I've had an interesting few weeks, starting with attending the Federal Reserve Bank of Kansas City's annual economic symposium at Jackson Hole to finalizing a cooperative research initiative between the Atlanta Fed's Americas Center and the Institute for Applied Economic Research (note that IPEA's site is in Portuguese) in Brasilia to a speech delivered at the BM&F Bovespa fourth International Financial and Capital Markets Conference in Campos do Jordao, Brazil, to sitting in on the Central Bank of Argentina's annual Money and Banking Conference. I tell you this not because I expect you to care about my schedule but because I plan to spend the next couple of blog entries trying to spill out of my brain some impressions and thoughts gleaned from some intensive listening to a wide variety of policymakers, businesspeople, and scholars explaining their impressions and thoughts.

On several points there appears to be a fair amount of unanimity: (1) the global economy has moved a substantial distance from the financial abyss and is now in the first phase of recovery from recession; (2) the financial crisis of the past two years originated, ironically, in the economies with the deepest and most advanced financial markets—the United States representing the epicenter—and is likely to have the longest-lasting negative impact in those economies; and (3) the time has come to begin building a market and regulatory infrastructure that can avoid or withstand the events associated with the recent turmoil.

There is a lot to be said on these topics, but a great starting point for conversation was provided at the Central Bank of Argentina's conference by Dr. Kiyohiko Nishimura, deputy governor of the Bank of Japan. Commenting on policy lessons learned, Governor Nishimura first offered two criteria that ought to govern the application of “unconventional” policy tools in times of stress, defined as those policies that entail microeconomic interventions and explicit risk taking by the monetary authority:

First criterion: Unconventional policies, as defined, inevitably distort resource allocation. The potential benefits of any such intervention must therefore be measured against these costs.

Second criterion: Because unconventional approaches entail taking more than the usual amount of risk onto a central bank's balance sheet, measures should be taken to guarantee adequate capital buffers in the event that substantial losses are incurred.

Assuming these criteria are met, what principles should guide the implementation of unconventional policies? Governor Nishimura offered four:

  1. Select what sorts of interventions are most important, and concentrate efforts in those areas.
  2. In any intervention, be careful to think through the broad implications and avoid creating further dysfunction. The example given was the decision, made by most central banks, to set the effective floor on short-run policy rates near but above zero in order to avoid eliminating all returns to participating in the market and hence driving all private players out of the market.
  3. Provide extensive safety nets to short-circuit panics.
  4. Design measures to be “self-fading” as conditions improve. For example, set lending rates and terms at levels above those that private providers will set once the acute phase of a crisis passes.

And finally the governor articulated a list of don'ts gleaned from the crisis experience:

  1. Don't assume the size of the balance sheet is, by itself, an adequate measure of monetary ease or support.
  2. Don't concentrate only on those markets where specific interventions are taken. Improvement in those markets cannot be assumed successful unless they improve the big picture.
  3. Don't underestimate the value of safety nets.
  4. Don't ignore the differences across countries or regions. One-size-fits-all strategies may fit no one.
  5. Don't assume things are going to return to the old status quo.

These thoughts strike me as a very nice departure point for debate and deliberation. If I were to add anything, I would start with one more item on the list of criteria: Unconventional policies should commence from a broad consensus on what level of risk is acceptable for a central bank to take in the “normal” course of business. In other words, the definition of unconventional policy assumes that there is a definition of conventional policy, and there should be collective agreement on how that is defined.

In the good old precrisis days, conventional policy was pretty straightforward: a buy-only-Treasury-securities policy for conducting open market operations and direct lending restricted to a narrow set of financial institutions (commercial banks) and largely on an overnight basis. There are good enough reasons to think that this is the way it should it be—Marvin Goodfriend, among others, has made this argument. But, in my view, I believe it is warranted to take to heart point number five on Governor Nishimora's list of don'ts and not yet assume the new conventional policy will look like the old one.

By Dave Altig, senior vice president and research director at the Atlanta Fed