The Fed’s quantitative easing programme has cost too much
AMERICA’S experiment with quantitative easing – and the ensuing quantitative tightening – is almost over.
This week, the US Federal Reserve will likely announce plans to slow the shrinkage of its balance sheet. This foreshadows the end of a long period in which it sought to stimulate the economy by holding large quantities of Treasury and mortgage securities, and later had to shrink its balance sheet.
So did it work? Yes, but at excessive cost.
Without doubt, quantitative easing had benefits. When the onset of the pandemic convulsed financial markets in March 2020, the Fed’s asset purchases kept funds flowing and stabilised prices.
Over the next two years, these purchases pushed down longer-term interest rates, providing economic stimulus at a time when the central bank determined it couldn’t lower short-term rates any further.
Yet, one must consider the costs of the Fed’s asset-purchase programme, too.
First, by lowering rates on mortgage loans, it overstimulated the housing market. Demand soared, driving up home prices and overall inflation.
Second, it contributed to last year’s regional banking turmoil. The cash that the Fed paid for securities turned up as deposits, a flood of money that banks had to manage.
The institutions – such as Silicon Valley Bank – that foolishly invested such runnable deposits in long-term, fixed-rate securities have themselves to blame. Still, the Fed bears responsibility for creating the flood in the first place – something it failed to acknowledge in its otherwise candid post mortem last May.
Third, the programme has been expensive in dollar terms. Since the Fed raised short-term rates to more than 5 per cent, its interest expense on liabilities (including bank reserves) has far exceeded its income from securities holdings. The Fed lost more than US$100 billion last year and cumulative losses could reach US$250 billion.
The true cost is considerably higher. In the absence of quantitative easing, the Fed would have turned a sizeable profit, given that its liabilities would have been mostly interest-free currency. In the final tally, the difference between what it actually earned and what it could have earned could exceed US$500 billion.
In hindsight, I can identify three mistakes that drove up the costs.
First, the Fed kept buying assets far longer than needed. After effective coronavirus vaccines were introduced in late 2020, officials should have recognised that growth would bounce back as the economy reopened, and that the federal government’s vast pandemic-related fiscal transfers – more than US$5 trillion – would render further extraordinary monetary stimulus unnecessary.
Second, the Fed decided to keep buying assets until substantial progress had been made towards its stated employment and inflation goals, and to phase out purchases slowly rather than abruptly. Thus, it was still buying in the first quarter of 2022, even though growth was strong and inflation was high.
Third, the Fed committed not to raise short-term rates until the economy was at full employment and inflation both reached its 2 per cent target and was expected to stay there long enough to offset previous downside misses.
This commitment ensured that it would respond too late to an overheated economy and higher inflation, necessitating a more radical reaction. If the central bank had begun to tighten monetary policy sooner, the peak in short-term rates would likely have been considerably lower, resulting in fewer losses.
The Fed’s excessive balance sheet expansion directly pushed up the costs of quantitative easing. The last trillion dollars of asset purchases, completed between June 2021 and March 2022, could end up costing the central bank – and the US taxpayer – US$100 billion, for very little benefit.
Fed officials should evaluate the quantitative easing programme as part of next year’s monetary policy framework review. The aim should be to learn from past mistakes and develop a guide for the future – focusing not only on the benefits, but also on the costs.