Please see below, an article from EPIC Investment Partners which discusses the health of the consumer credit market in the US. Received today – 29/05/2026
American consumers appear healthier than they feel. Payrolls are still growing, unemployment remains low and household wealth is close to record highs. Equity markets trade as if the economy has absorbed higher rates, sticky inflation and geopolitical shocks with little lasting damage.
The credit data are less comforting. Delinquencies on credit cards and auto loans are rising. Student loan defaults are climbing. Savings have fallen and more households appear to be borrowing simply to maintain spending.
At first glance, the numbers seem contradictory. Household wealth remains high, mortgage delinquencies are low and the banking system appears sound. Yet beneath the aggregate figures lies a growing divide. Auto loan delinquencies have already exceeded their financial-crisis peak. Credit-card delinquencies are moving in the same direction. The stress is concentrated among households with the least financial flexibility. America is not experiencing a balance-sheet recession. It is experiencing a cash-flow squeeze.
Normally, consumer credit weakens after the labour market turns. This time the order may be changing. Defaults are rising while payrolls still look respectable. That is the puzzle investors should care about.
Kevin Warsh’s enthusiasm for artificial intelligence offers one possible explanation. If AI allows companies to raise output while reducing hiring needs, employment data become less reliable. Firms do not need mass redundancies for labour income to weaken. They can hire less, replace fewer departing workers, cut overtime, slow promotions and use software to absorb work once done by junior staff.
None of that produces an immediate jump in unemployment. It does, however, reduce household income growth. Consumers do not experience the economy through GDP releases. They experience it through monthly cash flow. A worker may remain employed while real purchasing power falls. Another may keep a job but lose overtime, promotion prospects or bargaining power. Payroll statistics capture employment. They are less effective at capturing what is happening beneath the surface.
That makes consumer credit a potentially cleaner signal than payrolls. Credit cards and auto loans do not average together asset-rich households and stretched borrowers. They reveal where cash flow is failing. The upper-income consumer is still cushioned by equities, home equity and fixed-rate mortgages. The lower-income consumer is exposed to higher living costs and expensive unsecured debt.
This matters for the Federal Reserve. On the surface, policymakers face a familiar stagflationary problem: inflation remains too high while growth slows. But if Warsh is right, today’s inflation may be masking tomorrow’s disinflation. Energy shocks lift prices temporarily. AI-led productivity gains could lower costs more permanently, particularly if they weaken labour bargaining power and slow wage growth.
The danger is that the Fed spends too long fighting yesterday’s inflation while missing a demand problem already visible in household credit. Positive payrolls would then provide false comfort. Rising delinquencies would be the earlier warning.
This is not another 2008. The mortgage market is healthier, banks are better capitalised and household wealth remains high. But the absence of a banking crisis is not the same as the absence of consumer stress. The optimistic interpretation is that rising delinquencies merely reflect the final effects of an inflation shock. The more troubling possibility is that credit markets are detecting a labour-market transition before official employment statistics can see it.
For most of the past half-century, investors have looked to payrolls for the first sign of economic weakness. If AI is changing the relationship between output and employment, that habit may become expensive. The labour market still looks healthy. Consumer credit is beginning to suggest otherwise. The question for the Fed is which signal arrives first — and which one is telling the truth.
Please continue to check our blog content for the latest advice and planning issues from leading investment management firms.
Alex Kitteringham
29th May 2026