Monday, July 10, 2006

Pimco's Paul McCulley on Keynsian thought meeting Austrian thought!

A Kind Word for the Austrian School
By Paul McCulley

Bill Gross teases me that I'm not just a Keynesian, but a religious Keynesian. There is an element of truth to that, as evidenced by the fact that the only major piece of art that I own is a portrait of Keynes by Salisbury, the preeminent portrait artist of Keynes' time. And it hangs in my office, presumably watching over me. So, Bill does have a point.

As a practical matter, however, I also have deep appreciation for other schools of economic thought. And, yes, that includes the Austrian School. Indeed, there is much overlap between Keynesian thought and Austrian thought, even though many think they are antithetical lines of thinking.

I was reminded of this recently when reading a brilliant essay by William White, of the Bank for International Settlements - Is Price Stability Enough? Mr. White's core thesis is that low and stable inflation in goods and services prices, presumably the holy grail of central banking, is not a guarantee of steady growth in real economic activity.

To be sure, that has been the case over the last two decades, at least in the United States, a dynamic known as the Great Moderation: lower volatility in both inflation and growth. There is a rich professional literature arguing the sources of this blessed outcome, with the consensus being that it has been the consequence of greater reliance on private markets - both in the United States and 'round the world in the allocation of resources; better macroeconomic policies, and in particular better monetary policy; and old fashioned good luck.

This consensus, at least in the United States and especially in the halls of the Federal Reserve, leads to the conclusion that monetary policymakers should:
Always and everywhere target - implicitly or explicitly - low and stable inflation and perhaps even more important, low and stable inflationary expectations.

Do nothing in response to changes in asset price valuations unless those changes imply an undesirable course for aggregate demand growth relative to aggregate supply growth, implying an undesirable path for the output gap and, thus, inflation and inflationary expectations.
Taylor's Austrian Constant

This is standard Keynesian stuff, married to a Phillips Curve. Or, for those who remember the jargon from school days, it's the IS-LM model married to NAIRU. It is also the foundation of the ubiquitous Taylor Rule, with one major exception: Mr. Taylor has a "neutral" real rate as a constant in his Rule.

And the concept of a "neutral" real rate is a very Austrian notion: there exists some unobservable interest rate (called the "natural" rate by Austrians) that perfectly matches the time preference of lenders and borrowers and in the absence of fiat currencies, free markets would find this rate (as if by Adam Smith's invisible hand), bringing about just the right amount of investment and savings.

In turn, Austrians argue that if fiat currencies do exist and if policymakers peg rates below the natural rate, there will be an excess of investment relative to that level which would generate returns consistent with the natural cost of borrowing, producing an investment bubble, which will be revealed when policymakers lift rates to or above the "natural" rate, generating an investment bust.

This is where the Austrian School is in direct opposition to the Keynesian School. For Keynes, investment was not a function of savings, but rather savings was a function of investment, which drives income, from which savings flow. Yes, after the fact, savings and investment must, as an accounting matter, equal each other.

But there was no a priori reason, Keynes argued, that savings and investment would ex post equal each other at full employment. For Keynes, what mattered was ex ante investment desires, which were driven by the state of confidence in long-term return expectations, summarized as animal spirits.

Thus, for Keynes there was no magical natural - or neutral - rate of interest. Indeed, Keynes actually put more emphasis on the role of stock prices than interest rates in eliciting the ex ante desire to invest. Yes, the great James Tobin found his Q in Chapter 12 of Keynes' General Theory!

Which brings us back to the ironic crossover between the Keynesian School and the Austrian School: if investment is a function of animal spirits and if animal spirits get their mojo from asset prices, then boom and bust in investment is endemic to the capitalistic system. That doesn't mean that it's not the best system ever devised for allocating resources. It is. But it is also inherently given to cycles of euphoria followed by doom.

For the Austrians, the solution to this problem was to kill fiat money systems and in the absence of that prescription, to resist using fiat money systems to soften periods of busts. For Keynes, writing at a time when the Austrian "solution" was de facto being applied, called the Great Depression, this made no sense at all. If investment and savings were equaling each other at 25% unemployment, then it was the duty of the sovereign to incite animal spirits with monetary ease, while even more important, boosting public investment.

Could this prescription lead to ex post maldistribution of investment, as the Austrians argued? Yes, Keynes acknowledged that; indeed, his very own work - Chapter 12 again! - implied this outcome. But for him, what mattered most was to achieve full employment. And he didn't trust unfettered markets to consistently deliver that outcome, as the Austrians argued.

One Step From Targeting Asset Prices

Which brings us back to the Taylor Rule, with its "neutral" real rate constant, a very Austrian concept. Recognizing that "flaw," many researchers, including at the Fed, have been empirically trying to model a time-varying "neutral" rate. I applaud that work. But once you take that step, you are only one step away from targeting asset prices, as the key variable driving fluctuations in the neutral rate are fluctuations in risk premiums, in response to changes in animal spirits. Thus, in my view, asset prices should matter for monetary policy and not just through the output gap-inflation channel.

