In light of this cumulative progress, the FOMC concluded at our June meeting that the criterion of substantial improvement in the outlook for the labor market might well be met over the subsequent year or so. Accordingly, the committee sought to provide more guidance on how the pace of purchases might be adjusted over time.

The committee anticipated in June that subject to certain conditions it might be appropriate to begin to moderate the pace of purchases later this year, continuing to reduce the pace of purchases in measured steps through the first half of next year and ending purchases around mid-year 2014. However, we also made clear at that time that adjustments to the pace of purchases would depend importantly on the evolution of the economic outlook; in particular, on the receipt of evidence supporting the committee's expectation that gains in the labor market would be sustained and that inflation is moving back toward its 2 percent objective over time.

At the meeting concluded earlier today, the sense of the committee was that the broad contours of the medium term economic outlook, including economic growth sufficient to support ongoing gains in the labor market and inflation moving toward its objective, were close to the views it held in June. But in evaluating whether a modest reduction in the pace of asset purchases would be appropriate at this meeting, however, the committee concluded that the economic data do not yet provide sufficient confirmation of its baseline outlook to warrant such a reduction.

The general framework in which we’re operating is still the same. We have a three-part baseline projection, which involves increasing growth that’s picking up over time as fiscal drag is reduced, continuing gains in the labor market, and inflation moving back towards objective. We are looking to see—in the coming meetings, we’ll be looking to see if the data confirm that basic outlook. If it does, we’ll take a first step at some point—possibly later this year—and then continue so long as the data are consistent with that continued progress. And so that basic structure is still in place.

But what I want to emphasize is really two things. First, as I’ve said, asset purchases are not on a preset course, they are conditional on the data. They’ve always been conditional on the data. And, secondly, that even as we move from asset purchases to rate policy as the principal tool of monetary policy, it’s our intent to maintain a highly accommodative policy and to provide the support necessary for our economy to recover and to provide jobs for our citizens.

Well, I think part of the {financial market} reaction we've seen, I mean, comes from a number of sources. Part of it comes from improved economic news. And that's part of the reason why rates have gone up in other countries, as well as in the United States. And that -- to the extent that tighter financial conditions reflect a better outlook, that's a good thing. That's not a problem at all.

Part of it reflects views about monetary policy, in that -- that we want to make sure we get straight. And that's why -- to answer the earlier question again -- it's why communication is so important. We need to explain as best we can how we're going to move and on what basis we're going to move. It's much more difficult today than it was 20 years ago, because the tools are more complex, they're less familiar. But that's still very important.

I think the other factor which was at play was an unwinding of excessively risky and leveraged positions in the markets, and insufficiencies of liquidity in some cases meant that those unwindings led to larger reactions in prices and rates than might otherwise have occurred.

Now, the tightening associated with that is to some extent unwelcome, but on the other hand, to the extent that some of the riskier, more levered positions have been eliminated, I think that makes the situation more sustainable and reduces, at least, the risk that there will be an over-strong reaction to further announcements.

So we will do our best to communicate clearly. That is our goal and our objective. The more clearly we communicate, the better the chance that markets will understand our intentions and that we can avoid any -- any sharp movements. But, again, we're dealing with tools that are less familiar, harder to quantify, and harder to communicate about then the traditional funds rate.

Well, some days I wish that questions like these {about tapering} could be answered with a firm date, a single number and a confident “yes,” accompanied by a fist pump. Unfortunately, the answer to the first question is not as simple as giving a calendar date. Instead, uncertainty over the pace of the recovery means all of this depends on the progress that the economy makes toward our goals of maximum employment and price stability. In my view, this means continuing our current QE3 asset purchase program until three conditions are met. First, the unemployment rate must be in the vicinity of 7 percent with expectations for it to continue falling in a self-sustaining fashion. Second, other important labor market indicators must show a commensurate improvement. And third, we must have considerable confidence that inflation is moving back toward our target of 2 percent. Currently, my best assessment is that by the time these conditions have been met, our QE3 asset purchases since January of this year will total at least $1.25 trillion. By comparison, that would be twice the size of our QE2 purchases made in 2010 and 2011. This would mean that our System Open Market Account (SOMA) balance sheet will reach approximately $4 trillion.

