Same balance sheet, a different asset
The weekly bank balance-sheet release dated 14 August reduces July to a single sentence: total bank credit grew at a 5.9 per cent annual rate while commercial and industrial loans ran at minus 1.1 per cent, securities in bank credit rose 10.6 per cent and government and agency securities 14.6 per cent. In April, commercial and industrial lending was growing at 15.8 per cent. Within four months the line went from expansion to contraction.[1]
Start by dropping the loanable-funds story. A bank does not lend because it has collected deposits; it creates the deposit when it makes a loan or buys an asset. So the question July poses is this: which asset went on the other side of the deposits the banks created? Deposits reached 19,408.5 billion dollars in July, against 18,273.9 billion dollars a year earlier. The creation continued; what the banks wrote on the other side of it changed.[1]
That weakens the timing of what I argued a week ago. On 8 August I wrote that firm credit carried the banking system's expansion in the second quarter, and that the purchasing power created in that quarter originated on corporate balance sheets. July does not refute the mechanism, but it shows the carrier changing the moment the quarter closed.[1], [3]
Whose spending it becomes
A loan drawn by a firm and a government security bought by a bank can create the same quantity of deposits without producing the same spending. The borrowing firm pays wages, inputs and investment, and the deposit enters the production chain. The bank buying government paper credits the deposit to whoever sold it, and that seller usually rearranges a portfolio. The difference shows up in where that money lands.[1]
A third channel is growing. Loans to nondepository financial institutions reached 2,016.1 billion dollars in the week ending 5 August, against 1,674.7 billion dollars a year earlier. Here the bank places another balance sheet between itself and the ultimate borrower: both the credit risk and the knowledge of who the borrower is move off the bank's own book. In calm periods the arrangement looks cheap and efficient; when credit quality turns, it becomes harder to see where the loss comes to rest.[1]
The demand side supports the same reading. Retail sales in July rose less than economists expected, and the S&P 500 fell 13.23 points to 7,785.76, slipping from the record set the day before. A falling loan balance cannot be pinned on unwilling banks alone; in a month when sales slow, a smaller working-capital need produces the same figure. July's data on its own does not separate the two.[1], [2]
What to watch
One month of a break-adjusted series does not declare a turning point on its own. There is still a concrete threshold to watch: if commercial and industrial lending stays negative or near zero through the October data while securities holdings keep growing at a double-digit rate, then the purchasing power the banking system produces has moved out of the production chain and into government paper and non-bank intermediaries. If firm lending reaccelerates, July stays an interruption.[1]
The distinction says something beyond an accounting curiosity: the asset behind a deposit decides whose income that money becomes and which balance sheet absorbs the next shock. July's answer is not yet settled; what needs asking now is the composition itself.[1]