Discussion: (0 comments)
There are no comments available.
| American Enterprise Institute
View related content: Financial Services
With all of the knowledge and experience collected in financial markets and governments, why are we not able to avoid repeated disastrous mistakes like those that led to the US housing meltdown and the European sovereign debt crisis? It is because financial markets are governed by a recursive system of interconnected decisions, theories, strategies, predictions, actions, and expectations that lead to booms and busts. History shows that the complexity of these interactions makes it nearly impossible to predict accurately which way the markets will go. Regulators like the US Federal Reserve are themselves too enmeshed in the system to avoid its systemic risk. We must recognize the reality of uncertainty and the absence of any godlike regulatory guardians.
Key points in this Outlook:
Why do financial markets present us with recurring spectacles of very smart people making very big mistakes, so that something they had judged impossible or very improbable nonetheless happens, with disastrous results?
Booms and busts populate all of financial history. In 1584, a Venetian senator claimed that 103 banks had been established in the history of the city, of which 96 had failed. Carmen Reinhart and Kenneth Rogoff’s list of banking crises around the world since 1800 is forty-five pages long. They also count 250 defaults on government debt during this period. In the twentieth century’s one hundred years, banking crises started somewhere in fifty-four of them. More recently, 2,270 regulated commercial banks and thrifts failed in the United States between 1982 and 1992. During this period, more than one hundred per year failed for eight years in a row.
Considering all this experience, human learning, judgment, and ingenuity, surely we would have figured out how to prevent such unintended, unfortunate financial adventures! The regulators and legislators of two decades ago thought so. With the advice of expert regulators and academics, Congress enacted the Financial Institutions Reform, Recovery, and Enforcement Act of 1989; the Federal Deposit Insurance Improvement Act of 1991; and the GSE Safety and Soundness Act of 1992. (The last title is particularly ironic, since “GSE” refers to the now hopelessly insolvent government interventions Fannie Mae and Freddie Mac.) The then-secretary of the Treasury, Nicholas Brady, announced that all this effort meant that never again would we have such problems.
He was wrong, needless to say, and we know what happened afterward. The last fifteen years of finance have provided us with three huge bubbles and their subsequent collapses: the technology stock bubble, the US housing bubble (plus notable real estate bubbles in Great Britain, Spain, and Ireland), and the now-collapsing European government debt bubble.
Living in the wake of a bubble is no fun, and we do not even have a word in English that is the exact opposite of a bubble to convey the experience. “Boom” and “bust” go together, but “bubble” and—what? I believe a “shrivel” has the right image and emotional tone.
The continuing bubbles and shrivels do not mean that the people involved are stupid. Many are brilliant. But, then, why do they make so many and such disastrous mistakes? Some of the very smartest private and governmental financial actors contribute to the recursive interactions that create the bubbles. This includes the mathematicians or “rocket scientists” of Wall Street who designed complex structured securities and, it has been wittily said, built a financial missile that landed on themselves.
But it equally includes, among others, the government’s central bankers. Not only was the housing boom stoked by Federal Reserve actions, but transcripts released by the Federal Reserve Board make clear that they entirely missed the magnitude of the problem of the housing bubble and failed to anticipate the coming crisis. This is in spite of diligent and constant economic forecasting by scores of intelligent and well-intentioned PhD economists, armed with all the computers and databases they could desire, and earnest discussions of the outlook by senior officials.
Both the chairman of the Federal Reserve and the secretary of the Treasury proclaimed on multiple occasions in the first half of 2007, when things were already heading downhill, that the problems were “contained” and unlikely to cause wider economic damage—a really bad call! “I could kick myself. We were just plain wrong,” Treasury Secretary Hank Paulson wrote later, adding, “The crisis . . . came from an area we hadn’t expected—housing—and the damage it caused was much deeper and much longer lasting than any of us could have imagined.” (Note: not just worse than we had thought, but worse “than any of us could have imagined.”) The problem was not a lack of intelligence, but the nature of dealing with the future from within complex, recursive systems like financial markets.
Consider the nature of such systems. The following quotations are taken from a book by Robert Jervis about the “system effects” inherent in political problems, especially the diplomatic and military interactions among countries, but they lend themselves well to thinking about financial life:
[G]reater complexities are introduced with human beings whose behavior is influenced by their expectations of what others will do, who realize that others are influenced by their expectations of the actor’s likely behavior, and who have their own ideas about system effects. This is an area filled with paradoxes and self-reflective phenomena, and any discussion must be tentative and incomplete.
