On October 29th the Federal Reserve said it would end “QE3”：the programme of asset purchases it first announced in September 2012 and began shrinking last December. Quantitative easing, or the buying of assets with newly created money, has been the workhorse of monetary policy since rich-world interest rates fell almost to zero in 2008-09. Despite its expansive use since then, many still see it as an exotic and possibly dangerous monetary tool. They raise three pressing questions：did it work, did it have unacceptable side effects, and was the Fed right to stop? Though QE is often described as an “unconventional”form of monetary policy (as opposed to mundane adjustments to interest rates), it has actually been in use for some time. The Federal Reserve tried it from 1932 to 1936, and the Bank of Japan in the early 2000s. Both have used it again since the financial crisis, along with the Bank of England. QE is thought to work in a few ways. Early in the recession asset purchases were primary intended to ease credit by directing a firehose of liquidity into fearful markets. Other effects became more important over time, such as portfolio rebalancing. A bank that has sold bonds to the Fed will typically wish to spend the proceeds to buy some other asset. When banks buy new securities, the newly created money flows through the system, raising asset prices and reducing interest rates. That, in turn, should boost demand by making investment attractive. If banks buy foreign assets, that may weaken the exchange rate, giving a boost to exporters. Finally, QE is also thought to work as a signal：research suggests that it reinforces central banks' policy guidance, making promises to keep rates low more credible, for instance. QE's detractors point out that central banks around the rich world have expanded their balance-sheets by trillions of dollars in recent years, yet are still nurturing lackluster recoveries. Nonetheless, there is a consensus among researchers that QE has indeed lowered borrowing costs, and thus increased both economic output and inflation, as its advocates intended. In both America and Britain, for instance, several studies have concluded that it helped to lower interest rates. Those declines are generally held to have boosted economic growth. Recent work by Martin Weale and Tomasz Wieladek of the Bank of England found that for every 1% of GDP the Fed spent on bonds, both real output and inflation rose by about a third of a percentage point. Given how close so many economies now are to outright deflation, even a small boost to inflation is not to be sneered at. Why haven't central banks done more in that case? Excessive optimism is one reason；the Fed, for example, has repeatedly overestimated how quickly the American economy was likely to grow. The fear of possible side effects has also stayed central bankers' hands. In some cases, those are subsiding：claims that the new trillions coursing through the economy would lead to hyper-inflation no longer seem credible, given how static prices remain. Central bankers are also increasingly confident that their pneumatic balance-sheets will not interfere with conventional monetary policy. The Fed, for instance, has begun paying interest on the excess reserves banks keep with it. By raising that rate it can attract extra reserves and thus curb bank lending. The worry that QE is generating dangerous financial instability has been more persistent. Jeremy Stein, who served on the Fed's Board of Governors from 2012 to 2014, thinks that persistently low interest rates have caused investors to embrace frighteningly risky assets in a search of yield. If QE spurs the construction of a financial house of cards, the argument runs, its eventual collapse may hurt more than QE ever helped. These fears will be difficult to judge until more time has passed, yet there are good reasons to believe they are overstated. Most central bankers insist they have the regulatory tools to keep excessive risk-taking in check (although their recent record in that respect is inglorious). Recent research by Gabriel Chodorow-Reich of Harvard University finds that the “reach for yield”QE has prompted has been modest relative to its benefits. Asset purchases helped recapitalise ailing banks, he reckons, and higher asset values have dampened the incentive to make risky bets. Anyway, it is important to weigh QE's risks against those of other policies. Were the European Central Bank to resort to QE, it might lead to undue risk-taking, but if the policy also reduced the odds of catastrophic break-up of the euro it would be worthwhile. Critics' third complaint is that QE raises inequality. Asset purchases, naturally, raise asset prices. Since the rich own most financial assets, QE tends to increase inequality through the “portfolio” channel. Yet to the extent that QE boosts growth, it reduces inequality, for jobless is regressive：poorer workers suffer higher unemployment rates and more significant loss of income in weak economies. Higher inflation also makes debts easier to bear, because they can be repaid with money that is worth less. Since borrowers tend to be poorer than creditors, QE diminishes inequality in that way too. Recent research suggests that tight monetary policy tends to raise inequality while expansionary policy reduces it. If QE works and its potential costs are overstated, should the Fed not keep at it? Those cheering its demise note that growth has been strong and steady for most of 2014, and that unemployment, at 5.9%, is near its long-term average. Yet history suggests that leaving QE behind is never simple. As long as interest rates remain near zero any nasty new development will force the Fed to resume it or stand by while the economy deteriorates. And low and falling expectations of inflation suggest that markets doubt the Fed's ability to leave Japan-style stagnation behind. A premature exit from QE3 is a good way to make sure there is a QE4.