Federal Reserve Chairman Ben Bernanke and U.S. Treasury Secretary Timothy Geithner testify in Washington after the collapse of AIG, March 2009.
Federal Reserve Chairman Ben Bernanke and U.S. Treasury Secretary Timothy Geithner testify in the House Financial Services Committee Hearing in Washington, March 2009.
Kevin Lamarque / Reuters

Now, almost a decade after the Great Recession hit, the story of its origins and course has become familiar. It began in December 2007, soon after the U.S. housing bubble burst, triggering the widespread collapse of the U.S. financial system. Credit dried up, as banks lost confidence in the value of their assets and stopped lending to one another. Consumer spending plummeted. At first, the U.S. Federal Reserve tried to boost spending through traditional monetary policy, by reducing short-term interest rates. Yet this proved ineffective, even though short-term interest rates fell close to zero. The government then turned to fiscal stimulus, with Congress passing a package of tax cuts and spending increases in 2009, but this, too, proved ineffectual.

After both standard monetary and fiscal policy had failed, the Federal Reserve turned to what it calls “unconventional monetary policy,” aiming to lower long-term interest rates by purchasing long-term financial assets and promising to keep short-term rates close to zero for a long period of time. This policy paid off, and the U.S. economy has at last rebounded. Today, the overall unemployment rate has dropped below five percent, and unemployment among college graduates stands at just 2.5 percent. The European Central Bank (ECB) followed the United States’ lead with massive bond purchases and extremely low interest rates (short-term interest rates are actually negative in many of the eurozone countries). Europe thus also began to dig itself out of recession, but its policy is not proving nearly as effective as similar moves did in the United States.

What is now clear, however, is that unconventional monetary policy and extremely low interest rates have also created major financial risks that could hurt the European and U.S. economies in the years ahead. These policies have driven up the prices of stocks, low-quality bonds, and commercial real estate, potentially setting the stage for another asset price collapse—the very phenomenon that led to the Great Recession in the first place. Unconventional monetary policy does represent a powerful weapon in a central bank’s arsenal. But it is a dangerous strategy, one that policymakers should try to avoid when the next recession hits.


Economic downturns in the United States generally occur when the Federal Reserve raises interest rates to keep inflation in check. When the Fed’s Federal Open Market Committee believes that it has brought inflation under control, it can start to lower rates and reverse the downturn. But the recession that began in the United States at the end of 2007 was unusual: since the Fed did not cause that downturn by tightening monetary policy, it could not reverse it by lowering short-term interest rates. The downturn was the result of the market mispricing assets, including financial assets such as equities and mortgage-backed securities and real assets such as housing.

Between 2000 and 2006, house prices in the United States increased by nearly 60 percent above the long-term trend, driven by very low mortgage interest rates and by lenders who were willing to make loans to “subprime” borrowers. Lenders bundled together many of these subprime mortgage loans and then created investments, or tranches, of varying degrees of risk.

At a tax foreclosures auction of 9,000 properties in Detroit, Michigan, October 2009.
A registered bidder holds her bidders card before the start of the tax foreclosures auction of nearly 9,000 properties in Detroit, Michigan, October 2009.
Rebecca Cook / Reuters

For instance, a bank could take 1,000 subprime mortgage loans (issued to people with low credit scores and a limited capacity to repay) and create separate investment tranches. The riskiest tranche would require the buyer of that investment to bear the losses of the first 100 mortgages to default. If a borrower defaulted, the creditor would seize the house and sell it for a price that was generally much less than the amount owed on the mortgage. That tranche was risky enough that investors would regard it as a junk bond, bearing a high yield.

The second-riskiest investment would require the buyer of that tranche to bear losses only after 100 mortgages had defaulted, up to a total of 200 defaults. Since the probability that more than ten percent of the original 1,000 mortgages would default seemed low, investors considered this second tranche to be safer and gave it a higher rating; as a result, the second tranche had a lower expected rate of return. Banks created additional tranches, each one apparently safer than the last. Investors in the fifth tranche, for example, would bear a loss only if more than 400 of the original 1,000 mortgages defaulted. The probability that more than 40 percent of the mortgages would default seemed so low that rating agencies could assign this tranche a triple-A rating.