Further complexities are introduced when we look at the interactions that occur between strategies when actors consciously react to others and anticipate what they think others will do. Obvious examples are provided by many diplomatic and military surprises.
We can add many financial surprises, the happy booms and the painful busts, to the political surprises mentioned in this last quotation.
Jervis’s discussion highlights the difficulties of predicting, controlling, or even understanding such densely recursive or self-reflective, complex systems of human minds and interactions. He offers the following definition of a system: “We are dealing with a system when (a) a set of units or elements is interconnected so that changes in some elements or their relations produce changes in other parts of the system, and (b) the entire system exhibits properties and behaviors that are different from those of its parts.”
“Some of the very smartest private and governmental financial actors contribute to the recursive interactions that create financial bubbles.”This is all right as far as it goes, but the systems really of interest to finance are those in which the units are interacting minds, theories, strategies, predictions, actions, and expectations, all taking account of and thus changing with respect to one another. These systems often display nonlinear, indeed discontinuous, behavior and may have results not merely unintended, but previously considered impossible. This was summed up by George Soros as “the complicated relationship between thinking and reality” in financial markets.
“When the interactions are dense,” writes Jervis (as they are in finance in spades), “it may be difficult to trace the impact of any change after the fact, let alone predict
it ahead of time, making the system complex and hard to control.” Or in the less-formal language often attributed to John Kenneth Galbraith: “The purpose of economic forecasting is to make astrology look good.”
This is emphatically not because economists are not intelligent or not trying hard. It is because of something profound and elusive about complex, recursive interactions of human minds, expectations, theories, partial knowledge, strategies, and actions; the theories and forecasts of the economists, the business strategies, the regulatory rule making, and the central bank policies, all reflecting theories and forecasts, are all part of and enmeshed in the web of interactions. No one is outside the recursive system.
A related point was made by Friedrich Hayek about the results of entrepreneurial competition: you cannot know how the competition will turn out except by having it take place. “I call competition a discovery procedure,” he wrote, and “the results of a discovery procedure are in their nature unpredictable.”
Here is a clever analogy for recursive complexity from John Maynard Keynes:
Professional investment may be likened to those newspaper competitions in which the competitors have to pick out the six prettiest faces from a hundred photographs, the prize being awarded to the competitor whose choice most nearly corresponds to the average preferences of the competitors as a whole; so that each competitor has to pick, not those faces which he himself finds prettiest, but those which he thinks likeliest to catch the fancy of the competitors, all of whom are looking at the problem from the same point of view!
So everybody’s behavior is based on predicting one another’s predictions of one another’s behavior. (Keynes knew what he was talking about, having been himself an active financial speculator who personally experienced both making a lot of money and nearly going broke.)
This is an insightful analogy, but to make it really reflect financial markets we have to add some further complications. The state of the current “voting” (analogous to the prices in a market) is constantly reported, and the competitors are correspondingly changing their votes. Statistical models are built to predict the outcomes and the probabilities of their variation, and large numbers of the competitors use these models, which changes the way they vote. Government agencies are issuing rules about how the competitors can vote, including regulatory quotas on the photos chosen. Meanwhile, the central bank is continually printing new photos to add to the competition. A winner is declared at the end of each day, but the game resumes immediately and never ends.
These problems of system effects not only characterize economic and financial forecasting, but also interpretation of economic history. The causality may be hard to figure out even after the fact. The future is unknowable, we are confused by the present, and we misinterpret the past, I was once told. Theories of what caused the housing bubble of 1999–2006 continue to be hotly debated—and so do theories about the Great Depression of eighty years ago.
“No one is outside the recursive system.”
There is an illusion about history, as Jervis says: “Looking back, actors and observers may see the results as inevitable.” They were not, and “the course of events [is] more difficult to control than people believe.” Nowhere is this more true than in financial markets. Jervis’s discussion of complex systems thus rebuts naïve faith in financial regulation—“regulation did not work last time, but if we just regulate more, it will work this time”—so often displayed by academics, politicians, and especially journalists.
