high but - like the Fed - it could be considered in case of a significant slowdown in growth and jobs. While the bar to cut 50bp is high, the bar to skip a meeting is also high, especially now that the Fed is about to cut. Although both risks are low, our economist sees a higher risk of a 50bp cut than a skip.

CAD terminal path serves as guide for US terminal

After the string of sticky Canadian inflation prints to start the year, markets were pricing a terminal of 3.4%, but now that inflation risk has subsided and growth fears are rising, the terminal has repriced to 2.46% (Exhibit 22). Pricing of the Fed's terminal has followed a similar path recently as inflation has softened. Softer economic data in the US and expectations for Fed cuts have likely contributed to the market pricing of BoC cuts, especially as concerns about policy divergence constraints have faded. We would expect terminal rates in the US and Canada to continue to converge as the US jobs market softens. This would bring 10y US rates closer to CAD 10y.

QT forecast unchanged

Expectations for QT are little changed according to Senior Deputy Governor Rogers. The BoC is still aiming for settlement balances to fall below CAD$80bn from around CAD$100bn currently. The BoC does not attribute the recent upward pressure in CORRA to the reduction in settlement balances.

  Front end - US

•  Fed cuts won't unlock MMF cash unless much deeper cuts vs market pricing. Inflows likely to remain positive unless Fed cuts <2%

Fed cuts won't drive near-term MMF outflows

Clients have asked about MMF flows & Fed cuts. They ask two questions: (1) will cuts see MMF outflows? (2) where will MMF cash go? Our short answers (a) Fed cuts should see MMF inflows slow but outflows unlikely unless cuts to <2% (b) if outflows, MMF cash most likely shift to other higher yielding fixed income, not equities. Detail below.

Fed cuts won't see MMF outflows unless FF <2%

To assess MMF flows post Fed cuts we looked at the past 5 cut cycles, back to '89 (Exhibit 23). MMF AUM typically continues to grow in the first year of cuts & lower rates may not see MMF outflows. We observe only 3 periods of MMF AUM outflows over a series of months (1) late '01-'04 (2) early '09 to mid '12 (3) mid '20 to late '20.

Historically, MMF AUM growth y/y is typically positive unless front-end rates <2% or curve steep (Exhibit 25, Exhibit 26). Only two cutting cycles saw MMF AUM decline on net vs the start of the cutting cycle (1) '01-'04 (2) '07-'09 (Exhibit 24). In both cycles MMF AUM was positive for the first 2Y of cuts cycle even with FF rate <2% (Exhibit 27, Exhibit 28). Outflows over the '07 cycle were likely exacerbated by a prime fund "breaking the buck" in Sep '08 which reduced investor confidence in prime MMF.

Digging deeper, outflows are typically driven by retail and prime MMFs. Retail MMFs saw larger net outflows or smaller net inflows vs inst'l MMFs in each of the last 5 cutting cycles. Prime funds saw net outflows in 3 of the last 5 cutting cycles vs gov't MMF AUM which was flat or positive in all 5 cycles.

Prime retail MMFs seems most sensitive to cuts. Today prime AUM is impacted by prime institutional MMF reform in Oct '24. Prime vs gov't MMF yields have compressed making prime AUM even more vulnerable to the upcoming Fed cutting cycle.

MMF outflows shift to fixed income, not equities

MMF outflows are uncommon unless rates are <2%. Given the 3 MMF outflow periods EPFR data provides insight on where this money has generally moved. EPFR data is limited in the '01-'04 period so we focus on the latter 2 periods. Note: we cannot directly trace MMF outflows to other funds in EPFR so we simply compare time periods.

In both MMF outflow periods of '09-'12 & '20 EPFR data suggests strong flows into fixed income over equities. EPFR data suggests strongest inflows to "mixed allocation" funds (fixed income benchmark aggregate funds that invest across USTs, MBS, & credit) followed by credit & MBS. We find this intuitive since it represents a modest extension out of very safe MMF holdings & into higher yielding fixed income alternatives. Equity outflows in each period are not surprising since there were sharp equity declines in 6m of each relevant MMF outflow period.