When the housing bubble burst, of course, these ratings proved far too optimistic. Widespread defaults on subprime mortgages caused the prices of the seemingly safe mortgage tranches to fall sharply. That drop sent a signal to investors in other, very different securities that investors had underestimated risk, and the prices of many types of assets fell rapidly as a result. The S&P 500 dropped by nearly 20 percent in the 12 months after July 2007 and had lost half its value by March 2009.

Banks and other financial institutions found that it was often impossible to obtain market prices for mortgage-backed securities and other risky assets. Not knowing the value of their own portfolios, they could not judge the solvency and liquidity of other financial institutions either. As a result, they were unwilling to lend to them, and the financial system ground to a halt.


As the financial collapse reverberated throughout the economy, the Fed responded in the traditional way. It gradually lowered the short-term federal funds interest rate, the rate at which banks lend money to one another, from 5.3 percent in September 2007 to 4.3 percent in December 2007 (when the downturn officially began). By October 2008, the Fed had pushed the rate below one percent.

In most economic downturns, traditional monetary policy of this kind is enough to bring about a recovery. Those downturns tend to be relatively short—the average time from peak to trough in U.S. business cycles is just ten months—and relatively shallow. To bring about a recovery from these dips, the Fed employs conventional monetary policy, or “open-market operations”—essentially, buying short-term government notes. This policy works by lowering the short-term interest rate and temporarily lowering some longer-term rates as well. Spending that is sensitive to interest rates, such as housing construction, usually picks up. So there is no need to drive down long-term rates by buying bonds and committing to a long period of low rates.

Nor, in a typical downturn, is there any need for fiscal policy, which is both unnecessary and likely to destabilize the economy. Because it can take months for Congress to pass legislation, there are long lags between the start of a downturn and the potential launch of an effective fiscal stimulus. This timing problem means that a fiscal package aimed at ending a short downturn may hit too late, once the economy is already expanding, leading to economic overheating and rising inflation.

But not all downturns are created equal. In the dysfunctional financial environment that prevailed from the end of 2007 onward, traditional open-market operations to reduce short-term rates proved ineffective. Working with the U.S. Treasury, the Fed took a series of additional measures, such as guaranteeing money market funds and buying assets from investors, to prevent widespread financial failures and protect mutual funds. It also bought a large amount of mortgage-backed securities before the end of 2008 to lower mortgage rates and stimulate home buying. But these actions were not sufficient to revive the economy: in 2008, there was still a $700 billion annual shortfall in demand.

Because it soon became clear that this recession was going to be deeper and longer than usual, there was less risk that the government would mistime fiscal policy. So an expansionary fiscal policy was appropriate, and in 2009, U.S. President Barack Obama asked Congress for about $300 billion a year in spending increases and tax cuts for three years. But the collapse of consumer spending and home building had created a hole in demand that was much larger than the $300 billion that Congress provided. The administration also designed the stimulus poorly, spending most of it on temporary tax cuts and transfers to states rather than on real government spending that would add to demand, such as infrastructure projects and the repair and replacement of government military and civilian equipment. A better-designed fiscal stimulus might have been able to raise aggregate demand and kick-start a true recovery, but the 2009 Obama plan probably added more to the national debt than it did to economic activity.

After declining for 18 months, the U.S. economy finally began to grow in June 2009, but only very slowly. And so the Fed concluded that neither traditional monetary policy nor the fiscal policy of the Obama administration was going to achieve an adequate recovery. It responded with the unconventional policy of quantitative easing: buying billions of dollars’ worth of long-term bonds and promising to keep short-term rates close to zero for a long period of time.

As Ben Bernanke, then chair of the Federal Reserve, explained, the purpose of this new policy was to lower long-term interest rates so that investors would buy more equities and other riskier assets. The Fed also hoped that the unprecedentedly low interest rate on mortgages would boost house prices. The resulting uptick in household net worth, it expected, would increase consumer spending, thus speeding the recovery. By targeting net worth, quantitative easing would be very different from the traditional monetary policy the Fed had already employed, which aimed to encourage investment and other spending that is sensitive to interest rates.

