A chilly January in the job market

Unlock the Editor’s Digest for free

Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.

Good morning. It was a wild and strange week last week. On Friday the S&P 500 ended up about where it started on Monday, despite some ugly job market numbers and a software industry meltdown. Some thoughts on each of those below. The second piece is by a guest, the FT’s indefatigable US M&A reporter Oliver Barnes. Send us your thoughts: [email protected].

The job market

January’s jobs report, delayed by the partial government shutdown, will land on Wednesday, rather than last Friday as planned. In its absence, the market paid more attention to some less comprehensive labour market measures, and got spooked.

The Challenger Jobs report received a lot of attention: it counted 108,435 lay-offs last month, the highest level for any January since 2009:

But the report, pasted together from public lay-off announcements, is not scientific (to put it mildly). Unhedged is inclined to shrug it off. Less so the ADP report, which showed private payrolls only increased by 22,000, half of what economists had expected. And what increase there is continues to occur mostly in non-cyclical sectors such as education and health services, which do not speak to economic strength. But, again, Unhedged takes it with a grain of salt: ADP’s figures don’t correlate tightly with the BLS’s private payrolls estimates, which are what the market really responds to.

Which leaves a report that Unhedged does take very seriously: the Jobs Openings and Labor Turnover Survey. The latest JOLTS report showed just a 0.1 per cent increase in hiring to 3.3 per cent in December, which was mostly driven by large firms, while the hiring rate among small businesses fell. Historically, there’s been a solid positive correlation between changes in the hiring rate and economic activity. Chart courtesy of Matthew Luzzetti at Deutsche Bank: 

Labour demand has been cooling gently for months, but this has mostly manifested through fewer job openings rather than a significant rise in the unemployment rate. Troy Ludtka at SMBC Nikko Securities believes that this cannot last:

The unemployment rate is likely to rise more meaningfully now that the job openings rate has fallen below the critical 4.5 per cent threshold — a key level Fed Governor Chris Waller highlighted in January 2024. This 4.5 per cent level is also where the Beveridge Curve changes slope, indicating that a further reduction in job openings should translate more directly into higher unemployment. 

As we have noted many times before, strong overall growth in the economy sits uneasily alongside a stagnant job market. The crucial question is whether employment will pick up to match the GDP growth, or GDP will slow to match employment. The data from last week, showing deepening stagnation, makes the question more pressing. Still, one month is just one month, and we will learn more on Wednesday.

CoStar, AI, software and activist investors

AI panics come in different flavours. A year ago, in the DeepSeek panic, the worry was that competitive AI models could be built quite cheaply, so returns on big investments in AI would be low. Last week came the Anthropic panic. This time the worry is that AI tools will undercut the business models of software and analytics firms.

The DeepSeek panic faded fast. Will this one? Looking for an answer, you could do worse than follow the case of CoStar.

It was coincidental that hedge funds Third Point and DE Shaw went public with activist attacks on real estate data company CoStar just as the Anthropic panic blossomed. But the timing is revealing. CoStar is best understood as the Bloomberg terminal of real estate data, a repository of information about the commercial sector, from vacancy rates to construction permits and more — all packaged and sold to brokers, developers and the like. The shares were down 18 per cent last week, leaving it with a market value of just over $21bn, a victim of a sector rout that wiped $560bn off the S&P 500 software index.

Both Third Point and DE Shaw have held stakes in the company for at least a year, and they agree on where CoStar is going wrong. They argue that the company is weakly governed and made a mistake when it launched a residential real estate platform, Homes.com, at a cost of more than $3bn.

Last year, the hedge funds agreed to a brokered peace in exchange for two board seats and a review of the Homes.com investment, which decided to keep it. They have since concluded that a bigger reshuffle of CoStar’s board is necessary, setting the stage for a nasty proxy fight. CoStar barked back, saying the pair of hedge funds have “latched on to a manufactured tale of governance shortcomings”.

Homes.com has yielded just $80mn in annual revenue, far below the company’s previous target of $700mn-$1bn. Worse, Third Point argues “these direct financial costs have been compounded by broad management distraction which has prevented the core [commercial real estate] franchise from reaching its full potential”. It’s hard to argue that the company’s strategy has been a smashing success. Even before last week’s drop, the stock has been trading sideways since 2019.

Whether the activists or CoStar’s leaders were right about Homes.com before the events of last week, management’s plan to make CoStar more of a “software-as-a-service” company looks different now. CoStar’s secret sauce is its data. By pushing into residential real estate, it entered into an area where little or no data is proprietary. Homes.com is an interface on top of the same property listings available on Zillow and other sites. The margins in all of these businesses are thin. What is to be the fate of data processing — as opposed to data owning — businesses in the age of AI? Third Point and Shaw think it is time to get out.

This debate is likely be settled in a fight for control of CoStar’s board when the nomination window opens next month. At stake is not just the future of one company, but the question of what makes a software company valuable.

One good read

How we got so angry.