Finance and Astronomy
How different is this view from the unfulfilled dream of economics and finance as Newtonian science. The divide is nicely highlighted in the financial experience of the illustrious master and father of predictive mathematical science, Isaac Newton himself.
Newton may have been the greatest mind in history—in any case, his genius is beyond doubt. How did he do when faced with the bubble of his day, the South Sea Bubble of 1720?
Newton was an early investor, or speculator, in the shares of the South Sea Company. He doubled his money and sold out with a 100 percent profit. Very good performance indeed, but the market kept going up. He experienced seller’s remorse, bought back in, and lost a great deal of money when the bubble shriveled. He famously wrote in disgust, “I can calculate the motions of the heavenly bodies, but not the madness of people.”
Using Newton’s mathematical laws, astronomers can calculate with accuracy the paths of the planets for hundreds of years into the future. They can calculate the paths of rockets through and out of the solar system. All experts will agree with one another on these predictions.
But how about calculating the interactions of people? Obviously, three centuries years later, many very intelligent professionals have not done any better at this than Newton, and not for lack of trying. Unlike the astronomers, economists have forecasts that always disagree and engage in arguments about theory and interpretation that are never settled but become part of the financial market’s recursive interactions.
The recursiveness gives rise to Goodhart’s Law, that when you try to use the statistical relationships that econometrics has discovered to control financial behavior, your actions in doing so change the behavior and the relationships. As stated by banking expert Charles Goodhart in 1975, “Any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes.”
Here is a more recent application of this law: “A risk model breaks down when used for regulatory purposes.” Another related version of the problem: “Any change in policy will systematicallyalter the structure of econometric models.” The models helped form the policy, so the policy is part of the forecasts and expectations and must forecast the forecasts and expectations it will affect. We are back to Keynes’s photograph competition and far from the motions of the heavenly bodies.
The contrast between the mathematical models of physics and those of finance has been nicely articulated by Emanuel Derman, a physicist who also worked for years on Wall Street: “Newton’s laws and Maxwell’s equations are facts of nature,” he writes. “In physics one wants to predict or control the future. In finance, one wants to determine present value [and] turn those opinions about the future into an estimate of the appropriate price to pay today for a security that will be exposed to that imagined future.” So what are you trying to find, a fact or a price?
However, Jervis does issue a warning even about astronomy. “Henri Poincaré showed that the motion of as few as three bodies (such as the sun, the moon, and the earth),” he writes, “although governed by exact scientific laws, defies exact solution: While eclipses of the moon can be predicted thousands of years in advance, they cannot be predicted millions of years ahead.”
Only thousands of years in advance! How far ahead can financial events be accurately predicted? Not a thousand days, not a year—not even three months. The motions of heavenly bodies and the bubbles and shrivels of financial markets are two different types of reality.
A major economic and financial policy of the American government that ended up having the opposite of its intended effects was the promotion, sponsoring, and guaranteeing of Fannie Mae and Freddie Mac. Meant to improve the mortgage finance market, and making it at one point, it was claimed, “the envy of the world,” Fannie and Freddie acquired vast numbers of bad loans and went completely broke. Having made in the thirty-five years from 1971 to
2006 total profits between them of $106 billion, in the five years from 2007 to 2011, they lost $256 billion. Quite a performance. Because the government was and is guaranteeing every creditor of Fannie and Freddie, they will end up costing the taxpayers $150 billion or more.
Consider some forecasts concerning this memorable financial disaster by highly educated, well-informed, and senior voices:
“The risk to the government from a potential default on [Fannie and Freddie’s] debt is effectively zero.” A very bad call by distinguished economists Joseph Stiglitz and Peter Orszag in 2002, while the housing bubble was already inflating.
“Let me be clear—[Fannie and Freddie] have prudent cushions above the OFHEO-directed capital requirements.” Their chief regulator, James Lockhart, the director of the Office of Federal Housing Enterprise Oversight, six months before both the GSEs collapsed.
“What’s important are facts—and the facts are that Fannie and Freddie are in sound situation.” The chairman of the Senate Banking Committee, Christopher Dodd, two months before they collapsed.
“We have no plans to insert money into either of those two institutions.” Paulson, one month before they collapsed. The US Treasury has since inserted more than $180 billion into those two institutions.