Unconventional monetary policy and extremely low interest rates have created major financial risks.

The strategy worked well. House prices rose by 13 percent from December 2012 to December 2013. The Standard & Poor’s measure of equity prices rose by 30 percent during those same 12 months, and the net worth of households grew by $10 trillion. As a result, consumers spent more, which raised the profits and wages of the firms that sold the goods and services they bought. Those higher incomes then led to further increases in consumer spending, lifting overall GDP. In 2013, GDP rose by 2.5 percent, and the unemployment rate fell from 8.0 percent to 6.7 percent.

The impact of quantitative easing on asset prices also proved surprisingly large, because investors acted as if the decline in interest rates was not just temporary but would last more or less indefinitely. This suspension of disbelief was necessary to drive down the interest rate on very long-term bonds and increase the prices of equities. If investors had thought that interest rates would return to normal only a few years later, long-term rates would not have fallen so low, nor would equity prices have risen so much. But as it was, unprecedentedly low interest rates and the rise in equity prices helped end the Great Recession in the United States.


Across the Atlantic, the same strategy was not working as well. Like the Fed, the ECB was following a strategy of large-scale asset purchases and extremely low (even negative) short-term interest rates. But the purpose of the ECB’s quantitative easing was very different from what the Fed was trying to do. Since the eurozone lacks the widespread stock ownership that exists in the United States, quantitative easing could not stimulate consumer spending by raising household wealth.

Instead, a major unspoken reason for the ECB to lower interest rates was to depress the value of the euro and thus stimulate net exports. Financial investors in Europe sold euros and bought dollars to take advantage of the higher yield on securities denominated in dollars. The euro’s value fell from almost $1.40 in the summer of 2014 to $1.06 by the fall of 2015, before rising slightly to $1.12 in early 2016.

But even though a weaker euro has stimulated exports and discouraged imports, it has done little to raise the eurozone’s total exports and combined GDP. Most eurozone exports go to fellow members of the monetary union, which use the same currency. Nor have exports to the United States increased much, because European exporters generally invoice their exports in dollars and adjust their dollar prices very slowly. The overall effect has been disappointing: between September 2014 and September 2015, total net exports from the eurozone budged upward only slightly, rising from 17.4 billion euros to 20.2 billion euros in an economy with a GDP of more than 13 trillion euros.

Both the United States and Europe may well eventually pay for their use of unconventional monetary policy.

Nor has the ECB’s strategy of purchasing bonds to increase the amount of cash banks have to lend seen much success. In contrast to banks in the United States, which are paid a positive interest rate on reserves by the Fed, commercial banks in the eurozone actually lose money when they deposit funds at the ECB. Not surprisingly, then, they do not deposit the funds they acquire from the sale of bonds to the ECB at the central bank. The ECB hopes that the banks will use their cash to increase lending to businesses and consumers. So far, however, bank lending has barely increased, perhaps because of a lack of demand from borrowers and the weakness of the banks’ own capital positions.

The eurozone is also having trouble when it comes to inflation, with the rate now sitting at just 0.4 percent. The ECB, fearing a deflationary spiral of declining prices and wages, is eager to raise the eurozone inflation rate to its target of just under two percent. The Federal Reserve also wants to increase the rate of inflation (in addition to achieving maximum employment). The Fed expects its quantitative easing will lift U.S. inflation back to two percent: by increasing real demand and reducing the unemployment rate until employers have to start raising wages to hire new workers, the policy will accelerate inflation. But the ECB, by contrast, will likely struggle to raise the overall inflation rate in the eurozone so long as unemployment remains at more than ten percent, about three percentage points above where it stood before the recession. The ECB’s quantitative easing can probably achieve higher inflation only through the decline in the value of the euro and the resulting increase in import prices—a limited process that still leaves core inflation below one percent. The bottom line is that quantitative easing in the eurozone is likely to raise employment and inflation far less than it did in the United States.


