After an unexpectedly strong January, the broad-based rally that risk assets have enjoyed since October last year has finally paused for breath, and we have seen a reasonable retracement in many sectors within fixed income. The recent sell-off has been driven partly by a rise in the market’s terminal rate expectations, but perhaps more importantly by a re-alignment of market prices with the ‘higher for longer’ rhetoric coming from the major central banks.
The adjustment has been particularly acute in US Treasuries. The UST curve is now pricing in a terminal Fed Funds rate of 5.4% around the middle of 2023, up from 4.9% at the beginning of February. The move in longer term rate expectations has been even more pronounced, with the 5% Fed Funds rate now expected in January 2024 having been as low as 4.17% just a few weeks ago.
Only part of this shift is justified, in our view. Markets had previously been pricing in three or four cuts from the Fed by the end of this year, which always seemed rather ambitious even after a raft of good news regarding inflation. Investors have finally relented to the Fed’s hawkish messaging and it makes sense that yields react accordingly.
However, more than anything it was the blowout January non-farm payrolls data (517k jobs added vs. 190k consensus) that seemed to spark this mini-correction, and what was interesting about those numbers was the size of the seasonality adjustment.
In a nutshell, companies hire more people towards year-end to deal with the extra demand of the holiday period; more goods are purchased and need to be transported, and more services are also provided with more corporate and family events taking place. Come January, companies adjust their workforce back to ‘normal’ and this happens at the same time every year. The Bureau of Labour Statistics (BLS) therefore applies a seasonality adjustment. This is standard practice and essentially the objective is to give people a ‘cleaner’ number, i.e. one that strips out the obvious fluctuations due to the time of the year at which the data is being collected. The raw (non-seasonally adjusted) change in NFPs for January was a decline of 2.5m jobs, while the seasonally adjusted number showed an increase of around 0.5m, an adjustment of some 3m jobs.
We are not questioning the validity of the adjustment, but in our view the correct interpretation of January’s number was that companies let go of fewer people than usual, as opposed to the companies hired far more people than usual narrative that took hold in some quarters.
The former clearly sounds less bullish, and to us doesn’t represent the start of a new trend. The more likely explanation is that given the difficulties they faced when rehiring post-COVID, US firms have become more reluctant to reduce their workforce. Before the January NFP print the trend was a decline in job creation from the peaks of mid-2022 down towards a figure of around 200k-250k a month. That is still a high number, and it will need to keep coming down before the Fed is comfortable it has taken the heat out of the labour market, but it is very different from 500k per month. In addition, the NFP numbers have been subject to some big revisions over the last 12 months, and while it is almost impossible to predict what the figure for February will be when it is published on March 10, it is quite possible we will see a large reversal.
Inflation and activity data have also played a role here. While month-on-month core and headline inflation were both in line with expectations in January, the numbers at 0.4% and 0.5% respectively were relatively high compared to recent prints. Core goods prices disinflation continues but at a slower pace, while services inflation still showed elevated prints for Shelter and Owners’ Equivalent Rent (OER). If we look at the subcomponents of the ISM manufacturers’ index, we can see that Manufacturing Prices continue to decline but while Services Prices continue to weaken, the downward trend has stalled somewhat. We should point out that this is more or less what the Fed had anticipated in recent months. Goods disinflation still has some space to continue declining, but the pace of decline will likely slow as lingering pandemic effects wear off, while Services inflation will be stickier both due to Shelter and OER and due to other services having more to do with wages in the context of a hot labour market. Regarding growth, Retail Sales were strong and more recently the Markit PMI Services index unexpectedly bounced back above the 50.0 mark (the threshold indicating growth) for the first time since June. While we do not disregard this data, we do believe that warmer weather might have had an impact on Retail Sales, while the Markit PMI reading has now moved a little bit closer to the ISM Services survey after a very large recent divergence between the two.
What does this mean for our view on the US economy, and the Fed? We thought the rate cuts being priced into curves in January were likely to correct at some point, particularly given the consistent messaging we have heard from the Fed for the last few months. Inflation was never going to moderate in a straight line – the Fed chair, Jerome Powell, said himself at his press conference earlier this month that the monthly declines we saw at the back end of last year could strengthen slightly in the first half of this year as the rapid goods deflation we saw post-COVID began to normalise. However, we still expect inflation to move lower as the stickier parts of the inflation basket start to feel the effects of the 425bp of rate hikes put through the system last year, and we expect growth to be well below trend in 2023. Importantly, given terminal rate forecasts that are now significantly above the Fed’s estimate of the neutral rate, any change in these forecasts becomes more incremental, and markets have reacted well to this latest repricing in rate expectations. All-in yields remain elevated versus historical levels, and in our view they continue to look attractive here given the more settled macro picture that has emerged since the start of the year.
We have seen a reasonable retracement in asset prices on the back of a few data points, but we don’t think those data points will change the way the Fed will conduct policy, and it also looks a bit more likely than usual that we will see some reversion in February’s NFP number as well as some activity numbers such as retail sales. Data dependency remains the name of the game.