Risk, Uncertainty, and Macroeconomic Policy

There is a distinction between risk and uncertainty:

Uncertainty is a measure of our ignorance. Risk is what remains when we know everything that can be known.

Edmund Phelps says this distinction helps to explain why monetary policy rules and financial engineering based upon the neoclassical model, which relies upon the mistaken idea that we face known, quantifiable risks, are not holding up very well in the current environment where there is uncertainty about the nature of the risks that market participants face. Given this, what would he do different? Instead of lowering interest rates to combat unemployment, accept that the natural rate of unemployment has increased recently, and raise interest rates to prevent inflation:

Our Uncertain Economy, by Edmund Phelps, Commentary, NY Times: In recent times, most economists have pretended that the economy is essentially predictable and understandable. Economic decision- and policy-making in the private and public sectors, the thinking goes, can be reduced to a science. Today we are seeing consequences of this conceit in the financial industries and central banking. "Financial engineering" and "rule-based" monetary policy, by considering uncertain knowledge to be certain knowledge, are taking us in a hazardous direction.

Predictability was not always the economic fashion. In the 1920s, Frank Knight at the University of Chicago viewed the capitalist economy as shot through with "unmeasurable" risks, which he called "uncertainty." John Maynard Keynes wrote of the consequences of Knightian uncertainty for rational action.

Friedrich Hayek began a movement to bring key points of uncertainty theory into the macroeconomics of employment -- a modernist movement later resumed when Milton Friedman and I started the "micro foundations of macro" in the 1960s.

In the 1970s, though, a new school of neo-neoclassical economists proposed that the market economy, though noisy, was basically predictable. All the risks in the economy, it was claimed, are driven by purely random shocks -- like coin throws -- subject to known probabilities, and not by innovations whose uncertain effects cannot be predicted.

This model took hold in American economics and soon practitioners sought to apply it. Quantitative finance theory became a tool relied on by most banks and hedge funds. Policy rules based on this model were adopted at the Federal Reserve and other central banks.

The neo-neoclassicals claimed big benefits from these changes. They boasted that their statistical approach to risk made the financial sector much more effective in matching lenders with borrowers, with vast savings in labor and increases in profits. They asserted a decline in "volatility" in the U.S. economy and credited it to the monetary policy rules at the Fed.

Current experience is putting these claims to the test.

Subprime lending and the securitization of debt was an innovation that, it was believed, offered the prospect of increasing homeownership. But "risk management" was out of its depth here: It had no past data from which to estimate needed valuations on the novel assets, it did not allow for possible macroeconomic dynamics, and it took inadequate account of the system effects of unknown numbers of entrants into the new business all at nearly the same.

The claim for rule-based monetary policy is weak on its face. In deciding on the short-term interest rate it controls (the Fed funds rate) the Federal Reserve thinks about the "natural" interest rate -- the rate needed if inflation is neither to rise nor fall. Then the Fed asks whether the expected inflation rate is above or below the target. The Fed also asks whether the unemployment rate is above or below the medium-run "natural" unemployment level -- the level to which sooner or later the actual rate will return. ... [...continue reading...]

Comments

[Spectator]

Thanks for posting this. It's been hard to apportion blame between the two primary causes of this crisis, moral hazard and artificially cheap money. While some would blame private market players, and others lack of oversight/regulation, I feel that is misplaced.

That the economics establishment can place so much faith in models from an admittedly social science, defies belief. We're not talking small experiments here. We can only wait for the day when economists and their models are completely discredited, and they go the way of socialist central planners. Only the Austrian school has the necessary humility about how little economists know.

[Negative Natural Rate]

WSJ..."The Fed's view seems to be that the natural interest rate has decreased with the business downturn."

Is it really likely that the natural interest rate has decreased below the rate of inflation?

[paine]

"Uncertainty is a measure of our ignorance. Risk is what remains when we know everything that can be known."

knight actually conflated these things some
or at least got himself jumbled up over the interaction betwixt

brad seems to be crystal clear
a priori at least

scratch one up
for delongeurs

[PHYRON]

As usual Wittgenstein said it best:
Certainty...is just the inability to imagine the alternative....