The EPFR data confirms our prior intuition (see Back to school). If MMF investors to leave the product, we expect they will only inch marginally out the credit curve. We are skeptical MMF investors would be willing to shift from a risk free & intra-day liquid investment vehicle into much riskier assets simply because of Fed rate cuts. History supports our view & shows only MMF outflows if money markets <2%.

Bottom line: Fed rate cuts are unlikely to unlock MMF cash unless rates <2%. Fed cuts should see MMF inflows slow but outflows unlikely unless cuts much deeper vs market expectations. If MMF see outflows cash most likely shift to other higher yielding fixed income, not equities. MMF cash should remain sidelined from a risk taking perspective.

  Front end - EU

• We expect €50-65bn MRO/LTRO take-up this month as the refi-depo spread tightens to 15bp, and funding of lower quality assets to shift to the central bank
• We expect euro funding conditions to increase as QT continues, we stay in 2y EUR 3s6s widener and EUR FX-Sofr tightener

The corridor tightens

From 18 September 2024, the ECB's refi-depo corridor will tighten from 50bp to 15bp. This will make the spread between the refi and depo rate the tightest since the ECB's inception (Exhibit 29). We believe the tighter corridor will:

• Help banks meet reserve demand as the ECB continues QT and the remaining TLTROs mature. The current monthly pace of QT is c. € 37.5bn per month and there are €76bn of TLTRO outstanding, which will all mature by the end of the year.
• Reduce negative stigma associated of banks tapping into the ECB's open market operations.

We forecast euro area banks will borrow between €50bn and €65bn from the MRO/LTROs in September 2024. In our view, one justification for banks' borrowing from the central bank may be to satiate reserve demand for regulatory purposes. At the country level, we estimate banks in France, Ireland, Italy, Netherlands, and Spain may need to increase reserve holdings most by the end of the year, as reserve supply in these countries may be less than that demanded for LCR purposes (Exhibit 30).

Bank reserves in one country may be increased 1) by attracting reserves from another bank on a cross-border basis, 2) borrowing from the central bank. We believe banks' reliance on central bank funding will increase over time as QT continues and reduces total reserves in the system. This is likely to cause MRO/LTRO take-up to increase over time and account for c. 30% of total reserves by end-2026 (Exhibit 31).

One risk to our view is the announcement by the ECB of details of the structural longer-term refinancing operations. We expect these refinancing operations to have a longer maturity and higher price than the existing one-week MRO and three-month LTROs, both priced at the refi rate. The announcement of these details, if at attractive terms, may accelerate bank reliance on the central bank for funding. But we believe the ECB would not want to be lender of first resort from day one, may prefer to observe bank take-up of existing OMOs first, and would like to encourage banks to tap the market for funding more before offering a new lending operation to banks.

Some funding to shift to central bank

The tightening of the refi-depo corridor is likely to shift bank funding of some assets from the market to the central bank. Assets that would be funded at market rates sufficiently higher than the refi rate are likely to shift to the central bank, given the ECB applies the same lending rate across all collateral type (haircuts on the assets will vary). Given the €str-depo spread of c. 9bp, we believe the funding of ABS and non-IG assets will shift to the central bank first (Exhibit 32).

We believe the tighter corridor will indirectly cap further cheapness in GC repo, where current market rates imply they would still trading lower than the refi rate when the corridor tightens. We expect Germany one-day GC-€str spread to tighten to c. 5bp in 4Q24 vs 7bp currently.

