Empirical work done at the Fed and elsewhere suggests that QE worked in the sense that it reduced interest rates other than the federal funds rate, and particularly seems to have succeeded in driving down longer-term rates, which are the rates most relevant to spending decisions.
Critics have argued that QE has gradually become less effective over the years, and should no longer be used. It is extremely difficult to appraise the effectiveness of a program all of whose parameters have been announced at the beginning of the program. But I regard it as significant with respect to the effectiveness of QE that the taper tantrum in 2013, apparently caused by a belief that the Fed was going to wind down its purchases sooner than expected, had a major effect on interest rates.
More recently, critics have argued that QE, together with negative interest rates, is no longer effective in either Japan or in the euro zone.That case has not yet been empirically established, and I believe that central banks still have the capacity through QE and other measures to run expansionary monetary policies, even at the zero lower bound.Empirical work done at the Fed and elsewhere suggests that QE worked in the sense that it reduced interest rates other than the federal funds rate, and particularly seems to have succeeded in driving down longer-term rates, which are the rates most relevant to spending decisions.
Critics have argued that QE has gradually become less effective over the years, and should no longer be used. It is extremely difficult to appraise the effectiveness of a program all of whose parameters have been announced at the beginning of the program. But I regard it as significant with respect to the effectiveness of QE that the taper tantrum in 2013, apparently caused by a belief that the Fed was going to wind down its purchases sooner than expected, had a major effect on interest rates.
More recently, critics have argued that QE, together with negative interest rates, is no longer effective in either Japan or in the euro zone.That case has not yet been empirically established, and I believe that central banks still have the capacity through QE and other measures to run expansionary monetary policies, even at the zero lower bound.

Kashkari on quantitative easing before joining the FOMC:

• On CNBC in 2012: At the end of the day, this is not going to lead to real economic growth. Unfortunately, it likely leads to an inflationary outcome.
  
  
• On Twitter in 2013: Printing money is morphine. Makes u feel better but doesn’t cure.

Well I think its a bit much that I said anything like that. I said you might consider pausing in our taper program. Our taper program was supposed to be a state contingent program that would be adjusted as the data came in and as I was speaking we certainly had markets thinking that global recession was on the horizon and if there was ever a time to be state contingent maybe that was it. Now as it turns out that all kind of evaporated and the committee did a very sensible thing at the meeting last week; we cited lower market measures of inflation expectations which showed concerns for that issue, but we also cited survey based measures which are more stable about inflation expectations. And so I think that it shows we are keeping a watchful eye on this issue. We went ahead and closed the QE program; I thought that was the right judgment for that meeting.

Especially at the beginning of QE3, when the nonsense term QE infinity was being broadcast by otherwise responsible analysts, journalists and pundits, rates dipped to lows that helped creditworthy businesses bolster their balance sheets and strengthen their wherewithal for future expansion. Again, even if you were living under a rock, you know that this gift of near-cost-free debt as measured in inflation- and tax-adjusted terms has thus far been used primarily to finance stock buybacks, increase dividends and fatten cash reserves, and recently, finance mergers by the most creditworthy companies. For those with access to capital, it was a gift of free money to speculate with. (One wagI believe it was mequipped that there was, indeed, a positive wealth effect the wealthy were affected most positively.)

John Williams, president of the San Francisco Fed, said in an interview with Reuters that the first line of defense at the central bank, if needed, would be to telegraph that U.S. rates would stay near zero for longer than mid-2015, when he currently expects them to rise. If the outlook changes "significantly," with inflation showing little sign of returning to the central bank's 2-percent target, he said he would even be open to another round of asset purchases

"If we really get a sustained, disinflationary forecast ... then I think moving back to additional asset purchases in a situation like that should be something we should seriously consider," Williams said.
...
In the interview, Williams repeated he is comfortable with his call for a rate hike about nine months from now. But "if inflation isn't moving above 1.5 (percent) and we get stuck into that gear, that would argue for a later liftoff," he said. "If we don't see any improvement in wages, that would be a sign that we still have a lot of slack in the economy and we are not getting any inflationary pressure to move inflation back to 2 percent."

The Fed next meets on Oct. 28-29, when some policymakers are pushing to ditch a promise to wait a "considerable time" before raising rates, given a sharp drop in the U.S. unemployment rate to 5.9 percent. Williams wants to leave that phrase intact for now, but said the central bank could adjust it as the outlook evolves.