I have been more skeptical than the majority of the Committee that the benefits of continued easing through unconventional policy actions justify the risks, particularly when the economy is growing and financial markets are not under acute stress. While continued accommodative settings for monetary policy are warranted as the recovery proceeds, I continue to support slowing the pace of purchases as the appropriate next step for monetary policy. For example, an appropriate next step toward normalizing monetary policy could be to reduce the pace of purchases from $85 billion to something around $70 billion per month, then have purchases going forward split evenly between Treasury and agency-MBS securities.

“The decision when and if to taper later this year will depend on the data, and specifically are we still seeing signs of positive momentum,” Williams said. “I’m not going to speak about what meeting or not, but I do think that if the data continue to progress as we’ve seen, then I do agree that we should edge down or taper our purchases later this year.”

“I don’t think we have to be in any hurry,” Bullard said today in a CNBC interview from Jackson Hole, Wyoming. “Inflation is running low and we have got mixed data on the economy.”

“So I’d be cautious,” said Bullard, who votes on monetary policy this year. “We want to take our time and assess what’s going on before we make a move.”

“We can afford to be very deliberate in our decision making,” Bullard said.

“I would be supportive in September as long as the data between now and then basically confirm the path we’re on,” Lockhart, who doesn’t vote on monetary policy this year, said today in a CNBC interview from Jackson Hole, Wyoming. “I am confident in a continuation of this sort of moderate growth path.”

“The committee would not normally remove policy accommodation in an environment where inflation is below target and is projected to remain there,” said Bullard, who votes on policy this year, in Louisville, Kentucky.

Bullard suggested the Fed could exercise caution by cutting back by only a modest amount.

"There are a lot of ways this could go. You could do a small amount of tapering versus a larger amount," he said.

As of September, the FOMC will have in hand one more employment report, two reports on inflation, a revision to the second-quarter GDP data, and preliminary incoming signals about growth in the third quarter. I don't expect to have enough data to be sure of my outlook. For that reason, I don't think a decision that commits the Fed to a full phase-out of asset purchases and lays out a precise, beginning-to-end path for doing so would be advisable.

In my mind, the first adjustments to asset purchases, when they occur, should be the beginning of a process with steps that will be determined as later information arrives and certainty about the direction of the economy accumulates. As I see it, a decision to proceed—whether it is in September, October, or December—ought to be thought of as a cautious first step.

Policymaking is quite appropriately forward-looking because monetary policy actions affect the economy with a lag. The rolling outlook from here is what really matters in making future decisions on asset purchases. I will need to get comfortable that the employment progress we've enjoyed is not stalling and that disinflation pressures are not building.

All that said, in considering a decision to reduce purchases, I think it is important to acknowledge the progress that's been made since the launch of QE3. The most recent program of asset purchases has been in force for just short of a year. In August a year ago, the unemployment rate stood at 8.1 percent. A year later, the unemployment rate has fallen to 7.4 percent and monthly job gains, looking back over the year, are averaging just below 200,000. Consumer activity has grown, house prices and housing activity have picked up, and equity markets have shown strength.

We have continued to see steady progress in economic fundamentals, in my opinion. Progress is evident, and we should not lose sight of that.

“If we see the growth pick up in the second half and if we see job gains ... at a higher range, say 180-200,000 – I think with other fundamentals improving we probably are in a position to remove ... the extraordinary policy program over the medium term,” Mr Lockhart said in an interview with Market News International.

But Mr Lockhart noted that the timing of slower asset purchases would depend on the incoming economic data. “[With a] kind of ambiguous picture of mixed data that signal neither accelerating strength nor necessarily deterioration, but that kind of moping along in the middle, then I think it’s not a foregone conclusion that the asset purchase programme should be removed or be removed rapidly,” Mr Lockhart added.

As reported by The Financial Times.

“We’ve seen good improvement in the labor market, there’s no question in my mind about that,” Evans said in Chicago. “I’m still wanting to see greater evidence that it’s a sustainable improvement” and “I would clearly not rule” out a decision to begin dialing back the purchases in September.