Both the United States and Europe, however, may well eventually pay for their use of unconventional monetary policy. Although the Fed succeeded in boosting wealth and stimulating economic activity, it also increased several risks in financial markets—risks that may create instability when interest rates return to normal.

In the United States, very low interest rates and the Fed’s asset purchases drove down the yield on government bonds. In response, investors sought higher returns by buying other, riskier assets, and lenders did the same by issuing more speculative loans. The Fed wanted to promote such risk taking. But as investors have bought up these riskier assets, their prices have risen and their yields fallen—even though the risk of default remains the same. In other words, investors may have overpaid for these assets. And once interest rates return to normal, that mispricing could cause problems such as a stock market decline, as investors sell these riskier stocks and their prices fall.

One sign that investors have mispriced assets can be found by looking at the price-earnings ratios of companies in the S&P 500. The ratios have grown higher than they were before the downturn, and about 30 percent higher than their historic averages, suggesting that considering companies’ incomes, their stocks are overvalued. These very high share prices might make sense if interest rates stayed at today’s low level forever. But when interest rates rise, as they eventually must, the high price-earnings ratios will no longer be justified, and the stock market will have to correct.

Investors may be mispricing not only stocks but also commercial real estate. Commercial real estate prices are very high relative to the yields provided by the rents on those properties. When interest rates rise, those yields will be less competitive, causing the value of those properties to fall. And because those investments are financed with borrowed funds, the investors could lose a very large share of their net equity.

Banks and other lenders are also seeking higher returns by lending to riskier borrowers. U.S. Treasury bonds are much safer than low-quality corporate debt, so investors should demand much higher yields on low-quality debt. The large demand for investment in junk debt has for the most part driven its price up and its yield down, so that the spread between the yields on essentially risk-free U.S. Treasury bonds and the yields on low-quality corporate debt has been quite narrow. Banks are also making more loans to less reliable borrowers and more loans that entail few restrictions for the borrowers. If there is an economic downturn, many of these high-risk loans will fail.

Low interest rates have also led to increased lending from the United States to corporate borrowers in emerging markets. Companies in those countries have been tempted by the opportunity to borrow in dollars at historically low rates. Lenders and investors in the United States, meanwhile, like the higher yields available on loans to emerging-market borrowers.

These loans are risky for two reasons. First, rising interest rates will hurt the profits and even the viability of the borrowers. The prospect of rising interest rates in the United States has already caused U.S. creditors to raise interest rates on loans to borrowers in emerging markets. And the prospect of rising U.S. rates is strengthening the dollar relative to the currencies of those other countries, making it harder for companies whose earnings are denominated in their local currency to repay their loans.

In the eurozone, too, risk has increased. Italy’s ten-year sovereign debt yields investors just 1.4 percent, even though the country’s debt equals more than 100 percent of its GDP. When interest rates return to normal, these investors will lose money. In the eurozone’s private debt market, loans with few requirements for borrowers now constitute nearly half of all institutional loans.

Investors and lenders have clearly taken substantial risks. But whether this puts the entire financial system in jeopardy remains unclear. Nor is it clear whether the normalizing of interest rates by the Fed will trigger a systemic downturn. Still, there is no doubt that unconventional monetary policy has increased the risk of such systemic instability.

Despite the close link between quantitative easing and financial instability, the Fed continues to downplay the connection.

Unfortunately, the Fed seems oblivious to this possibility. In December 2015, it finally began the process of raising the overnight federal funds interest rate. The minutes of the Federal Open Market Committee meeting showed that members had little interest in taking financial stability risks into account when setting interest rates in the future. The committee said nothing suggesting that it would raise rates more rapidly to discourage the mispricing of assets. Instead, it stressed that the federal funds rate will rise only very slowly to increase aggregate demand and will remain low even after the economy has achieved full employment and the target rate of inflation.