The problem with QE is it works in practice, but it doesnt work in theory.

The preponderance of evidence suggests that the Fed's asset purchases raised the prices of the assets purchased and close substitutes as well as those of riskier assets.

Importantly, evidence--including the evidence of our eyes--shows that foreign asset markets have been significantly affected by the Fed's purchase programs.

Although much of the recent commentary on spillovers has focused on the United States, it bears mentioning that other countries' monetary policy announcements can leave an imprint on international asset prices, with market reactions to new initiatives announced by the European Central Bank (ECB) in the past few weeks the most recent example. However, event studies tend to find larger international interest rate spillovers for U.S. policy announcements than for those of other central banks.

It is also worth emphasizing that asset purchases are merely one form of monetary accommodation, made necessary when policy interest rates hit their zero lower bound Studies that have compared the spillovers of monetary policy across conventional and unconventional measures generally conclude that the effects on global financial markets are roughly similar.

Because it looks very likely that the program will have fulfilled its objectives, I expect to support a decision to end the asset purchase program at the end of this month.

Richmond Fed President Jeff Lacker said he believes the central bank will not raise its target for short-term interest rates until next year. In an interview in his office this week, Lacker called that scenario “pretty likely,” adding that he does not think the Fed is behind the curve in withdrawing its support for the nation’s economic recovery.

“I don’t see signs of that,” he said.

...

Lacker is a prime example. His dissents in 2012 were driven by two factors: discomfort with the Fed’s official commitment to keeping rates near zero and its purchases of mortgage-backed securities to help the housing market. The latter was an inappropriate use of Fed power, Lacker has long argued, in which the central bank effectively picked winners and losers in the credit markets.

“I think matters of very important principles are at stake,” he said in the interview. “We’re in uncharted territory with regard to the expansiveness of how people talk about what the central banks can do and ought to do.”

Many financial pundits protest that weaning the markets of ber-accommodation, however gradually, risks wreaking havoc, so dependent on central bank largess have the markets become. As a former market operator, I am well aware of this risk. As I have said repeatedly, a bourbon addict doesnt go from Wild Turkey to cold turkey overnight. But even if we stop adding to the potency of the financial punchbowl by finally ending our large-scale asset purchases in October, the punch is still 108 proof. It remains intoxicating stuff.

I believe the time to dilute the punch is close upon us. The FOMC could take two steps to accomplish this after ending our large-scale asset purchases.

First, in October, we could begin tapering our reinvestment of maturing securities and begin incrementally shrinking our portfolio. I do not think this would have significant impact on the economy. Some might worry that paring our reinvestment in MBS might hurt the housing market. But I believe the demand for housing is sufficiently robust to continue improving despite a small rise in mortgage rates. The economy is improving. An acceleration of income growth that will result will likely buttress a recovery in housing and compensate for the loss of momentum that occurred during the winter freeze. (As a sidebar, I note that according to the National Association of Realtors, overseas buyers and new immigrants accounted for $92 billion, or 7 percent, worth of home purchases in the U.S. in the 12 months ended in March, with one-fourth of those purchases coming from Chinese buyers. As Californians, you might find it of interest that Los Angeles was the top destination for real estate searches from China on realtor.com; San Francisco was second; Irvine was third.)

Importantly, I think that reducing our reinvestment of proceeds from maturing securities would be a good first step for the markets to more gently begin discounting the inevitable second step: that early next year, or potentially sooner depending on the pace of economic improvement, the FOMC may well begin to raise interest rates in gradual increments, finally beginning the process of policy normalization.

I was uncomfortable with QE3, the program whereby we committed to a sustained purchase of $85 billion per month of longer-term U.S. Treasury bonds and mortgage-backed securities (MBS). I considered QE3 to be overkill at the time, as our balance sheet had already expanded from $900 billion to $2 trillion by the time we launched it, and financial markets had begun to lift off their bottom. I said so publicly and I argued accordingly in the inner temple of the Fed, the Federal Open Market Committee (FOMC), where we determine monetary policy for the nation. I lost that argument. My learned colleagues felt the need to buy protection from what they feared was a risk of deflation and a further downturn in the economy. I accepted as a consolation prize the agreement, finally reached last December, to taper in graduated steps our large-scale asset purchases of Treasuries and MBS from $85 billion a month to zero this coming October. I said so publicly at the very beginning of this year in my capacity as a voting member of the FOMC. As we have been proceeding along these lines, I have not felt the compulsion to say much, or cast a dissenting vote.

