Markets live and die by economic data, but what happens when investors start to doubt the numbers?
That question is at the heart of Goldman Sachs’ latest Top of Mind, which tackles growing concerns about the reliability of US statistics after hefty downward revisions to payrolls and the dismissal of the Bureau of Labor Statistics (BLS) Commissioner.
Goldman brings in voices from across the spectrum. Erica Groshen, a former BLS Commissioner, insists revisions are “features, not bugs” of a system built to balance timeliness with accuracy.
Large changes, she argues, often signal turning points in the economy rather than flaws in the process. But she also warns that budget cuts and declining survey participation are eroding the granularity and resilience of official statistics.
Harvard’s Alberto Cavallo draws a starker parallel, likening the politicisation of the BLS to Argentina in the 2000s, when official inflation data became a national joke.
He points out that survey response rates have plunged, with less than half of US businesses now returning employment surveys, a structural weakness compounded by shrinking budgets. Trust, he says, once lost, is painfully slow to recover.
Arthur Laffer, by contrast, sees merit in disruption. He praises Donald Trump’s shake-up of the BLS, arguing that entrenched processes needed re-examining after the pandemic scrambled seasonal adjustment models. In his view, statistical agencies have enough money and talent; the problem lies in stale methodology, not resources.
For markets, the implications are not academic. Goldman’s Michael Cahill warns that scepticism over US data adds another reason for investors to diversify away from the dollar, already on the back foot this year.
William Marshall, its rates strategist, highlights the risk for Treasury Inflation Protected Securities, which are directly tied to official inflation figures. If investors stop trusting the numbers, their value as a hedge diminishes.
Currently, global statistics remain broadly reliable, but the warning is clear.
If trust in the numbers fades, the economic cost will be real and lasting.
Markets can price risk, but not if they cannot believe the data in the first place.