Fannie and Freddie were an important part of the overall “systemic risk,” as we now call it, of the complex financial system. Arising from recursive financial interactions and expectations, systemic risk is the possibility that the whole system may bust or shrivel. Most of the time, the probability of this will be viewed as extremely small and remote, a so-called “tail risk,” but when a bubble has become fully inflated, the subsequent fall has become unavoidable—the tail risk is virtually 100 percent. Tail risk is not a fact of nature, but is created by financial interactions.
Regulators, central bankers, academics, and financial companies around the world are trying to understand, develop approaches to, and cope with present and future systemic risk. This is difficult to do, although it is easy to appoint a committee and tell its members they are supposed to figure it out.
“The essential problem is that the unknowable human future is worse than risky: it is uncertain.”The most feared result of systemic risk is a financial panic, when maturing debt, especially short-term debt that needs to be constantly rolled over, cannot be renewed, and when everybody wants their money back at once. Of course, it is impossible for everybody to get their money back at the same time. In its best-known form, this is a run on a bank.
“On extraordinary occasions, a general panic may seize the country . . . against such panics banks have no security on any system,” David Ricardo wrote two centuries ago, and he is still right. In the course of the most recent financial crisis, a cartoon showed two Tinker Bell figures with magic wands, one saying to the other, “If you don’t believe in banks, they die.” This reminds us that “credit” comes from “credo” (“I believe”). The relevant belief in this case is whether other people do not believe in the banks and therefore will panic, in turn affecting others’ beliefs.
This shows us why governments constantly harp on the need to make the people “confident.” But when the government pleads for you to be confident, does that increase your confidence?
Systemic financial risk turning into a run or bust or shrivel is always unintended. But how are we doing at defining “systemic risk”? National Mortgage News reported that “The chief obstacle to heading off systemic risk turns out to be agreeing on a definition for it, including regulators both here and abroad and financial firms . . . what’s needed is a public-private partnership between regulators and industry to aid the process of defining specific standards.”
This looks like a fine example of Goodhart’s Law in process. Can we successfully define “systemic risk” and the factors which create or increase it so that it can be controlled? US Supreme Court Justice Potter Stewart’s famous quote is that he could not define obscenity, but he knew it when he saw it. With systemic risk, however, we cannot define it and we do not know it when we see it—only afterward, when it is too late.
Interest Rates and Systemic Risk
Interest rates are the price put on the future by financial markets, often with heavy interventions by central banks. “In the charts and tables of interest rates over long periods,” wrote the leading historian of interest rates, Sidney Homer, “students of history may see mirrored the rise and fall of nations and civilizations, the exertions and tragedies of war, and the enjoyments and abuses of peace . . . the fluctuations [in] the progress of knowledge and technology, the successes and failures of political forms.” This last point is now being pointedly demonstrated by the interest rates on Greek government debt.
Interest rates over time, as system effects, display surprising behavior—indeed, behavior that previous market participants considered simply impossible. Why? This history is a lesson in how little anyone knows of the complex financial future.
In March 1950, interest rates on ten-year US Treasury notes were 2.25 percent, about the same level as in March 2012. They have since last March gone even lower, although most professionals a year ago thought that they had to go up. As one thoughtful investment manager wrote in June 2012, “This remarkable bond market continues to surprise . . . there is no question we are in extraordinary and bizarre times.”26 Six decades ago and now, the Federal Reserve was and is manipulating bond prices to keep long-term interest rates low. In 1950, this manipulation had been going on since the Second World War. How long can it go on now? Guessing interest rate movements turns out to involve guessing the
theories, strategies, beliefs, and expectations of the Federal Reserve, and a lot of effort is spent doing so.
How low or high can interest rates go? Ten-year Treasury rates peaked at over 15 percent in 1981—a level unbelievable in 1950 and again so now. The early-1980s interest rates were the death knell of the savings and loan industry and the postwar American mortgage finance system.
An old colleague of mine worked on a strategic planning effort for a great American bank during the early 1960s. Interest rates had been rising, so they asked themselves, “What is the highest that US interest rates could possibly go?” The answer of these professional bankers: 6 percent.
