Of course, MOAR is coming. But not before an epic rout in everything! This is not bullish bonds enough for me but it is in the right direction. BofA with the note.
US rate forecasts revised lower, with less-aggressive Fed
The US rates team is making substantial downward revisions to our rate forecasts following our US economics team’s new call for a mild 2022 recession and lower Fed funds rate path. We are lowering our 10yT end ’22 forecast from 3.50% to 2.75% and end ’23 forecast from 3.25% to 2.50%. Our new forecasts are very bullish vs the forwards given the expectation for Fed rate cuts in 2H23 and 1H24 (Exhibit 1).
Despite these material US rate revisions, our near-term core rate views are unchanged: (1) stay in flatteners because of sticky front-end Fed pricing and lower long-end rates with downside growth risks, (2) remain constructive duration expressed through 6m10y receiver spreads, and (3) hold bullish longer dated real-rate positions.
One important shift in the US rate view is on quantitative tightening (QT). Our base case is for the Fed to cease QT with the first rate cut in Sept ‘23 and reinvest mortgage proceeds into USTs, similar to its approach in July ’19. This would support UST richening and spread widening; as a result, we close our recent 2yT vs OIS cheapening trade.
US economist revisions take rate forecasts lower
Our US economics team recently changed its forecast to reflect a mild US recession in 2022 and materially lowered the Fed funds rate path. The terminal Fed funds rate was lowered from 4.00-4.25% to 3.25-3.50%, and our economics team now expects 100bp of Fed rate cuts between Sep ’23 and Jun ’24 (Exhibit 2).
Our US economics team revision justification: incoming data point to a loss of momentum, suggesting the US economy may have contracted in the first and second quarters this year. The largest driver of this revision is the decline in services spending, likely due to the erosion of real spending power from higher-than-expected inflation.
Financial conditions have also tightened faster than expected in response to a more aggressive Fed path. The mortgage market in particular has seen a slowing of home sales and housing starts due to a rapid increase in mortgage rates, especially relative to 10y UST yield. We still expect a 75bp hike in July and a 50bp hike in September, though the risks are now skewed higher following Wednesday’s CPI print. The slowing economic outlook means the Fed will likely not need to keep hiking next year, hence the lower terminal rate. Eventually, the Fed may need to move from a contractionary stance to neutral, starting 2023, by cutting its policy rate (see US Economic Viewpoint).
The lower Fed rate path and rate cuts in ’23-’24 are the most relevant aspects of our US economics team changes for the US rate path. We are lowering our 10yT end ’22 forecast from 3.50% to 2.75% and end ’23 forecast from 3.25% to 2.50%. Our new forecasts are very bullish vs the forwards given the expectation for Fed rate cuts in 2H23 and 1H24 (Exhibit 4). Our prior rate forecasts initially reflected a bullish rate bias vs the forwards but these revisions make us more constructive vs the current market.
US rate forecasts lower, but key convictions remain
The shift in views does not materially alter our core macro views. We continue to recommend the following:
(1) Curve flattening: we have recommended clients remain in curve flatteners because of sticky front-end Fed pricing and building downside growth risks that would be concentrated in UST back-end. The recent softening of US economic data likely brings forwards the timing when the UST curve will shift from a bullflattening bias to a bull-steepening bias as the Fed pivots dovish.
We have recommended curve-cap and cap-spread positions for a potential UST curve shift towards steepening at forward horizons (see: summer liquidity drought and UST curve nearing inflection).
(2) Constructive duration bias: the BofA rates team has been early in recommending clients lean long duration at the UST back-end to position for a US growth slowdown (see lean long 10Y USTs). We have retained a generally bullish duration bias, expecting an eventual inflation moderation and reduction in demand from a Fed-induced sharp tightening of financial conditions. We have also flagged expectations for the market to gradually shift focus from inflation to recession concern. We recently reiterated this view via 6m10y receiver spreads (currently +2bp, risks limited to the upfront premium – see vol views).
(3) Hold bullish long-dated real-rate positions: we continue to hold our 10y20y real rate long and believe back-end real yields should decline in a stagflation environment (see: Got stagflation? Go long 10y20y TIPS). With the revisions to our nominal 10y forecast for the year-end, we also take down our 10y real and breakeven forecast to 0.25% and 2.50%, respectively. While breakevens have fallen materially with commodity prices, we believe the market is optimistic about how fast inflation will converge to target (see: Inflation priced to TIP over).
Fed cuts => QT early end, bullish for spreads and USTs vs OIS
Our new base case for Fed cuts in 2H23 is expected to occur with an end to QT. We expect the Fed will stop QT with rate cuts due to the contradictory signal it sends on monetary policy and to simplify policy communications; the Fed will likely not want to be easing with rate cuts but tightening with QT.
The Fed has also established a playbook for such an action in 2019. Recall, the Fed cut started cutting rates in July ’19 and simultaneously announced a cessation of QT.
The ’19 QT cessation resulted in the Fed shifting MBS reinvestments up to the monthly redemption cap into USTs purchased in the secondary market.
It is possible the Fed chooses not to follow the ’19 cut and QT end playbook, but we believe this is unlikely. A few potential arguments for not ending QT with a 2H23 rate cut: (1) the Fed will only be “normalizing” front-end rates after a period of elevated inflation, (2) the Fed balance sheet will likely still be far from reserve scarcity. We see the logic of each argument but believe it will be difficult to prove ex-ante. We also expect the Fed to opt for the simplest communication: stop QT when cutting rates.
An earlier QT end would have several UST implications: (1) reduced front-end UST cheapening pressure with more cash / less collateral in financial system, (2) potentially more UST coupon cuts due to lower financing needs, and (3) fewer secondary market UST dislocations due to a re-start of Fed purchases from MBS => UST reinvestment.
Each of these would tend to bias UST-SOFR spreads wider across the curve.
Fed QT that is stopped in Sept ’23 will result in $1t+n less balance sheet reduction vs our prior estimates through end ’24 (Exhibit 5). Over a similar period, early QT end would result in $780b less UST financing need (Exhibit 6) + $350b of additional Fed UST demand (from Fed MBS paydowns reinvested into USTs). Such a large UST financing shock would likely result in initially less bill supply and eventual additional coupon cuts.
The risk of early Fed QT end may marginally encourage the UST to continue with its current slate of coupon reductions and even larger 20yT cuts (see UST 20Y dislocations).
The sharp change in our US economist Fed view and rate cut / early Fed QT end risk leads us to close our recently initiated 2yT vs OIS cheapening trade (see dislocated US front end). This trade was initiated when the 2yT vs OIS was -9bp and has since cheapened to near zero. We close the trade since it has rapidly moved in our expected cheapening direction and since the logic is not as compelling with risk of an earlier QT.
The position is now less dislocated according to our relative-value measures. We still expect the US front-end to cheapen with current QT, increased FHLB supply, and dealer balance sheet constraints, but prefer to trade this theme more tactically in the near term.
Bottom line: our US economist revisions to the growth outlook and the Fed path have caused the US rates team to materially revise its rate forecasts. However, our near-term core macro views are unchanged: (1) stay in flatteners because of sticky front-end Fed pricing but downside growth risks concentrated in UST back-end, (2) remain constructive duration expressed through 10y receiver spreads, and (3) hold bullish longer dated realrate positions.
The largest change in market view comes as a result of a potentially earlier QT cessation. We close our recently established 2yT vs OIS cheapening position. We still believe in near-term dynamics to support 2yT cheapening but prefer to trade the position more tactically with reduced QT tailwind through end ’24.