Stay in 2y EUR 3s6s widener, EUR FX-Sofr tightener

We expect euro funding conditions to tighten over time as the ECB continues to reduce reserves via QT and does not offer banks a full solution to their term funding needs, e.g. to the meet NSFR requirement. In our view, this will make term premium in money markets increase and we stay in 2y EUR 3s6s widener (current: 5.3bp, target: 14.0bp, stop: 5.0bp). Risks are ECB's structural longer-term operations are priced meaningfully below market rates, the ECB reduces QT, bank demand for reserves is lower than expected, and widening pressures only reflected at tenors above 6M.

We also expect the QT outlook between the ECB and the Fed to diverge and cause euros to become more expensive vs dollars. The ECB is set to accelerate QT in 2025 whereas the Fed is expected to have stopped QT by then. We stay in our 9Mx12M EUR FX-Sofr basis tightener (initiated in May 2024 and has rolled since, current: -5.6bp, target: -2.0bp, stop: -10.2bp). Risks are: a strong global risk-off event that creates strong demand for USD, the Fed continues QT for longer than expected, the ECB does not accelerate QT, and the ECB announces details to its structural credit operations that allows euro area banks to satiate their reserve demand under very attractive terms.

 Front end - EU II

• Assets of EUR MMFs at new highs as ECB rate cuts loomed, the decline in WAM may keep Euribor fixing curve steep and cap further cheapening in repo
• We stay in 2y EUR 3s6s widener while being mindful of fixing 

Assets of EUR MMFs at new highs

Assets of EUR MMFs in the euro area rose to record highs. Assets of short-term EUR MMFs, which follow similar rules as US SEC 2a-7 MMFs, exceeded €300bn since July 2024 and continued to rise (Exhibit 33). Assets of French MMFs, mainly standard MMFs that can hold longer maturity assets than short-term EUR MMFs and are mostly denominated in euros, also recorded a record high of €421bn in 1Q 2024.

We believe the inflows in EUR MMFs over the past year reflect growing expectations of rate cuts by the ECB. Curve flattening associated with such expectations may have also reduced the attractiveness of longer-dated assets to investors vs parking cash at MMFs. Bank interest rates on new deposits have also fallen below yields of short-term EUR MMFs (Exhibit 34).

Short-term EUR MMFs reduced their WAM since July 2024 to c. 30 days (Exhibit 35). The reduction in WAM may raise MMF demand for assets with shorter maturities, including CP and CD that collectively account for c. 50% of short-term EUR MMF assets (Exhibit 36). Newly issued CP and CD by Euribor panel banks at shorter maturities to meet this demand may feed into the Euribor fixings and keep the fixing curve steep (Exhibit 37).

The reduction in WAM has also been achieved by increased allocation of MMF assets to repo, which tend to have among the shortest maturities. The share of short-term EUR MMF assets in repo rose to c. 15%, a decade high; while the share of TD declined (Exhibit 38). Increased demand of MMFs to lend cash in repo may cap further cheapening in repo rates. We expect Germany one-day GC to tighten to c. 5bp vs €str in 4Q 2024 from 8.5bp currently.

We stay in 2y EUR 3s6s widener because we expect bank demand for longer term funding to rise and cause term premium to build at the front-end of the curve (current: 6.7bp, target: 14.0bp, stop: 5.0bp). Risks are ECB's structural longer-term operations are priced meaningfully below market rates, the ECB reduces QT, bank demand for reserves is lower than expected, and widening pressures only reflected at tenors above 6M.

The decline in short-dated EUR MMF WAM may keep the Euribor fixing curve steep and be a source of support for our 2y 3s6s widener recommendation. But we are also mindful of days with no or low eligible Euribor transactions, as it is possible that fixing risks to our recommendation may be exacerbated by the current phasing out by mid-November 2024 of level 3 contributions to Euribor submissions.

  Inflation - US

•      Express long bias in nominal vs real yields: expect that drivers of lower RY will also likely weigh on BEs, making nominals more efficient trade
• Recommend initiating an inflation swap curve steepener trade 2 years forward using 10y & 30y inflation swaps

How do you take your long?