The ability of highly educated, trained, and experienced professionals to anticipate what interest rates will do is distinctly limited. “They had believed in 1946 that 21⁄2% was a fair rate of interest,” James Grant writes in his sardonic style. “They had entertained similar delusions about 31⁄2% in 1956, about 41⁄2% in 1959, and about 51⁄2% in 1966. When successive Presidents, Treasury Secretaries and Federal Reserve Board Chairmen had promised balanced budgets, lower interest rates and sound money, they had believed them.” Because bond prices fall when interest rates rise, “It was one of the longest losing streaks in the annals of investments.”
As the Federal Reserve ran the monetary printing presses during the 1970s, interest rates trended higher. Finally came the 1981 peak, with the Federal Reserve, under chairman Paul Volcker, famously “breaking the back of inflation”—the inflation it had itself created. Then began the great three-decade-and-still-going bull market in bonds. For the entire adult memory of people now in their thirties and forties, interest rates have been, on average, falling. What are the system effects of that experience?
Regulators in the wake of the twenty-first century crisis are busy designating large financial firms as SIFIs—“systemically important financial institutions”—or those that can create systemic risk. What the history of interest rates over the last six decades makes indubitable is that the biggest SIFI of them all is the Federal Reserve itself. Then-senator Jim Bunning is said to have asked Federal Reserve Chairman Ben Bernanke, “How can you regulate systemic risk when you are the systemic risk?”: a great and unanswered question.
It reflects the fact that neither the Federal Reserve nor any regulator nor any government is in a godlike position above the system of financial interactions, looking down and able to understand and decree from on high. Rather, they are all enmeshed within the system of recursive interactions. This means there will always be mistakes.
Uncertainty and Mistakes
Europe’s banks and entire monetary system are in an extended and agonizing crisis from the sovereign debt of financially weak governments. But the capital requirement for banks to hold such debt was zero. Being erroneously considered “risk free” caused the risk of this debt to increase, helped push it into a bubble, and led to massive and painful ongoing losses.
What an amazing set of blunders, it now seems, both by those who bought the debt and who wrote the capital requirement—especially given the blatant historical record of defaults by governments on their debt! But the actors who made what are so obviously blunders viewed after the fact, were intelligent and well educated, not stupid, and surely in most cases, well intentioned.
Considering financial crises, Jeffrey Friedman and Wladimir Kraus emphasize the “cognitive limitations” of regulators, central bankers, and private financial actors. In other words, they stress mistakes. “Honest mistakes . . . are inherent in the human condition,” they write. Very true. I have long thought that the question of incentives is overemphasized, while the question of knowledge versus ignorance is greatly underappreciated, by economists.
The essential problem is that the unknowable human future is worse than risky: it is uncertain. It is uncertain precisely because of recursive system effects. As Frank Knight pointed out long ago, uncertainty means not only that you do not know what will happen in the future; it also means you do not even know the odds and, moreover, that you cannot know the odds.
Of the kinds of problems created by system effects, Jervis rightly says, “Problems are almost never solved once and for all.” This applies well to systems of financial interactions. A very costly example of system effects and systemic risk is the savings and loan industry collapse in the 1980s. The collapse notably included the insolvency and failure of the government’s savings and loan deposit insurance fund. The result was a $150 billion bailout at the expense of the taxpayers, partially financed by the sale of noncallable forty-year bonds yielding 9 percent: the taxpayers will be paying on those bonds for eighteen more years, until 2030.
In an insightful essay, Arnold Kling discusses how in the 1990s the regulators carefully, diligently, and reasonably studied the savings and loan disaster to learn from it. They drew three principal lessons: You had to expand the securitization of mortgages. You had to have mark-to-market accounting. And you had to have risk-based capital requirements. These were plausible, though arguable, conclusions. All three lessons were applied and implemented. And all contributed significantly to the inflation of the housing bubble and the subsequent crisis of the 2000s. “Each era of regulation seems to contribute to the next era of euphoria,” Kling concludes. The buildup of systemic risk in the new situation from applying the lessons of the previous crisis was not apparent to most observers.
I am often asked, “Will we learn the lessons of the financial crisis?” “Yes, we will,” I answer. We learn the lessons after every crisis. But it does not stop the next one from happening.
In complex, densely recursive systems, uncertainty is an unavoidable fact of life. Yet those enmeshed in such systems still have to make decisions and act. Mistakes inevitably follow and will continue to do so.
Alex J. Pollock ([email protected]) is a resident fellow at AEI.