Despite the close link between quantitative easing and financial instability, the Fed continues to downplay the connection. In a 2014 speech, Janet Yellen, Bernanke’s successor as chair of the Federal Reserve, explicitly limited the goals of monetary policy to the two congressional mandates of maximum employment and price stability, saying that financial stability should be left to “macroprudential policies,” or measures the government can take that limit the risks of instability across the whole financial system. But it is not clear what those macro­prudential policies should be, in the United States or the eurozone.

The one clear example of a potentially useful macroprudential policy in the United States is the increased capital requirements that the Fed has imposed on commercial banks. Because banks now hold more capital, they are capable of absorbing greater losses. The dysfunction that afflicted the banking system in 2007 and 2008 is less likely to recur. But since the high capital requirements force banks to have more capital relative to their total portfolios, they are less likely to hold bonds, which will make interest rates more volatile when investors want to sell their bonds.

Janet Yellen discusses the Fed's recent decision to raise interest rates in Washington, D.C., February 2016.
Not on my watch: Janet Yellen in Washington, D.C., February 2016.
Olivier Douliery / Abaca (SIPA via AP Images)

Although the Fed has subjected banks’ portfolios to stress tests to measure the effect of rising interest rates and investment losses, the results might not have shown the whole picture: U.S. banks, particularly small and medium-sized ones, may not have enough capital to maintain solvency and liquidity should their risky loans and investments fail in large numbers. The same is true of the stress tests that the ECB has performed in Europe, especially because it is hard to be confident of the value of the sovereign bonds of some of the peripheral countries, such as Portugal and Spain. In both Europe and the United States, therefore, macroprudential policy has done little to mitigate the growing risk of financial instability that unconventional monetary policy has caused.


In the future, central banks should respond to normal downturns with the best tool available: traditional monetary policy alone. But in the face of a more severe downturn, a combination of fiscal policy and traditional monetary policy will likely prove better than the more extreme option of relying on unconventional monetary policy, given the accompanying risks of financial instability.

Ideally, if countercyclical fiscal policy is needed to deal with a future downturn, it should combine a short-term fiscal stimulus with changes in entitlement programs to stabilize the long-term level of national debt. Otherwise, financial markets may fear sustained future deficits and rising government debt, a worry that could raise long-term rates to the point where they offset some or all of the favorable effect of a fiscal stimulus. Unfortunately, it may not be politically feasible to touch entitlement programs. But it is possible to design a fiscal policy that provides a net stimulus to the economy without increasing the size of the national debt—something that is particularly important now in the eurozone, where debt levels are already very high.

The key to such a revenue-neutral stimulus strategy is to recognize that there are two possible types of fiscal stimulus policies. The first entails cutting taxes and increasing government spending, which raise fiscal deficits and boost demand through the traditional Keynesian channels. The second includes specific investment incentives, such as the investment tax credit, that make new investments more profitable and thereby encourage firms to invest.

This second type of fiscal stimulus can raise aggregate demand without expanding the national debt, if the government also enacts a temporary increase in the corporate tax rate. An investment tax credit would make new investments more profitable, and a higher corporate tax rate would apply only to existing capital. Businesses would have more incentive to invest during the years when the investment tax credit was available.

A similar revenue-neutral strategy could stimulate consumer spending in countries that have value-added taxes, as is common in the eurozone. The government in a given country could commit to raising the value-added tax rate each year for the next several years, balancing the higher tax burden with reductions in the personal income tax. Individuals would have an incentive to spend sooner to avoid paying more in the future. In the eurozone, such revenue-neutral fiscal policies have the additional advantage that, unlike monetary policy there, they can be tailored to suit individual countries.

These revenue-neutral fiscal policies might not be enough to deal with a crash as large as the one that hit the United States and Europe in 2007. But they would serve as a useful supplement to conventional monetary policy, limiting the need for unconventional monetary policy, which causes investors and lenders to misprice assets and increases the risk of financial instability.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now
  • MARTIN FELDSTEIN is George F. Baker Professor of Economics at Harvard University and President Emeritus of the National Bureau of Economic Research. He was Chair of the Council of Economic Advisers from 1982 to 1984.
  • More By Martin Feldstein