Our long bias in US rates remains focused in nominals vs real yields. The macro case for preferring real yield longs vs nominals comes in the scenario where stagflation is a greater risk. While recession/ hard landing is not our economists' modal scenario, the market is likely to continue to price downside risk to growth as the economy cools which can weigh on inflation compensation (see: Too soon to re-engage inflation longs) and makes owning nominals a better trade.

What will change our mind? In short, a dovish Fed that is encouraged to cut with still sticky inflation data. This could stem from a more persistent CPI/ PCE wedge driven by persistently higher shelter components or an emergence of commodity pressures that the Fed may look through. A pickup of material tariff/ inflation re-acceleration risks on election events, which we now think the market assigns lower likelihood to (see: Close 1y1y inflation short), could also support a clearer justification for owning real yields vs nominals. This is especially true if upside inflation risks coincide with more downside growth risk as we discuss in US election section of Back to school cheat sheet.

If we were to pick a spot to favor long real yields, we would recommend further out the curve. Exhibit 39 shows the percentile of the real yield / nominal yield ratio which is elevated across the curve, particularly at the 30y tenor (95^th percentile since 2005). As discussed in Stretched positioning a risk to UST market's favorite trade, most of the strong bid we have observed since last month's payrolls report has been concentrated in the front end of the curve and it will likely require a worsening of the growth outlook for investors to extend further out. This is a core factor that makes us hesitant to recommend long RY vs nominals: a worsening growth outlook that could see longer-term RY rally would likely see BEs compress as well.

This historically high RY component of the nominal has been the result of a pickup in breakeven beta to the change in nominals alongside the recent rally. As shown in Exhibit 40, breakeven betas are now at some of the highest levels observed recently which was concurrent with the rally in rates and market re-assessment of downside risks.

Inflation term structure has room to steepen

With greater conviction in lower near-term inflation and investors pricing greater potential for a slowing US economy, the inflation curve has steepened from historically inverted levels observed in 2022 (Exhibit 41). Much of this repricing has come from a compression in spot inflation, as inflation forwards are little changed by comparison.

We think there is room for the inflation term structure to continue to steepen, but are wary of near term upside risks to spot. These risks to higher spot include oil carry trade unwind driving oil higher and data stabilization reducing downside risk.

Our preferred inflation swap steepener expression is a 2y forward 3s28s steepener (expressed using 2y, 5y, 30y inflation swaps). As shown in Exhibit 42, levels are still historically cheap and have room to widen closer to the average term structure observed since 2005. We initiated this trade at 0bps (current level = 4bps), targeting 30bps with a stop of -15bps. The key drivers of this trade form a macro perspective are the market pricing more risk for inflation over the longer term (as it historically has) and/ or perceiving less upside risk in belly forwards. The key risk would be a re-emergence of inflation risk in 2-5 years that the market views as only temporary.

Bottom line: While real yields are elevated vs nominals, particularly further out the curve, we maintain our long bias in nominal yields. What will drive longer term real yields to rally more will likely also drive a compression in inflation compensation. We initiate a forward starting inflation swap steepener given a historically flat term structure that can correct on market perception of upside inflation risk further out the curve vs levels in 2-5 years.

 Inflation - EU

 OATei 30y to outperform 20y on ASW, real yield and BE

• Recent moves and our new Bund call make us more constructive 30y linkers and inflation. Expected supply tilt and index events favour OATei 2053s over 2043s.

We now like 30y Bunds. What does it mean for linkers?

We have recently been expressing our Euro inflation views in derivatives. Firstly, there's our belief that the inflation curve should steepen, with our economists' call for inflation to average 1 ½% next year and our view that the inflation curve will retain a large term premium (because it will see some of the current inflation decline as cyclical and because it will perceive the tail risks of future inflation shocks as skewed to the upside). To us, it's the 2s5s (or 1y1y/2y3y) piece of the curve that needs to steepen the most. Secondly, there's our desire to receive real swap rates in Euro - outright and relative to UK - believing that falling inflation will reset perceptions of "real neutral" rates lower.