[paine]

witt also said
un-certainty is the inability to see the real

see out there versus imagine in here

spooly shit man

[ken melvin]

One can not a silk purse from the sow's ear make; unless of course they use innovative tailoring.

[david]

It's like hearing jug band music make its way from the college radio station to the public radio oldies show, seeing risk and uncertainty pop up like this in the papers. Which I suppose is why neo-neoclassicism gets lumped in with that post-funky disco in Phelps' mind.

[Steve J.]

But "risk management" was out of its depth here: It had no past data from which to estimate needed valuations on the novel assets, it did not allow for possible macroeconomic dynamics, and it took inadequate account of the system effects of unknown numbers of entrants into the new business all at nearly the same.

So why did the ratings agencies hand out AAAs like they were popcorn?

[hari]

I've been dealing with risk analysis in global politicl economy and there're no models to depend on to tell you how the story will impinge on markets - positively or negatively. However, if you've some critical historical knowledge of the actor(s)/country(ies) in question, it's possible to eliminate socalled uncertainty - atleast in the short run. Long term, it gets a bit more dicey and one doesn't or simply can't forecast developments accurately.
But one gets paid for country-specific analysis for use by market makers.

Here, what we've is a serious bottleneck because "risk and uncertainty" in macroeconomic models - even if quantifiable - are NOT likely to be (always) accurate or reliable. Simply because the data base from which these socalled throughputs are aggregated are not based on rocket science. They're more often than not guess estimates and whatnots.

[Terms like "natural rate of unemployment" and whatnot are baffling for purposes of polical analysis!]

Of course, current neo-classical determinates will come in for serious discourse and criticism. Certainty in the sphere of economic management, by definition, must belong to the modelling gurus. The relevance to what happens in the real economy is sadly more difficult to discern, since we're dealing with all of its imperfections.

My take, at this point in time, is to simply and accurately define the nature of economic uncertainty in dealing with monetary economic policy based on an imperfect econometric model of anaylsis.

We've to admit our error of judgement and try to go forward and see if there are still better ways to deal with the real world or not. Back to the drawing board!

[hari]

I've been dealing with risk analysis in global politicl economy and there're no models to depend on to tell you how the story will impinge on markets - positively or negatively. However, if you've some critical historical knowledge of the actor(s)/country(ies) in question, it's possible to eliminate socalled uncertainty - atleast in the short run. Long term, it gets a bit more dicey and one doesn't or simply can't forecast developments accurately.
But one gets paid for country-specific analysis for use by market makers.

Here, what we've is a serious bottleneck because "risk and uncertainty" in macroeconomic models - even if quantifiable - are NOT likely to be (always) accurate or reliable. Simply because the data base from which these socalled throughputs are aggregated are not based on rocket science. They're more often than not guess estimates and whatnots.

[Terms like "natural rate of unemployment" and whatnot are baffling for purposes of polical analysis!]

Of course, current neo-classical determinates will come in for serious discourse and criticism. Certainty in the sphere of economic management, by definition, must belong to the modelling gurus. The relevance to what happens in the real economy is sadly more difficult to discern, since we're dealing with all of its imperfections.

My take, at this point in time, is to simply and accurately define the nature of economic uncertainty in dealing with monetary economic policy based on an imperfect econometric model of anaylsis.

We've to admit our error of judgement and try to go forward and see if there are still better ways to deal with the real world or not. Back to the drawing board!

[GREG RANSOM]

Hey Mark. How is your study of "Austrian" economics going?

[kio]

Thank you for the link. Phelps raises many important questions on the "residuals" left after application of economic theories. Since the problems are still open, everybody (including me) has an opportunity to put forward own concepts, which cover the holes left by the conventional approach.

