In such circumstances, the most likely outcome ... for the economy may be fairly benign, but there may be a significant risk of more severe adverse outcomes. In such circumstances, the central bank may prefer to take out insurance by easing the stance of policy further than if the distribution of probable outcomes were perceived as fairly symmetric...

...

However, it is important to recognize that financial markets can also turn around quickly, thereby reducing the drag on the economy as well as the degree of tail risk. Therefore, the central bank needs to monitor credit spreads and other incoming data for signs of financial market recovery and, if necessary, take back some of the insurance; thus, at each stage of the episode, the appropriate monetary policy may exhibit much less smoothing than would be typical in other circumstances.

Most of the quantitative studies of optimal monetary policy have also assumed that the shocks hitting the economy have a time-invariant Gaussian distribution, that is, a classical bell curve with symmetric and well-behaved tails. In reality, however, the distribution of shocks hitting the economy is more complex. In some instances, the uncertainty facing the economy is clearly skewed in one direction or another; again, this is likely when there are significant financial disruptions. The Federal Reserve often reports on our judgments regarding the degree of skewness and the associated economic costs by giving assessments of the “Balance of Risks” in the press releases that are issued following FOMC meetings.

In addition, at least in some circumstances, the shocks hitting the economy may exhibit excess kurtosis, commonly referred to as tail risk because the probability of relatively large disturbances is higher than would be implied by a Gaussian distribution. In that light, one element of the recent enhancements to the Federal Reserve’s communication strategy is that FOMC participants now provide assessments of the relative degree of uncertainty. For example, in the “Summary of Economic Projections”issued in late November, FOMC participants indicated that the degree of uncertainty regarding the economic growth outlook was relatively high compared to the average degree of uncertainty over the past two decades. This account could be interpreted as a statement that the Committee perceived the tail risk as unusually large.

With a nonquadratic objective function (consistent with the importance of uncertainty for the course of monetary policy) as well as nonlinear dynamics and non-Gaussian shocks, optimal monetary policy will also be nonlinear and will tend to focus on risk management.

 In practice, policy decisions are likely to depend on more than the modal forecast used by the authors, such as the degree of uncertainty, the risks around the central tendency, and a weighting of the costs to public welfare from missing the forecast on one side or another. 

We clearly must be vigilant about these risks to economic growth. However, overly accommodative liquidity provision could endanger price stability, which is the second component of the dual mandate. After all, inflation is a monetary phenomenon. Indeed, one of the many reasons for the Fed's commitment to low and stable inflation is that inflation itself can destabilize financial markets.

...

That is, the Fed must adjust the stance of policy to guard against the risk of events that may have low probability but, if they did occur, would present an especially notable threat to sustainable growth or price stability...  But while the risk is still present of notably weaker-than-expected overall economic activity, given the policy insurance we have put in place I don't see this as likely.

Just as an analytical risk-management framework is fundamental to the safe and sound operation of large banking and financial institutions, it is also, I believe, essential for sound monetary policy-making. 

The Federal Open Market Committee recently announced that it will increase the frequency and expand the content of its economic projections.  A clear understanding of the risk-management framework should help improve the public's comprehension of these expanded announcements and thereby, I believe, improve the efficacy of monetary policy actions and the overall functioning of the economy. 

To me, the uncertainties about how financial conditions might evolve and affect the real economy mean that risk management considerations have an important role in the current policy environment. The cutback in nonconforming mortgage originations and the continued high level of inventories of unsold homes will result in further weakness in housing markets. Under one scenario, the effects on overall growth will be fairly isolated to declines in residential construction similar to our experience in 2006 and early 2007. However, there is a less benign possibility. Housing demand and prices could weaken a good deal more than we expect either because a new shock hits the sector or because we have underestimated the weakness already in train. A more pronounced downturn could weigh more heavily on consumer spending. In addition, further delinquencies and foreclosures could add to the problems with mortgage-backed securities. This, in turn, could generate further adverse effects on financial conditions that support economic activity. Together, such events would pose a more serious downside risk to growth. I want to emphasize that I do not see this extreme outcome as likely. But it is one of those high cost outcomes that we should guard against.

The last item on my list of limitations was that simple rules do not take account of risk-management considerations.  As shown in Figure 2A, the core CPI inflation rate for 2003 was falling toward 1 percent.  The real-time reading of the core PCE inflation rate (not shown) was on average even lower than the comparable CPI figure.  Given these rates, the possibility of deflation could not be ruled out.  We had carefully analyzed the Japanese experience of the early 1990s; our conclusion was that aggressively moving against the risk of deflation would pay dividends by reducing the odds on needing to deal with the zero bound on nominal interest rates should the economy be hit with another negative shock.  This factor is not captured by simple policy rules.

[T]he fourth principle is that any monetary policy rule should be robust to uncertainty.5 Indeed, the specification of the original Taylor Rule was not chosen to be optimal in any one particular model, but was based on its "good" performance in monetary-policy-rule evaluation exercises using a variety of macroeconomic models.6 This approach places greatest weight on getting the "basics" right; that is, it emphasizes policy prescriptions in which we have the most confidence. This approach is purposefully modest in that it does not attempt to take advantage of all the potential benefits of the optimal policy in a given model. In fact, subsequent research has shown that the cost of insuring against model misspecification is relatively small because fully optimal rules yield typically only small stabilization benefits over simple rules like the Taylor Rule.7

Unlike the other principles which are uncontroversial, this last principle is still the subject of research and debate. But, based on my experience, John's position on the benefits of robustness seems the right one to me.

