Arnold Kling  

Memo to Krugman and DeLong: Start a Little Differently

Warren Buffett, Five... The Other Demographic Suicide...

Paul Krugman writes,

Brad DeLong offers a neat little model of speculative fluctuations in asset prices, based on the idea that investors gradually switch strategies based on what seems to work for other people: if people buying stocks seem to be doing well, more people move into stocks, driving up prices and making stocks look even more attractive...

What's missing, as Brad himself points out, is the asymmetry of booms and crashes. What he doesn't say is that what we really need is a model that can produce a Minsky moment -- the point at which margin calls force deleveraging.

Separately, Krugman writes,

I have to say that the Platonic ideal of Minsky is a lot better than the reality.

Indeed. As Robert Solow once said of Schumpeter, he was all problems with one big idea. Minsky's actual macro analysis consists of holding everything in the national income accounting identity constant except for the budget deficit and corporate profits, thereby making them the mirror image of one another. Not promising.

Where I would suggest that DeLong and Krugman start differently is to focus on financial intermediation. Investor beliefs can differ, but I think the main issue is what they believe about financial intermediaries, which I think can be described in terms of the signals that intermediaries send to induce people to invest.

My story is this:

1. Ordinary folks in the nonfinancial sector want to issue risky liabilities (shares of investments in fruit trees, mortgages on houses) and wants to hold riskless assets (demand deposits, money market fund shares).

2. The financial sector obliges by having a balance sheet with risky assets and supposedly riskless liabilities. The bigger the financial sector gets, the more euphoric the investment climate, because nonfinancial folks get to hold more riskless assets and issue more risky liabilities.

3. The financial sector's expansion is based in part on signaling. That is, banks signal that their liabilities are low risk. Signals include fancy buildings, the "FDIC insured" sticker, balance sheets filled with AAA-rated assets, and so on.

4. Financial expansions are gradual, because it takes a while to come up with new signaling mechanisms and to get the credibility of those mechanisms established with investors.

5. Financial crashes are sudden, because once investors lose a little bit of their confidence in financial institutions, their natural instinct is to ask for safer assets (they withdraw money from uninsured banks, or they ask AIG to post collateral). This behavior weakens the financial institutions further, leading to a rapid downward spiral. Today, we are seeing that all sorts of signals are discredited.

So, my advice to Paul and Brad is this: don't start with a model that focuses on investor beliefs about real economic variables. Instead, start with a model in which financial firms use signaling to expand, and the credibility of those signals increases over time as long as nothing adverse happens. It should be easy to develop a model in which signaling devices gain credibility slowly but lose credibility suddenly. That will (a) produce the asymmetry between euphorias and crashes and (b) tell a story that puts the fragility of the financial sector in the middle, where it belongs.

Incidentally, one thing I got from reading Galbraith's book on financial euphoria is that you can fit the late 1920's into this model. Holding companies bought stocks, and the holding companies were bought by other holding companies, and so on. You get the spectacular leverage, and on the way up people think that the assets of the holding companies are pretty low risk. Then when people get a little nervous...

Comments and Sharing

COMMENTS (14 to date)
Lord writes:

Sort of neglects the real side though how relaxed lending leads to higher asset values that lead to even more relaxed lending, not that this is easy to accomplish without signaling, but that there are real distortions to asset values induced by it and increases in debt due to it and that there are real values underlying them dependent on long term interest rates and money supply as the case may be. Were they signaling investors, each other, or the central bank? Was the acquiescence of the central bank a signal to relax further? Was it the confidence in the central bank's willingness to bail them out that really mattered to them?

Greg Ransom writes:

Arnold, ind a way to put this into math and you'll have yourself a Nobel Prize -- especially if the math is just difficult enough than anyone who didn't take undergraduate math class would find it impossible to understand.

Greg Ransom writes:

Arnold, let add the microeconomics of heterogeneous and time-variable production goods and heterogeneous labor, and some sound money and interest theory, and you've got yourself a powerful explanation of the boom and bust cycle.

Gary Rogers writes:

I like it. Stress tests might be an example of a recent signal. It certianly adds more to economic understanding than a lot of models I have seen. The next question is whether it just explains the economics, or can it become a predictive model that might warn of growing bubbles?

Eric writes:

I prefer my pet theory, which can account for the asymmetry in a slightly different manner (slightly Ponzi-style):

The population of people willing/able to participate in any given boom is limited by the total population, the population of people with cash on hand/ability to borrow to trade, etc.

As buyers enter the market, liquidity increases and price increases accelerate. Earlier investors sell back into the liquid market, and only some of them buy back in. As the group with "dry powder" (and the opinion that price increases will continue) dwindles, price increases slow.

As the price increases slow, the most leveraged investors are forced to sell or try to sell, because their cost of capital becomes higher than their return. Prices aren't increasing anymore, so it's hard to attract new investors, and...

Suddenly the investment is illiquid and prices plummet.

Is it a predictive model? No. If it was I would be rich. You might be able to get a sense of it in "real" asset markets (like houses) simply by counting transactions.

Maximum Liberty writes:


I think Arnold's concept complements the idea that the euphoria is based on credit injection that leads to malinvestment. The firms that signal best will get the biggest benefit from credit injection. When the malinvestments caused by credit injection start becoming apparent, that triggers the doubt about the validity of the signals being sent. The firms that epitomize the signaling of the euphoric phase would then suffer the most to the extent that the signals they sent fail to correlate with strong investments.

Arnold, would you agree?


Bill Woolsey writes:

The problem with Kling's theory is that it is too "financial." Where does the nonfinancial sector get the money to buy the low risk assets? Where is the rest of the economy?

Implicit in all of Kling's theorizing is that the demand for money changes. Money is the hidden variable in these "models."

