Deposits have continued to flood into banks, but a red-hot inflation report that stoked rate hike expectations has the industry contemplating whether outflows might be on the horizon.

The Federal Reserve announced at the beginning of November that it plans to begin scaling back its quantitative easing operations, under which it buys long-term bonds to keep rates low. And an inflation report a week later showed the biggest year-over-year jump in consumer prices since 1990, adding pressure on policymakers to hike interest rates.

"People went from thinking rising rates were way beyond their planning horizon to being squarely inside their planning horizon in a matter of a couple of weeks," said Pete Gilchrist, head of retail deposits and commercial banking at Curinos, a data company for financial institutions. "Most of the effort right now is around trying to understand what's going to happen both to industry balances and then how that flows to specific institutions."

For now, the Fed is still adding to its bond holdings, however, and deposit inflows have been strong. In the third quarter, deposits across the industry grew $469.76 billion sequentially to $17.633 trillion, a 2.7% increase that was faster than the second quarter increase of 1.3%, according to data from S&P Global Market Intelligence. Funding costs across the industry ticked down 1 basis point sequentially to 16 basis points in the third quarter.

From the end of the third quarter through early November, banks added an additional $163.51 billion of deposits, according to seasonally adjusted, weekly data from the Fed.

But banks and analysts widely expect the torrid pace of deposit growth to cool as the Fed pivots away from stimulating the economy. And some banks anticipate an eventual reversal of deposit flows.

"We do expect that systemwide liquidity is going to remain elevated for some time," said Joseph Pucella, a senior vice president at Moody's. Deposits "may continue to grow but maybe at a slower pace as the Fed adjusts its policy until there is some contraction in the balance sheet, which would reduce liquidity in the system overall, and then, in our opinion, lead to some deposit outflow."

Deposits did not decline the last time the Fed shrank its balance sheet. The central bank launched a program to reduce its bond holdings in October 2017 by opting not to reinvest increasing amounts of principal payments. Its total assets fell by about $700 billion to a low of about $3.8 trillion in August 2019, shortly before it started expanding its balance sheet again after an episode of intense illiquidity in short-term money markets.

As the Fed allowed its balance sheet to shrink from 2017 through 2019, deposits increased by about $938 billion while loans increased about $846 billion, increasing the loan-to-deposit ratio 1.1 percentage points to 76.6%.

Some analysts believe that the Fed is laying the groundwork to again "normalize" its balance sheet, perhaps following the 2017 playbook of reducing its bond holdings after implementing several rate increases. Banks are divided about what that might mean for deposits this time around.

Executives at JPMorgan Chase & Co. said they anticipate a long lag between the end of quantitative easing and the beginning of quantitative tightening, but that they do anticipate systemwide deposits will ultimately contract. "We're still going to increase deposits for a year, and then there'll be a fairly large reduction over a two- or three-year period, which we should be prepared for," said Chairman and CEO Jamie Dimon during the bank's earnings call.

By contrast, Paul Donofrio, a vice chairman at Bank of America Corp., said deposits might never decline outright, citing the multiplier effect from money banks pump into the economy as they increase lending. With the Fed still adding to its balance sheet as it tapers its bond purchases, "deposits are really not likely to decline until many quarters, if you look back at historical data after QE ends, if they ever do," Donofrio said during his last earnings call in the CFO role. "Because as the economy expands, [with] the multiplier effect we could see growth in deposits."

Curinos' Gilchrist said that the extraordinary circumstances of the current environment are difficult to model, but he expects inflation and economic growth, which drive money demand, to overpower runoff of the accumulated "surge" deposits.

"It's an unprecedented time, so it's hard to get numbers that you really feel super confident in through just analytics," he said. "But what we're anticipating for the next several years is that we don't have a net decline in the overall amount of bank deposits. But what we do have is much, much lower growth than we have experienced lately." He said he then expects year-over-year deposit growth to return to a more normal level of 3% to 5% by around 2023 or 2024.

With the market expecting rate increases in the near-term — futures trading on Nov. 16 implied a 54.7% chance for at least three 25-basis-point rate hikes by December 2022 — analysts have also returned their attention to the question of deposit betas, or how much of the rate hike passes through to higher funding costs.

"There's a school of thought out there that deposit betas will be lower, at least initially, even after a few rate hikes, even versus last time just given the sheer amount of money the Fed has pumped into the system," said Michael Rose, an analyst at Raymond James.

But the prospect that inflation might force a rapid series of rate increases introduces a lot of uncertainty, Rose added. "You tell me what inflation does to the pace of rate hikes, and I'll give you a better indication of what the betas will end up being."

Even though he does not anticipate a systemwide decrease in deposits, Gilchrist said he does expect outflows in specific sectors and pockets, which would create difficulties for certain banks.

"One of the biggest challenges in very, very low-rate environments is it can be difficult for banks to tell which funds are going to be around for a long period of time and which aren't," Gilchrist said. "If you take any random consumer and any random business out there, chances are there are two or three banks that think that that business or a person has their primary relationship with them. And one to two of those is wrong."