Of course, gradualism and model averaging may not be appropriate in all circumstances.  For example, it may be necessary for monetary policy to respond to what might be called "tail events," along the lines suggested by recent work on "robust control."  To simplify greatly, this approach often amounts to choosing policy settings to minimize the maximum possible loss across different models of the economy, in contrast to the standard Bayesian approach, which (loosely speaking) seeks to minimize the average loss across models.  Much of the research on robust control has been a bit technical and esoteric.  But the notion that policymakers may at times base policy settings on especially pernicious risks has an important ring of truth. 

For example, in 2003 the FOMC noted that a continued fall in inflation would be unwelcome largely because such an eventuality might potentially lead to persistently weak real activity with interest rates stuck at zero.  Partly in response, the FOMC reduced the federal funds rate to an unusually low level and kept it there for an extended period, in a manner that perhaps would not have occurred in the absence of concerns about the "worst case" effects of deflation.  This type of risk management--in which the central bank takes out some insurance against a bad but improbable event--has been an aspect of policymaking for some time and does seem to respond to extreme risks in a way reminiscent of the literature on robust control.

[T]he judgments associated with monetary policy are not always as obvious as they might appear. Economic conditions can be unpredictable, and we are under no illusion that we are perfect forecasters. To paraphrase former Federal Reserve Chairman Alan Greenspan, we are essentially risk managers. We need to make policy choices that stand the best chance of moving us toward our objectives, given our imperfect understanding of the changing environment around us.

The Bank’s baseline forecast assumes that if energy prices stabilize as indicated in the futures market, core inflation will gradually subside.  That is my best guess at this point and, based on inflation expectations measured in a variety of ways, private-sector individuals, businesses, and financial markets appear to agree.  But, as others have said repeatedly, monetary policy is about risk management.  A key risk is that inflation will continue to rise or persist at high levels and embed itself in consumer and business plans.  Managing that risk is clearly important, and a matter about which central banks need to be quite vigilant – as I believe the FOMC has been and will continue to be.

By 2030, almost one in five U.S. residents will be 65 years or older.  Well before then, beginning in about 2018, Social Security will start to pay out more in ben­efits than it receives from payroll taxes.  Even before that, -- in the neighborhood of 2010 -- Social Security will start exerting upward pressure on the unified federal budget deficit as its surplus diminishes, with a consequent reduction in net public saving, absent changes in the program itself, increased taxes, or reduced spending on other government programs.  

The situation for Medicare is similar and, potentially even more serious.  Payroll taxes to cover Medicare expenditures are currently in surplus.  Over time, however, Medicare spending is expected to increase more rapidly than related tax revenues, creating a deficit prob­lem that analysts see as potentially greater in size and more difficult to deal with than that associated with Social Security.   Thus, despite the relatively benign federal deficit we currently see, it is clear the situation will worsen dramatically over the next decade.  And, unlike the late '80s when deficits became a national concern, there seems to be no political consensus on the nature of this problem or its resolution -- a fact that should be a concern to all of us.

While it's likely that the slowdown in the housing sector will have only moderating effects on economic activity and will continue to unfold in an orderly way, I should note that we can't ignore the risk that a more unpleasant scenario might develop. In particular, we have heard a lot in recent years about the possibility that there is a house-price "bubble," implying that prices got out of line with the fundamental value of houses and that the current softening could be just the beginning of a steep fall. While I doubt that we'll see anything like a "popping of the bubble"—in part because I'm not convinced there is a bubble, at least on a national level—it is a risk we have to watch out for.

My final observation is that all of this acquires greater significance at the current juncture. The world economy has been experiencing, and may now be emerging from, an unusually long period of historically low inflation-adjusted policy interest rates, unusually strong expansion in global liquidity, exceptionally buoyant asset prices, and strong global growth. And this has occurred at a time when globalisation should, if anything, have raised the world's natural interest rate, by boosting global growth potential while helping to keep a lid on inflation. But just because inflation has remained remarkably quiescent so far, should we assume that all is fine?

I say this in that spirit of humility which should underpin all policies. We must always remain alert and avoid complacency. The major policy mistakes in history that I highlighted earlier were made precisely when policymakers felt they had finally come to master the secrets of the economy. Preserving macroeconomic and with it, financial stability, is essential. We should not forget that it was financial stress at the core of the global economy, and its international ramifications, that proved fatal to the first wave of globalisation. I certainly do not have the answers. But these questions are worth asking.

From a comment on a paper by Frederic Mishkin on the benefits of globalization and the need for emerging markets to embrace more open capital markets

Given this reality, policymakers are well advised to follow two principles familiar to navigators throughout the ages:  First, determine your position frequently. Second, use as many guides or landmarks as are available In the context of monetary policy, these principles suggest that policymakers should monitor bond yields carefully in judging the current state of the economy--but only in tandem with the signals from other important financial variables; direct readings on spending, production, and prices; and a goodly helping of qualitative information. Ultimately, a robust approach to policymaking requires the use of multiple sources of information and multiple methods of analysis, combined with frequent reality checks. By not tying policy to a small set of forecast indicators, we may sacrifice some degree of simplicity, but we are less likely to be misled when a favored variable behaves in an unusual manner.