If it isn't the demand for money, then it is saving--that is, less current consumption. Households are inclined to spend their income on consumer goods because saving is too risky. They are willing to lend to financial intermediaries who issue low risk liabilities to finance the fundamentally risky real projects.

Under this scenario, the cycles should involve rising consumer demand during downturns, as people lose faith in intermediaries and dissave. The reduced investment reduces the production of consumer goods in the future.

Fundamentally, recessions should involve shortages of consumer goods (and surpluses of capital goods.)

If, on the other hand, it is a "demand for money" story, then when people switch from the liabilities of financial intermediaries, there is an incease in the demand for base money. We are in the usual place where this requires an incease in the real quantity of base money through a general deflation of prices and incomes or else an increase in the nominal quantity of base money. Monetary disequilibrium.

As an aside, households do invest in equities, corporate bonds, and the like. So, it is more like firms are willing to issue more risky securities than households are willing to directly hold, and financial intermediaries issue safe liabilities, and so incease the amount of risky assets that can be fiananced in total--both directly and indirectly.

And, by the way, the liabilities of the financial intermediaries are less risky because some of the risk is shifted to their stockholders. And, for some investors, there really are gains from additional diversification. And people do hold the liabilities of financial intermediaries as a convenient way to make payments rather than as a "safe" investmnet.

So, people lose trust in the financial intermediaries and hold base money causing a general glut of goods. The market solution is an increase in the real quantity of base money through a general deflation. This is horribly painful, and as a fly in the ointment, it hurts all debtors, including financial intermediaries.

What is new?

In my view, fixing the financial system so that there will never be a loss of faith and an incease in the demand for money (or, after the fact, shoring up faith in the financial system to reduce the demand for money) is a fools errand. Figuring out a way to increase the quantity of base money enough to meet whatever amount is demanded at a target growth path for nominal income is the way to go. And then, if there really are cycles where people demand more consumer goods because they are too fearful to invest in capital goods--directly or indirectly--then consumer prices will rise and real incomes fall during these "recessions" all in the context of growing nominal incomes and unemployment limited to shifts between consumer and capital goods industries.

Mike S writes:



"Where does the nonfinancial sector get the money to buy the low risk assets? Where is the rest of the economy?

Implicit in all of Kling's theorizing is that the demand for money changes. Money is the hidden variable in these "models."

If it isn't the demand for money, then it is saving--that is, less current consumption. Households are inclined to spend their income on consumer goods because saving is too risky. They are willing to lend to financial intermediaries who issue low risk liabilities to finance the fundamentally risky real projects."

is backwards. Savings are related to the the demand for currency. Savings are the avoidance of consumption. If demand for currency becomes too high, then prices go down causing simply holding currency to pay implied interest, which then spurs further hording of currency as it increases in value relative to goods. Like for example, in todays real estate market where 200,000 buys more house than it did 3 years ago.

Arnold misses this point in his proposed model. Cycles are not exclusively the result of signaling. They are directly related to the relationship between desired savings and the availability of savings in addition to what Arnold outlines above. 1/2 of Minsky's macro is spot on, because the government deficit is funding something very important. Because HM misidentified the object of funding (as corporate profits) instead of private sector savings as is proper, Kling throws the entire money creation process out the window as unimportant. I will say this explicitly: govt. deficit spending = private sector savings. Avoiding this accounting identity is a mistake of monumental proportions, as any macro model that does not take into account money creation and non-financial activity of any sort is doomed to be incomplete and flawed.

That said I agree with the above in that if
Arnold could make this model work, it is Nobel prize quality work. However, so is Brads work, if he can find a rational mechanism that explains and works with quick drops within his model.

fundamentalist writes:

If you're looking for a strictly psychological explanation of cycles, Arnold's isn't bad. It needs an explanation of how the money supply grows to meet the increasing demand for investment. Also, research has shown that fear drives people more than any other emotion, and people fear losing money about ten times more than they enjoy gaining the same amount. That's why people are slow to invest and quick to cash out.

mike writes:

I think Arnold's model needs a little more justification starting from point (1). Here he seems to be assuming that savings originate from ordinary folks who always want to buy safe assets. The problem is that this implies (a) people are too stupid to invest in high return riskier assets that, over the long run, more than compensate for the extra risk, even if one is insanely risk averse; (b) that the typical savings dollar comes from an 'ordinary' person likely to be this stupid.

There's a lot of stupidity in the universe. The second most common element, after hydrogen, according to Harlan Ellison. Even so, I think you need a little more than a blind assertion regarding this level of stupidity. Consider that the typical dollar of savings come from an "ordinary" wealthy and highly educated household.

You need to flesh out how stupid expectations evolve in a way that makes them plausible. And I think that is what Krugman/Delong are doing.

Michael writes:

I'm somewhat with Mike S here.

Brushing off Minsky leaves me wanting more discussion about this relationship between deficits and net financial savings in the private sector.

The role of the vertical component of money creation seems crucial here. What are we missing?

gnat writes:

Doesn’t financial intermediation work like an actuarial insurance: Activities that payoff have to offset the costs of activities that fail. If everyone/anyone can intermediate then it is possible that the returns on activities that payoff may be eroded and losses cannot be offset. If you the investor expect other risk adverse investors will sell, you had better sell.

gnat writes:

Let me add:
the trigger is when diversifiable risk is changed to a game theory-type (systematic risk) payoff. How does this happen? It can be for a number of reasons: moral hazzard emerges, adverse selection, asymetric information and gaming, or market concentration. Game theory payoff reflects risk interdependance: your payoff depends upon what you expect others to do.

How is it possible to determine legitimate market signals when government is meddling with market risk?

Comments for this entry have been closed
Return to top