lunes, 23 de junio de 2014

lunes, junio 23, 2014

June 19, 2014 12:39 pm

The perils of returning a central bank balance sheet to ‘normal’

Asset holdings enable policy makers to regulate the economy, writes Benjamin Friedman

(FILES) File photo dated August 09, 2011 shows the US Federal Reserve building in Washington, DC. The Federal Open Market Committee's (FOMC) regular policy-setting meeting starts September 20, 2011 in Washington. The threat of inflation is expected to temper any decisions on new pro-growth measures when Federal Reserve policy-makers meet Tuesday on the US economy's poor health. With growth stagnating and unemployment stubbornly above nine percent, analysts say they expect the Federal Open Market Committee to take some action to stimulate the US economy. AFP PHOTO/KAREN BLEIER (Photo credit should read KAREN BLEIER/AFP/Getty Images)©AFP


With the US Federal Reserve on its way to bringing its bond-buying programme to an end, many are asking how to return the central bank’s balance sheet to “normal” – that is, to its pre-crisis size and composition. The same debate is under way at other central banks. Should they sell their bonds, or hold on to them until they mature? And if they are going to sell, which securities should go first? Yet there is another question that is equally important but seldom asked: is it sensible to return central banks’ balance sheets to “normal”? There are good reasons not to.

At the beginning of 2007, the Federal Reserve System’s assets totalled $880bn. Today, the balance sheet stands at $4.3tn, including $2.4tn of Treasuries and $1.7tn of mortgage-backed securities. The reason for buying these assets was not to reduce the federal funds rate, which had reached zero by late 2008, but to lower the interest rates at which loans are extended to people and businesses, stimulating demand.

The evidence shows that these bond purchases indeed lowered long-term rates relative to short-term rates, and lowered rates on more-risky compared with less-risky obligations. A conservative estimate is that a $600bn bond purchase (the size of the Fed’s second round of bond buying) lowered long-term interest rates by about 25 basis points: not enormous, but a worthwhile contribution to the US economic recovery. And the effect of lower long-term rates was probably reinforced by higher equity prices and a cheaper dollar.

The composition of the assets that the central bank buys matters too. Buying mortgage-backed securities narrowed the difference between the interest rate American homeowners paid on their mortgages and the rate at which the US government could borrow. This helped stop house prices from falling and spurred residential construction. Buying or selling bonds gives the Fed a way of influencing longer-term interest rates in general, and mortgage rates in particular. This lever will remain useful long after short-term rates begin to rise. But it will be out of reach if the central bank returns its balance sheet to its pre-crisis state.

Before the crisis, many people urged the Fed to tighten policy to arrest the developing bubble in the mortgage and housing markets. An increase in short-term interest rates would have helped cool asset markets, but it also risked impeding growth in other sectors. If the Fed’s balance sheet had included mortgage-backed securities, it could have sold them, scooping froth out of those particular markets without depressing the rest of the economy.

What are the potential drawbacks of maintaining a balance sheet significantly larger than the pre-crisisnormal”? First, the central bank may suffer losses if it buys securities that fall in value. So far, this has not happened; central banks’ bond buying programmes have made them – and, therefore, taxpayers record profits. Moreover, while such losses might ultimately impose costs on taxpayers, there is no risk from insolvency of the central bank itself. Unlike at private banks, which must keep cash on hand to meet withdrawals, central bank liabilities are not redeemable for anything else.

Second, because asset purchases have to be paid for, the huge increase in central banks’ holdings has required a corresponding increase in their outstanding liabilities. Those economists who think of prices and wages as determined by the liabilities of the central bank expected hyperinflation to follow

Yet no increase in inflationnot even a few percentage points – has yet appeared in any economy that has pursued this course. That is because the central banks in question have made it advantageous for banks to redeposit the additional reserves instead of lending against them. This has prevented these asset purchases from triggering what might otherwise be an inflationary flood of credit.

For decades, it has been commonly understood that the central bank’s policy interest rate is the only independent instrument of monetary policy. We now see that there are two: the policy interest rate and asset purchases or sales.

But the central bank cannot sell what it does not own. To keep this additional policy tool available, the Fed and other central banks should hold on to an ample supply of assets. They should not shrink their balance sheets to the pre-crisis size.


The writer is a professor of political economy at Harvard University 


Copyright The Financial Times Limited 2014

0 comments:

Publicar un comentario