Now it's time to add a bond linker trade. In this we take a lead from our new call to buy 30y Bunds outright (see Rates-EU section). However, it's worth making a few broad observations about the long end of the inflation and real swap curves first.

Exhibit 43 shows that although the long end of the inflation curve maintains a meaningful premium, it has moved sideways as the front of the curve has steepened. The steepness of the inflation curve beyond 10-years no longer looks odd (to us) relative to the steepness out to 10-years. Exhibit 44, showing 30y "real €str" rates, suggests a stickiness around zero, which (with our bullish bias) we tend to regard now as a soft ceiling. This is particularly true after recent long-end underperformance versus US real yields - a sub-100bp real spread vs. US is not something we regard as wide, given the US/Euro potential growth differential, the divergence in their International Investment Positions (i.e., the US's increasing dependency on imported capital), and the scarcity value of long-dated Euro linkers (and its corresponding impact on swap availability).

France's comparative advantage in long linkers

Light issuance of long linkers means that investor need for French paper, notably for periodic index rebalancing, dominates over an inflation swap receiving bias. Italy and perhaps now Germany are the preferred sources of manufactured inflation swap flows. And Germany's market withdrawal has compounded the long-end scarcity issue.

The zero iota for OATei 2043 is striking compared to Italian and German peers (Exhibit 47). That understates its richness because its comparator has a much shorter duration, as Exhibit 45 shows. Also noteworthy is that the cheapening of OATei z-spreads and shape of the spread curve has catapulted the OATei 2043/53 forward z-spread to 160bp.

First supply, then demand to richen 30y versus 20y OATei

New long-dated French linkers tend to be left untapped for a time to season, and the new 2043s hasn't been auctioned since its launch in May. OATei 2053s, by contrast, has been tapped three times since May. We expect 2043 auctions to begin soon, perhaps displacing 2053s at times, which should rebalance supply in favour of the '53s. Then, in early 2025, we have index exits of OATei 2026, DBRi 2026, and BTPei 2026 in consecutive months from March. We expect this to drive bullish flattening in the real curve and an additional force for OATei 2053 outperformance versus 2043s on ASW.

On Tuesday, in the above-linked note, we recommended buying OATei 2053 on asset swap versus OATei 2043. Monitoring the trade as the difference between the two z-spreads, we entered at +29bp, setting a target of +15bp and a stop-loss at +37bp (currently +28bp). Risk to the trade is poorly digested 30y supply. We also expect OATei 2043s/53s real flatteners and breakeven steepeners to do well.

 Inflation - UK

•  We rationalise UKTi trades, closing three related ones to replace with a 2032/36/47 cash-and-duration neutral barbell (long wings) that captures the essential elements of the ones closed. Stay short UKTi 2036s on ASW versus UKT 2042s.

Three into one

Over time we have accumulated four different trade ideas with related features, all involving short positions in UKTi 2036s (an issue we regard as particularly expensive on the real and asset swap curves).

Exhibit 49 shows the real yields of UKTi maturing in November, with forward real yields between issues. We show only Nov maturities in order to both simplify the chart and sidestep the need to seasonally adjust the real yields to calculate fair forward yields. The three circled forward rates are: UKTi 2027s to 32s, at 49bp (green); 2032s to 36s, at 129bp (purple); and 2036s to 47s, at 190bp (orange).

The first step up of 80bp, from green to purple, largely represents market pricing for the impact of RPI reform in 2030, but the second step of 61bp, from purple to orange, is almost wholly about asset swap cheapening (i.e., not observed in the real swap curve). The fact that forward rates step down sharply beyond 2047 increases the perception that the 2036 to 2047 forward yield plateau is a large anomaly (albeit a longstanding one), presumably reflecting investor habitat preferences. It is surprising to us that the popularity of asset swap repackaging of the 2047 issue has done little to temper this.