Illustrations, as an important part of any explanation, do not give any chance to present full responce here. So, if interested, try http://inflationusa.blogspot.com/2008/03/edmund-phelps-on-uncertainty-in.html

I have put a link to this post, but do not see it in the trackback

[Bruce Wilder]

ep: "In deciding on the short-term interest rate it controls (the Fed funds rate) the Federal Reserve thinks about the "natural" interest rate -- the rate needed if inflation is neither to rise nor fall. Then the Fed asks whether the expected inflation rate is above or below the target. The Fed also asks whether the unemployment rate is above or below the medium-run "natural" unemployment level -- the level to which sooner or later the actual rate will return."

Witt via paine:
"un-certainty is the inability to see the real

see out there versus imagine in here"

Inability or unwillingness?

Phelps is all about imagining "natural" rates -- unemployment rates, interest rates, every one of them a Goldilocks number, though it may gyrate, and always a subjective assessment, never an objective statistic.

What Phelps practices does not seem to me to be economics, so much as it resembles masturbation alone in the dark. (Maybe not alone; are there U of Chicago Jamborees for U of Chicago boy scouts?) He wants to get the "natural" unemployment rate (or excuse me, not just any natural u.r., but the medium-term natural unemployment rate, right with the "natural" interest rate. These things are always changing though, so I guess judgment and technique are critical. Dealing with the actual, unnatural economy -- well, that would just be "unnatural".

[Bruce Wilder]

Uncertainty is not knowing what will happen.

Risk is the consequence for welfare of not knowing what will happen.

Uncertainty is reduced, to the extent that you know or can control what happens.

Risk is reduced, to the extent that it matters less what happens.

What happens is the outcome.

The consequence of what happens is the payoff.

Uncertainty about the outcome makes the payoff risky.

Knowledge reduces the uncertainty.

Knowledge facilitating control changes the outcome.

Insurance changes the payoff.

[anonymous]

As far as uncertainty goes, remember this blast from the past:

March 16, 2006 press release (http://www.federalreserve.gov/releases/h6/ 20060316/)

“Discontinuance of M3---As announced on November 10, 2005, the Board of Governors will cease publication of the M3 monetary aggregate on March 23, 2006. The Board will also cease publishing the following components: large-denomination time deposits, repurchase agreements (RPs), and Eurodollars....
M3 does not appear to convey any additional information about economic activity that is not already embodied in M2 and has not played a role in the monetary policy process for many years. Consequently, the Board judged that the costs of collecting the underlying data and publishing M3 outweigh the benefits."

At the time, I remember that there were many who found this troubling although the Fed obviously didn't.

[Bruce Wilder]

myself: "Risk is the consequence for welfare of not knowing what will happen."

No, not quite right. Let's try again.

Risk is the expected cost of what will happen, given that one knows that one does not know what will happen.

Somewhat better, I think.

I wonder if a realized payoff can be correctly termed, a cost, given the resolution of contingency into events also wipes out choice.

[dirtyal]

Having been in one sector of the risk business for my whole career, I recall some of the early platitudes from one of my teachers.

1) Don't risk a lot for a little,
2) Don't risk more than you can afford to lose,
3) Consider the odds.

Pretty good advice, no? If some of our financial wizards had thought in those terms, life might have been a lot more prosaic and a lot less exciting.

But when you think about it, of course, they were not risking more than they could afford to lose. They were risking more than others could afford to lose.

[Bruce Wilder]

dirtyal: "They were risking more than others could afford to lose."

I remember what my father taught me:
"In a casino, the house is not gambling."

[dirtyal]

Bruce: Right On and well said.

[cb]

Let me see if I understand this: the neo-neo-classicals who pretended models were good representations of reality and all that was left were random shocks got it wrong and didn't account for uncertainty?

And this point is being made by Ned Phelps, who won his Nobel Prize for being one of the founders of that very RBC rational expectations school of thought?

So when is he handing back the prize money?

[Lafayette]

Nobel Prize

cb: Let me see if I understand this: the neo-neo-classicals who pretended models were good representations of reality and all that was left were random shocks got it wrong and didn't account for uncertainty?

Show us an econometric model that accounts for uncertainty and I shall propose you for a Nobel Prize in the subject.