However, given the rapidly improving employment picture, developments on the inflationary front, and my own background as a banker and investment and hedge fund manager, I am finding myself increasingly at odds with some of my respected colleagues at the policy table of the Federal Reserve as well as with the thinking of many notable economists.

[W]ith low interest rates and abundant availability of credit in the nondepository market, the bond markets and other trading markets have spawned an abundance of speculative activity. There is no greater gift to a financial market operatoror anyone, for that matterthan free and abundant money. It reduces the cost of taking risk. But it also burns a hole in the proverbial pocket. It enhances the appeal of things that might not otherwise look so comely. I have likened the effect to that of strapping on what students here at USC and campuses elsewhere call beer goggles. This phenomenon occurs when alcohol renders alluring what might otherwise appear less clever or attractive. And this is, indeed, what has happened to stocks and bonds and other financial investments as a result of the free-flowing liquidity we at the Fed have poured down the throat of the economy.

Many financial pundits protest that weaning the markets of ber-accommodation, however gradually, risks wreaking havoc, so dependent on central bank largess have the markets become. As a former market operator, I am well aware of this risk. As I have said repeatedly, a bourbon addict doesnt go from Wild Turkey to cold turkey overnight. But even if we stop adding to the potency of the financial punchbowl by finally ending our large-scale asset purchases in October, the punch is still 108 proof. It remains intoxicating stuff.

I believe the time to dilute the punch is close upon us. The FOMC could take two steps to accomplish this after ending our large-scale asset purchases.

First, in October, we could begin tapering our reinvestment of maturing securities and begin incrementally shrinking our portfolio. I do not think this would have significant impact on the economy. Some might worry that paring our reinvestment in MBS might hurt the housing market. But I believe the demand for housing is sufficiently robust to continue improving despite a small rise in mortgage rates. The economy is improving. An acceleration of income growth that will result will likely buttress a recovery in housing and compensate for the loss of momentum that occurred during the winter freeze. (As a sidebar, I note that according to the National Association of Realtors, overseas buyers and new immigrants accounted for $92 billion, or 7 percent, worth of home purchases in the U.S. in the 12 months ended in March, with one-fourth of those purchases coming from Chinese buyers. As Californians, you might find it of interest that Los Angeles was the top destination for real estate searches from China on realtor.com; San Francisco was second; Irvine was third.)

Importantly, I think that reducing our reinvestment of proceeds from maturing securities would be a good first step for the markets to more gently begin discounting the inevitable second step: that early next year, or potentially sooner depending on the pace of economic improvement, the FOMC may well begin to raise interest rates in gradual increments, finally beginning the process of policy normalization.

Although theres general consensus that the measures we took in the immediate wake of the crisis were necessary, critics of the Feds policies believe that weve been too accommodative since then, and that after 2010, we shouldve stepped back and let the economy move on its own. I often hear that the economys recovering, so why is the Fed still intervening? Or, in other words: enough is enough.

Ending accommodative policy prematurely would have been a major mistake. In 2010, the economy wasnt yet back on trackin fact, it had barely begun to recover. When we initiated the second round of asset purchases, or QE2, in November 2010, the unemployment rate was around 9 percentonly slightly down from its peak of 10 percent.

The latest round of asset purchasesor QE3was announced in September 2012, when the economy was better, but still well short of healthy. At around 8 percent, the unemployment rate had improved, but was still very high by historical standards, and inflation was running below the preferred 2 percent longer-term goal. In both situations, the very real danger of the recovery stalling and the economy slipping into a state of prolonged stagnation called for additional monetary stimulus.

When you break a leg, you dont just snap the pieces back into place; you leave the cast on until the bone heals. Otherwise, you risk doing even greater damage. And in this case, the economy wasnt ready to walk on its own. Not doing anything, or not doing enough, would just have led to more pain and the need to take even stronger measures down the road.

I was recently in Japan, which offers a real-life example. They shied away from sufficiently aggressive policy and the Japanese economy remained mired in deflationary stagnation for 20 years. Only now are they starting to put more forceful policies into place, and, happily, theyre workingbut those policies are much more forceful than they wouldve been had they been instituted 15 or 20 years ago. In keeping with the patient analogy, you can keep the cast on for a few weeks and let it heal, or you can go without and require extensive surgery later. So if we take the longer view, the Feds actions are in line with people who prefer a light policy touch: were essentially doing less now to avoid having to do more later.