The efficacy of this effort is the subject of significant debate, even internally within the FOMC. Some who question the efficacy, including myself, note that the effect of our purchasing MBS and driving down mortgage rates has certainly assisted a robust recovery in housing, and with it, construction jobs and manufacturing and transportation of materials that go into homes… [T]he Fed’s muscling of the yield curve has brought what has been a 30-year-long bond market rally to a crescendo…

Counteracting whatever benefits one can trace to the Fed’s unorthodox policies are some obvious costs. First, savers and others who rely on retirement monies invested in short-maturity fixed-income investments, such as bank CDs and Treasury bills, have seen their income evaporate while the rich and the quick, the big money players of Wall Street have become richer still.

Second, the standard return assumptions of 7.5 to 8 percent for retirement pools, as you well know, have been dashed (though I have always felt they were already calculated on an imaginary and politically convenient basis rather than a realistic one).

Third, accompanying the Fed’s growing balance sheet we have seen a dramatic expansion in the monetary base—the sum of reserves and currency. A basic understanding of demand-pull inflation is “too much money chasing too few goods.” Thus, the excess, currently nondeployed money could prove the kindling of an inflationary conflagration unless the Fed is nimble in managing its effect as it works its way into the economy’s production and consumption of goods and services.

A corollary of reining in this massive monetary stimulus in a timely manner is that financial markets may have become too accustomed to what some have depicted as a Fed “put.” Some have come to expect the Fed to keep the markets levitating indefinitely. This distorts the pricing of financial assets, encourages lazy analysis and can set the groundwork for serious misallocation of capital.

…

Whereas before, our portfolio consisted primarily of instantly tradable short-term Treasury paper, now we hold almost none; our portfolio consists primarily of longer-term Treasuries and MBS. Without delving into the various details and adjustments that could be made (such as considerations of assets readily available for purchase by the Fed), we now hold roughly 20 percent of the stock and continue to buy more than 25 percent of the gross issuance of Treasury notes and bonds. Further, we hold more than 25 percent of MBS outstanding and continue to take down more than 30 percent of gross new MBS issuance. Also, our current rate of MBS purchases far outpaces the net monthly supply of MBS.

The point is: We own a significant slice of these critical markets. This is, indeed, something of a Gordian Knot…

There is no Alexander to simply slice the complex knot that we have created with our rounds of QE. Instead, when the right time comes, we must carefully remove the program's pole pin and gingerly unwind it so as not to prompt market havoc. For starters though, we need to stop building upon the knot. For this reason, I have advocated that we socialize the idea of the inevitability of our dialing back and eventually ending our LSAPs. In June, I argued for the Chairman to signal this possibility at his last press conference and at last week’s meeting suggested that we should gird our loins to make our first move this fall. We shall see if that recommendation obtains with the majority of the Committee.

Mr. Bullard noted that inflation is “low” and “near the lower edge of acceptable outcomes.” He added “on balance, inflation expectations have declined since March,” and said the Fed “would not normally remove policy accommodation in an environment where inflation is below target and is projected to remain there.”

The central banker said it is also challenging to determine the health of the labor market. Based on the unemployment rate alone, declines in joblessness suggest improvement, while other measures indicate the labor market might not be as healthy as the unemployment rate indicates. But he also said “payroll employment growth has generally been strong.”

On the growth front, recent data have been “weak,” but there are reasons to be hopeful for the future. “Real estate markets are improving, equity markets have rallied, the European sovereign debt crisis remains subdued for now, U.S. fiscal brinkmanship has been less of a problem, and household deleveraging is further along,” he explained.

That said, Mr. Bullard said “I have generally been too optimistic” about the economy’s outlook, and because of this, “I think caution is warranted in taking policy action based on forecasts alone.”

In other comments, the central banker noted the Fed’s balance sheet, while very large in absolute terms, “is not particularly large when scaled by GDP and compared to other major central banks, or when compared to historical data on the Fed’s balance sheet.”

As reported by WSJ.