517,000 jobs and the end of the rate-cut dream

The Bureau of Labor Statistics reported 517,000 net new payroll jobs for January, and the unemployment rate fell to 3.4 per cent, its lowest since May 1969. Bond traders had spent the winter building a tidy story of painless disinflation and Federal Reserve rate cuts by the middle of the year. That story did not survive the morning. Hiring in America is speeding up, and a bond market with no protection against that had to reprice where it thinks rates will peak, and fast.

A week of mixed signals

The report landed two days after the Fed raised its policy rate by a quarter of a point, to a range of 4.5 to 4.75 per cent, slowing the pace of increases for a second meeting in a row. Jerome Powell, the Fed's chairman, had acknowledged at his press conference that disinflation had begun, and markets took that as a signal that the tightening cycle was almost over. Stocks and bonds rallied. By the end of the week the same markets were confronting evidence that the economy the Fed is trying to cool is heating up instead.

A surprise too large to dismiss

The size of the surprise wrecked every forecasting model on the Street. Even allowing for seasonal benchmark revisions and odd weather, hiring was strong across hospitality, healthcare and professional services, which points to an economy with plenty of momentum left.

The quirks are real, and worth taking seriously. January figures are always distorted by the end of holiday hiring and by seasonal adjustment, which this year produced a larger than usual boost, and a mild winter kept construction and outdoor work going. The annual benchmark revision also changed the history. But the previous two months were revised up, not down, and other data told the same story: the ISM services index, released on the same day, bounced back above 50 after a brief dip into contraction territory in December.

Paychecks keep demand alive

The figure also undermines the claim that earlier rate rises had pushed financial conditions far into restrictive territory. With half a million people added to payrolls in a single month, household income keeps growing quickly, and that income supports the kind of spending that stops core services inflation from cooling.

The mechanism is straightforward. Services such as restaurant meals, health care and travel are labour-intensive, so their prices track wages closely. When employers are competing for workers, pay keeps rising and the extra income is spent on the same services, which keeps their prices rising too. Job openings still number close to two for every unemployed American, which suggests the competition for workers is far from over.

The Fed will not cut rates while the labour market is this tight. Its biggest worry remains a wage-price spiral that slips out of control. Officials have also said repeatedly that they intend to hold rates at their peak for some time, and their own projections in December showed no cuts this year. Markets had simply chosen not to believe them.

The 1970s lesson

That fear has a history. In the 1970s the Fed repeatedly eased policy as soon as unemployment rose or inflation dipped, and each time inflation came back higher than before. The cycle was broken only when Paul Volcker pushed rates to punishing levels at the start of the 1980s, at the cost of two recessions. Mr Powell has invoked that period often, and the current Fed would rather keep rates high too long than repeat the mistake of cutting too soon.

Traders reprice the peak

Fixed-income portfolios paid the price immediately. Short-term swap markets dropped the 50 basis points of cuts they had expected for late 2023, and forecasts for the peak rate moved well past 5 per cent.

The move was sharpest at the short end. Two-year Treasury yields, which track expectations for Fed policy over the next couple of years, rose more on the day than ten-year yields, deepening an inversion of the yield curve that was already the most extreme since the early 1980s. An inverted curve has preceded most American recessions, but it also means the market still expects the Fed to cut eventually, just later than it hoped.

Yields rose along the Treasury curve and dragged equity valuations down with them. The losers are the investors who bet most heavily on an early pivot: holders of long-duration bonds bought for a rally in rates, and owners of unprofitable growth companies whose valuations depend on cheaper money. Borrowers with floating-rate debt, including private-equity-backed companies and households with adjustable mortgages, face higher costs for longer. Savers holding cash and short-dated Treasury bills are the winners, earning yields not seen in fifteen years with little risk of capital loss.

Where the optimists may be right

The best counterargument is that jobs and inflation can part company. Average hourly earnings have been rising more slowly than a year ago, and if productivity picks up or more workers return to the labour force, employment could keep growing without pushing wages and prices higher. Goods inflation has already fallen sharply as supply chains have healed, and rents on new leases have been cooling. A soft landing with strong hiring is possible. Annual wage growth in the same report also slowed a little, a hint that employers are finding staff more easily than they were a year ago.

But that outcome would still not bring rate cuts this year. It would simply mean the Fed holds rates at their peak for longer without doing much damage. The data that could shift the call are the January consumer price index, due next week, the next jobs report in early March, and the Fed's March meeting, which will bring fresh projections from its policymakers.

The 517,000 print has probably killed hopes of a rate cut this year. The Fed is unlikely to declare victory over inflation until the jobs market shows real cracks, and on this evidence that will take several more months. Expect at least two more rate rises by the summer and no cut before 2024, with long-duration bonds and unprofitable growth stocks bearing the brunt of the repricing.