I am aware that not everyone is a fan of the Fed or of accommodative policy. Im not deaf to criticism, and reasonable people disagree on policy all the time. But the bottom line is, it has worked. And the asterisk is that its not permanent. We wont raise interest rates for some time, which is the real marker of tightening policy. However, weve already considerably reduced the pace of our asset purchases, which will likely end this year. Were moving towards normalization, and as the economy continues to improve, well take off the cast; when its able to move on its own, well take away the walking stick. The events of the past several years demanded strong policy action, and we were right to take it. But it doesnt reflect a fundamental shift in our goals or strategy.

Once the U.S. economy began to improve and the Federal Reserve began to discuss a gradual reduction in purchases, many market participants who had been invested “alongside” the Federal Reserve became active sellers.

This was particularly true for highly leveraged investors who had borrowed “short” and invested “long” by engaging in the so-called “carry trade” – an issue of particular relevance in emerging markets where interest rates were considerably higher than in the United States. The result was a fairly significant increase in long-term rates and a repatriation of funds that had been previously borrowed short and invested long in emerging-market securities.

The unwinding of the carry trade was a particularly thorny issue for countries with high dollar-denominated yields and fixed exchange rates, and for countries that had high yields and a policy of gradual currency appreciation relative to the dollar. The speed and magnitude of the readjustment highlighted just how low long-term rates had been pushed. It also suggested that many models of the impact of Federal Reserve asset purchase programs had not fully accounted for potential investor reaction to the policy, or the impact of a sudden reassessment of investor desire to engage in the carry trade.
…
In sum, this episode made clear the importance of Federal Reserve communication and market understanding about programs and policies. The episode also underlined the importance of calibrating investor reactions into models of policy impact.
…
Importantly, I suspect that the benign reaction to the gradual tapering of stimulus over the last 6 months may be instructive as we consider how best to wind down the Federal Reserve’s balance sheet once the tapering of asset purchases is complete. While the optimal program for reducing the Fed’s balance sheet will need to be dependent on the state of the economy, the recent tapering experience suggests to me that a predictable, transparent reduction in the balance sheet could be done in ways that may minimize the risk of financial disruption.

Let me offer one scenario that I have been considering, in light of the sorts of financial stability concerns that I and my colleagues keep in mind. In one scenario, a reduction in the balance sheet, when that becomes appropriate, could be implemented as a basically seamless continuation of the tapering program used for reductions in the purchase program. For example, the Committee could decide to reinvest all but a given percentage of securities on the balance sheet as they reach maturity, and increase that percentage at each subsequent meeting, assuming conditions allow.

Such a tapering of the reinvestment program could allow for a gradual and transparent reduction in the Fed’s balance sheet. As the economy moved closer to the Federal Reserve’s 2 percent inflation target and full employment, there could be a gradual reduction in the reinvestment policy – which would allow for a predictable reduction in the size of the balance sheet. However, the pace of reinvestment should always be considered in the context of the economic outcomes we are seeking to achieve. If the economy was substantially stronger or substantially weaker than was expected, the reinvestment program would need adjustment. Again, I mention this as simply one scenario for consideration.
…
[A] positive collateral impact of using reverse repos and interest on excess reserves in these ways is that the ability to control short-term rates would not be tied to actions that impact the size of the Federal Reserve’s balance sheet. This means the Federal Reserve could have additional financial stability tools at its disposal, if it chose to maintain a larger-than-traditional balance sheet with a greater mix of assets. Naturally, maintaining a large balance sheet comprising both U.S. Treasury securities and mortgage-backed securities would provide the Federal Reserve with continuing options for impacting long-term interest rates and the spread between mortgage-backed securities and U.S. Treasury securities. For example, if a bubble seemed to be developing in housing markets generally, the Federal Reserve would have the option of addressing it by selling MBS or long-duration U.S. Treasury securities. This again is a notion that I consider worthy of further exploration and debate.

In any event, it will be important that the signals conveyed by balance sheet policies are consistent with the forward guidance about future interest rate policies. This has been a difficult communications challenge at times for the FOMC. And it will likely remain a communications challenge as the Committee coordinates the unwinding of the Feds balance sheet with the gradual increase in the policy rate.