In order of introduction, these are the four trade ideas that relate to this market feature in different ways. We will close the first three of them now:

In the September 2023 Inflation Strategist, we recommended a UKTi 2036s/2047s real yield curve flattener at 55bp, targeting 30bp, with a stop loss at 70bp. We close at the current spread of 51bp.

In the 2 February 2024 Global Rates Weekly, we noted that the forward real yield plateau between UKTi 2036 and 2047 in the November maturity issues curve (still present in Exhibit 49) was materially higher than the closest equivalent in the March issues curve. We recommended investors pay the forward yield between UKTi 2034 and 2046 and receive the UKTi 2036-to-2047 forward, entering the trade for a pick-up of 24bp, targeting 8bp with a stop-loss at 32bp. We close the trade at the current pick-up of 16bp.

In the April 2024 Inflation Strategist, we recommended a UKTi 2032-36-42 cash and duration neutral barbell (+35%/-100%/+65% weights) - effectively a forward real curve flattener - at a pick-up of 13.6bp, targeting 5bp, with a stop loss at 18bp. We close the trade at the current pick-up of 11.8bp.

Finally, and most recently, in the July Inflation Strategist, we recommended switching out of UKTi 2036 to buy UKT 2042 on asset swap, monitoring the trade as a spread of z-spreads. Although the bonds were six years apart in maturity terms, their duration difference was only 0.4. This is UKTi 2036's closest comparator in duration terms. We entered the trade at a spread of -21bp (UKTi less UKT), targeting -8bp with a stop-loss at -28bp. We keep this trade open (current spread -21bp). Risk to the trade is poor demand for nominal auctions in the area.

The replacement trade

The trade which we think captures the essential attractions of the three we now close is a UKTi 2032/36/47 cash and duration neutral barbell, which effectively "sells the step" between the two forward real rates circled in purple and orange in Exhibit 49, because it is a forward real flattener when weighted this way. Exhibit 51 shows the full history, with duration weights adjusted daily to keep the spread cash-and-duration neutral throughout.

We recommend a UKTi 2032-36-47 cash-and-duration neutral barbell, long wings, with +43.8%/-100%/+56.2% risk weights, picking-up 14.8bp. We set a target of 5bp and a stop-loss at 20bp. Risk to the trade is idiosyncratic moves around supply events in illiquid conditions.

     Technicals

•  Friday matters. US U-rate is in an uptrend and has risen four times in a row pushing yields lower, curves steeper, gold & euro higher, USDJPY lower. US10Y yield's short-term consolidation is breaking to the downside ahead of NFP. An NFP miss pushes it lower now vs later while a beat creates an opportunity to buy later this month.
• In H2 we remain bullish USTs and still see steeper curves. Buy UST dips for landing. Softer landing is US 10y to 3.50% / 3.22%. Harder landing is 10Y < 3.00%. Risks: US election year and ten-year seasonal trends support 10Y yield. A positive NFP revision and/or beat may lead to daily chart yield bottoms and September upside.

US 10Y Yield: Pushing lower ahead of NFP

The decline in US 10Y yield to 3.66% on Monday August 5^th marked a key low for the short-term trend. In our view, this was likely the end of a wave 3 or subwave (iii) and the start of a consolidation period in the 3.75-4.00% area. A wave 5 or subwave (v) should follow next to lower yield lows such as a retest of 3.66% and likely 3.50%. Recall the 50d SMA crossed below a declining 200d SMA on July 25th at 4.24%. Yield tends to be below this level 25-65 trading days later or August 29 through October 2024. For US 2y yield and US 2s10s, please see the Technical Advantage 04 Sept 2024. For seasonal risks to our view, please see the technicals section of the US Rates Viewpoint: 04 Sept 2024.