Indeed, this is precisely the point of Mr. White's essay, as well as that of a speech I gave to the Money Marketeers Club of NYU in February: too much focus - and too much success - in stabilizing goods and services inflation on one- to two-year horizons is a prescription for boom and bust in asset prices and, from a starting point of low inflation, is actually a prescription for both increased volatility and deflation on longer horizons.

Ironically, Mr. Greenspan more or less said the same thing last September when he declared:
"In perhaps what must be the greatest irony of economic policymaking, success at stabilization carries its own risks. Monetary policy - in fact, all economic policy - to the extent that it is successful over a prolonged period, will reduce economic variability and, hence, perceived credit risk and interest rate term premiums."
This is especially the case when a central bank - such as the Fed - explicitly promises not to lean against bubbles unless they portend a negative impact on the inflation outlook for goods and services prices, while also having a clear track record of "mopping up" after bubbles burst. This asymmetric reaction function - commonly called the Greenspan Put - is a form of moral hazard, which actually makes bubbles more likely.

Mr. Greenspan rejects, of course, the notion that the Fed put a floor under asset prices during his tenure, notably the stock market, noting that if it was a floor, it was a floor in the cellar of a tall building. Narrowly, that is correct. But broadly speaking, it is an incontrovertible truth that the Fed under Greenspan ran an asymmetric reaction function, with market participants fully aware that was what the Fed was doing. By definition, such a policy should be expected to contract risk premiums and lift asset price valuations.

But, as Mr. Greenspan himself said last summer in his valedictory address at Jackson Hole, "history has not dealt kindly with the aftermath of protracted periods of low risk premiums." Thus, it is hard for me to avoid the conclusion that too much success in stabilizing goods and services inflation, while conducting an asymmetric reaction function to asset price inflation and deflation, is a dangerous strategy.

Yes, it can work for a time. But precisely because it can work for a time, it sows the seeds of its own demise. Or, as the great Hyman Minsky declared, stability is ultimately destabilizing, because of the asset price and credit excesses that stability begets. Put differently, stability can never be a destination, only a journey to instability.

Keynes knew that. The Austrians knew that. On that, they agreed. What they disagreed about was whether instability was caused by fiat money makers holding real rates away from the natural level. For Keynes there was no ex ante natural rate level, only an ex ante imperative for policy-makers to pursue full employment. And the key to that was maintaining investment - private and public. For Keynes, there was no natural rate of interest at which that would happen, because investment is inherently the product of animal spirits, which fluctuate in animal-spirited ways.

Thus, the notion that the Fed can ignore asset prices except to the extent they matter in a Taylor Rule framework is, to my way of thinking, loosing its moorings. The putative neutral real rate is indeed time varying, not a constant as Taylor assumed. Actually, I think most Fed officials agree with that proposition, including Chairman Bernanke.

What they haven't embraced, however, is that the dominant variable driving fluctuations in the neutral rate is fluctuations in risk appetites in the private sector, which have a pro-cyclical reflexive quality, fueled by moral hazard. And that is a prescription for both Austrian maldistribution of resources as well as long term deflationary risk.

To be sure, those outcomes haven't bitten either the U.S. or global economy in the bum yet. But that doesn't mean they won't. In fact, the logic of the serial-bubble game suggests that the longer it is played, the more difficult it is to play, as the number of under-levered assets declines every time the game is played.

How to get off this treadmill? First and foremost, I think, as discussed in my Money Marketeers speech, the Fed needs to set its inflation target - or comfort zone, as it is called - high enough so that the U.S. economy - and indeed the global economy - could absorb a shock to aggregate demand, perhaps from a bursting property market bubble, without generating palpable risks of deflation, triggering once again the Fed's asymmetric policy of aggressive easing. Second, I think policymakers should be more wary of excessive pre-commitments to the markets.

This is not, I want to stress, the same thing as reducing transparency. As New York Fed President Giethner articulated brilliantly in a speech last month, transparency involves candidly presenting what you know and also what you don't know. Acknowledging genuine uncertainty is not being opaque, so long as the sources and nature of the uncertainty is transparently communicated. In contrast, articulating certainty when uncertainty is reality is not transparency, but a disservice called moral hazard.

Bottom Line

Yes, I am a Keynesian. But more precisely, I'm a Keynesian wearing Minsky clothing, and doffing Austrian shoes. In the fullness of time, I expect Chairman Bernanke, a brilliant Keynesian, to rediscover that the Austrians were not all wet in their diagnosis of the potential for maldistribution of investment, even though they were soaking wet about what to do about it. The Austrians said let the asset price and credit excesses purge themselves. A much better way, I believe, is to lean against the excesses preemptively, using all available tools, including regulatory tools.

Yes, inflation targeting is fine. Myopic inflation targeting is not. Asset prices matter, and not just in the context of their influence on aggregate demand relative to aggregate supply and, thus, inflation. Asset prices matter in their own right, because wild swings in asset prices, even in the context of "stable" goods and services inflation, are a source of both volatility and maldistribution in investment.

And, in the long run, a source of deflationary, not inflationary risk.