1. Bernard Shull and Gerald A. Hanweck, Bank Mergers in a Deregulated Environment (Westport, CT: Quorum Books, 2001), 19.
2. Carmen Reinhart and Kenneth Rogoff, This Time Is Different: Eight Centuries of Financial Folly (Princeton, NJ: Princeton University Press, 2009), 348–92.
3. Ibid., 111.
4. Alex J. Pollock, “There’s Usually a Banking Crisis Somewhere!” The American, September 21, 2011, www.american.com/archive/2011/september/theres-usually-a-banking-crisis-somewhere (accessed July 13, 2012).
5. Federal Deposit Insurance Corporation, “Failures and Assistance Transactions, 1934-2011,” n.d., www2.fdic.gov/hsob/HSOBSummaryRpt.asp?begYear=1934&endYear=2011 (accessed July 18, 2012).
6. Henry M. Paulson Jr., On the Brink: Inside the Race to Stop the Collapse of the Global Financial System, (New York: Business Plus/Grand Central Publishing, 2010), 66, 61.
7. Robert Jervis, System Effects: Complexity in Political and Social Life (Princeton, NJ: Princeton University Press, 1997), 253.
8. Ibid., 44.
9. Jervis, System Effects, 6.
10. George Soros, “Do Not Ignore the Need for Financial Reform,” Financial Times, October 25, 2009.
11. Jervis, System Effects, 17.
12. Friedrich August Hayek, “Competition as a Discovery Procedure,” in New Studies in Philosophy, Politics, Economics and the History of Ideas (Chicago: University of Chicago Press 1978), 182, 184.
13. John Maynard Keynes, quoted in Jervis, System Effects, 253.
14. Jervis, System Effects, 165.
15. Sir Isaac Newton, as quoted in C. P. Kindleberger, Manias, Panics, and Crashes, 3rd ed. (New York: John Wiley, 1996).
16. C. A. E. Goodhart, “Problems in Monetary Management: The U.K. Experience,” in Inflation, Depression, and Economic Policy in the West: Lessons from the 1970s, ed. Anthony S. Courakis (London: Mansell and Alexandrine Press, 1981).
17. Jon Danielsson, “The Emperor Has No Clothes: Limits to Risk Modeling,” Journal of Banking and Finance 26, no. 7 (2002): 1273–96.
18. Robert Lucas, “Econometric Policy Evaluation: A Critique,” 1976.
19. Emanuel Derman, Models.Behaving.Badly. (New York: Free Press, 2011), 193.
20. Jervis, System Effects, 6–7.
21. “Systemic Risk and Fannie Mae,” Wall Street Journal, December 1, 2009.
22. Three preceding statements quoted in Alex J. Pollock, Boom and Bust (Washington, DC: AEI Press, 2011), 50–51.
23. Quoted in ibid., 27–28.
24. Donna Borak, “Can Regulators Prevent Next Systemic Risk Crisis?,” National Mortgage News, December 19, 2011.
25. Sidney Homer and Richard Sylla, A History of Interest Rates (Rutgers, NJ: Rutgers University Press, 1991), 1.
26. David Kotok, “The ‘Good Guys’ and the ‘Bad Guys,’” Market Commentary (Cumberland Advisors), June 6, 2012, www.cumber.com/commentary.aspx?file=060612.asp (accessed July 18, 2012).
27. James Grant, Money of the Mind (New York: Farrar, Straus, and Giroux, 1992), 287.
28. Alex J. Pollock, “Yet Another Sovereign Debt Crisis,” AEI Financial Services Outlook (March 2012), www.aei.org/outlook/ economics/financial-services/banking/yet-another-sovereign-debt-crisis/.
29. Jeffrey Friedman and Wladimir Kraus, Engineering the Financial Crisis (Philadelphia: University of Pennsylvania Press, 2011), 153.
30. Jervis, System Effects, 294.
31. Arnold Kling, Not What They Had in Mind: A History of Policies That Produced the Financial Crisis of 2008, (Arlington, VA: Mercatus Center, 2009), http://mercatus.org/publication/not-what-they-had-mind-history-policies-produced-financial-crisis-2008 (accessed July 13, 2012).
There are no comments available.
1150 17th Street, N.W. Washington, D.C. 20036
© 2014 American Enterprise Institute for Public Policy Research