Watching deposits leave in real time

March payrolls rose by a healthy 236,000, but the Federal Reserve's weekly H.8 release showed deposits still pouring out of small and mid-sized banks. In normal times the H.8, which tracks commercial bank assets and liabilities, is read by a handful of bank analysts and money market economists. Since the March banking collapse it has become the most closely watched release on Wall Street, used by fund managers as a live gauge of the credit squeeze. The deposit drain it records offers an early look at the downturn to come.

The figures show money moving from small banks to big ones. In the three weeks after SVB failed, small domestic banks lost hundreds of billions in deposits. The twenty-five largest banks took in modest inflows, and money market funds absorbed the rest.

A labour market still running hot

The jobs report on its own would have pleased the hawks. The unemployment rate edged down to 3.5 per cent, close to its lowest in more than half a century, and average hourly earnings were up 4.2 per cent on a year earlier. Hiring has slowed from the pace of last year but remains well above what the economy needs to absorb new entrants to the workforce. On payrolls alone, the Fed would have every reason to keep raising rates at its meeting in early May.

That is the problem with relying on the monthly figures. Employment responds to credit conditions with a delay, as firms finish projects already financed and hold on to staff they struggled to hire. A labour market that looks strong in March tells investors little about what happens once the loans that fund next year's expansion are refused.

How to read the H.8

The H.8 is a blunt instrument, and reading it well takes care. The Fed builds the weekly figures from a sample of banks and benchmarks them to the fuller quarterly reports that every bank files, so early numbers are revised. It splits domestic lenders into the twenty-five largest and everyone else, which is why analysts can see money shifting from small banks to large ones. Bank failures complicate the series, since the balances of a failed lender pass to a bridge bank run by the FDIC and must be classified consistently.

Seasonal patterns are the other trap. Deposits normally fall in April as households and firms pay their taxes, which moves cash from bank accounts into the Treasury's account at the Fed. Analysts watching for signs of stress will have to separate that routine outflow from anything more alarming. Four-week changes, compared with the same period in earlier years, give a steadier signal than any single week.

Small banks keep bleeding

Small banks are losing more than large uninsured institutional deposits. Ordinary business accounts are leaking too, as owners spread their cash across several banks to stay below the FDIC insurance limit.

To cover the outflows, regional lenders have turned to costly Federal Home Loan Bank (FHLB) advances and the Fed's emergency facilities, paying 5 per cent for funding that used to cost 50 basis points. The hit to margins is forcing them to cut new lending, and fast.

The FHLB system, set up in 1932 to support mortgage lending, has become the first stop for banks in need of cash, because its advances carry none of the stigma of the Fed's discount window. It raises money by selling bonds, and its loans are secured on mortgages and securities. That makes it a reliable source of liquidity but not a cheap one, and advances are short-term, so a bank that leans on them must keep refinancing at whatever rate prevails.

Loans turn down

Weekly loan data already show commercial and industrial lending turning down, while new commercial property loans have stalled. Bank lending usually leads business hiring and investment by two to three quarters, which makes the weekly deposit figures one of the better guides to where the economy is heading. Hiring plans are usually the last thing to change, after investment budgets have already been cut.

The chain of causation is straightforward. A bank that loses deposits either pays more to replace them or shrinks its assets, and new loans are the easiest thing to stop making. Borrowers turned away by a small bank may find credit elsewhere, but often at a higher rate and on stricter terms, and some will simply postpone investment or hiring. Small firms, property developers and farmers depend most on regional lenders, and they have the fewest alternatives in the bond market.

The case for calm

The optimistic reading deserves a hearing. The largest banks and money market funds have kept the money inside the financial system, and some of it will be lent again, just by different institutions. The deposit guarantees for SVB and Signature and the Fed's new lending facility may slow the outflow now that the initial panic has passed. If the drain from small banks stops within weeks, the damage to lending could prove modest.

The trouble with that view is the cost of funding. Even if deposits stop leaving, the banks that kept them have had to pay more to do so, and a lender with a narrower margin will lend less or charge more. The redistribution toward big banks also matters, since large lenders are less active in small-business and commercial property lending than the regional banks they are displacing.

What to watch next

Several releases will test the thesis. First-quarter bank results in mid-April will show how much deposits fell and how far funding costs rose. The Fed's Senior Loan Officer Opinion Survey, due in early May, will reveal whether banks tightened standards after the failures. The H.8 will keep arriving every Friday, and a sustained return of deposits to small banks would weaken the case for a credit-driven slowdown.

Solid payroll numbers describe the past. The deposit data describe the credit that will be available over the next two to three quarters, and on that evidence private borrowing and hiring are likely to weaken by the end of the year.