MNI: Fed QT May Have Deeper Tightening Effect - Ex-Officials

The Federal Reserve's unprecedented effort to whittle down a record USD9 trillion balance sheet could have a much larger than expected impact on financial conditions if markets become disorderly, ex-Fed officials told MNI, casting doubt on the chances that the asset reduction plans will run quietly in the background as current policymakers hope.

The Fed's second go at shrinking its balance sheet, which it calculates will be equivalent in impact to just one rate hike a year, is expected by officials to proceed more smoothly than its first attempt in 2018 when key lessons were learned about banking system reserve requirements. But pandemic-era QT is twice as aggressive and comes at a time of rapidly rising yields. If markets turn bumpy, ex-officials fear the tightening effect from the Fed’s bond runoffs could be multiplied several times over.

“There certainly is potential for financial market tumult,” former Richmond Fed President Jeffrey Lacker said in an interview. “This is another looming possibility for the Fed to have to face, which is a scenario in which the inflation fight has not yet been won and yet financial markets exhibit some volatility, some asset price declines, some credit spreads widening, which is what you’d expect if growth slows down."

If forced to decide between halting runoffs and loosening policy in response to market downturns and widening credit spreads, and keeping policy tight to tame inflation, "it’s going to be a tough choice, and speculation about that choice is going to likely roil financial markets by itself,” Lacker said.

REPO VOLATILITY

When it reduces its balance sheet, the Fed in essence asks the private sector to hold more Treasuries and MBS, putting upward pressure on those rates and potentially also leading to more activity in repo markets as levered investors seek to fund their positions. That raises the possibility that repo rates could one day rise in a disorderly way, resulting in stresses that could exacerbate existing structural weaknesses within the Treasury market, ex-officials said.

In 2018 and 2019, QT contributed to spikes in repo rates, forcing the Fed to reverse course and add liquidity through market operations.

This time, to ensure a smooth transition, the FOMC will first slow then stop runoffs when reserve balances are "somewhat above the level it judges to be consistent with ample reserves," it said when announcing QT this month.

"No one at the Fed has a clear understanding of what the rolloff will do to financial markets and the economy," said Rick Roberts, a former New York Fed staffer and ex-adviser to the Kansas City Fed. Hopefully QT will run in the background without a shock to markets or the economy, but the process will likely slow or stop "if the Fed identifies a meaningful disruption that can be tied to the rolloff," he added.

REPO FACILITIES

Current and former officials have pointed to massive inflows into the Fed's reverse repo facility as a sign the banking system has an abundance of cash this time.

Money market funds have deposited as much as USD2 trillion in the facility, "suggesting the Fed can reduce the balance sheet by that much and nothing would happen," former New York Fed research director Stephen Cecchetti told MNI. The FOMC has signaled it intends to allow its portfolio to decline by roughly USD3 trillion over three years to about 20% of GDP from 36% now.

"It’s hard to find evidence large scale asset purchases do a lot when financial institutions are operating normally," he said. "The first 2 trillion are really easy, they got that back already in the overnight reverse repo facility. The next trillion maybe might do something, and if they sell MBS, which is a thinner market than Treasuries, then maybe something will happen."

The new standing repo facility could offer an early warning of strains in money markets, current and former officials said. Usage of that borrowing channel could be a more valuable signal that the Fed is approaching the end of QT than any survey or model-based estimate.