Is there such a thing as too much deposit?

Deposit growth outpaced loan growth, loan-to-deposit ratios fell, and NIM compressed. The loans that did grow came from fintech warehouse lending and riskier unsecured consumer credit. Then the pandemic turned the trend into a flood.

Originally published as a LinkedIn article, February 2020, combined with follow-up posts from 2020–2022.

Net interest margin (NIM) for banks, as represented by the KBW Bank index, has started to decline again. The cause is more nuanced than lower Libor or Prime rates, the indices behind most variable-rate assets on bank balance sheets.

Lower index rates account for most of the compression. But another factor is hurting banks: they can't profitably lend the growing deposits they are gathering.

Deposits are outrunning loans

Over the last twelve months US banks have seen significant deposit growth, while loan growth has lagged. The gap widened in Q3 2019, and given what happened with rates in Q4 I expect it to continue.

The result is a falling loan-to-deposit ratio (LDR). On its own that isn't bad. It improves liquidity and keeps the risk team happy. The second-order effect is the problem: when treasury and market securities yields are at historic lows, unlent deposits drag NIM down. The KBW index's underperformance against the S&P has captured some of this pain. (Small LDR moves matter here: total KBW-bank deposits are about $7.3T as of Q3 2019.)

Where the loan growth is coming from

The loan growth banks have managed has come mostly from two places:

  1. Indirect lending to fintechs, meaning loans to non-depository financial institutions (NDFI). This has been the fastest-growing asset on bank balance sheets for the last 18 months. It says less about a bank's own franchise and brand, and may carry more operational risk. The OCC raised similar concerns in its Spring 2019 Semiannual Risk Perspective.
  2. "Other consumer" loans, the second-fastest-growing category, mostly higher-risk unsecured credit.

No single fix is restoring loan growth, and banks are trying different strategies. Chase is growing its card book, visibly. Citi is expanding NDFI. Bank of America is expanding residential mortgages. Regionals like Citizens, KeyBank and Regions are pushing into other consumer lending.

One thing seems certain: big bonuses for opening checking and money market accounts may be a thing of the past.

Source: Y9C filings.

Follow-ups: the pandemic deposit flood (2020–2022)

Within months the trend became a flood.

  • Deposits up ~23%, loans barely moved. Large banks (over $50B in assets) saw deposits jump right after the Q2 2020 stimulus, while loan balances barely changed.
  • NIM squeezed from both sides. Deposit costs fell quickly toward crisis-era lows, but asset yields fell faster. Combined with uncertain credit on loans in COVID forbearance, bank P&Ls had a tough few quarters.
  • "Banks are just like us." Handed money by a rich uncle (cash-heavy depositors), they stashed it under the mattress: in Treasuries, and in balances at other banks.
  • Second-order effects. My expectation at the time: near-term cost-cutting from low NIM, then a slow expansion of the credit box once the vaccine took hold.
  • Unexpected winners. Branchless online banks, usually single-product lenders with high LDRs and deposit costs before the pandemic, drove deposit costs down sharply. The open question in 2022 was whether those lower costs would hold as rates rose. We now know the answer.