8: Debt and Bubbles
Charles P. Kindleberger was a giant in economics. He joined the economics department at MIT in 1948, but not as a freshly minted PhD. Instead, his academic career was preceded by close connections to policy making. He had served as a major in the U.S. Army, worked as an economist in the Federal Reserve System, and was a leading architect of the Marshall Plan in the State Department after World War II. So by the time he arrived at MIT, he had already made a major impact on western European economies. Kindleberger’s research style was a bit unusual compared to many of his contemporaries, and this likely reflected his real-world experience outside the ivory tower. Instead of proposing theory, he approached economic phenomena as a natural scientist would. His colleague and Nobel laureate Robert Solow compared Kindleberger to Darwin on the Beagle: “collecting, examining, and classifying interesting specimens . . . it was Kindleberger’s style as an economic historian to hunt for interesting things to learn, not pursue a systematic agenda.” The culmination of Kindleberger’s massive data collection on bubbles—Manias, Panics, and Crashes: A History of Financial Crises—is one of the most influential books written in economic history. The book is a tour de force: it covers bubbles going back to the tulip mania in the seventeenth-century Netherlands to the commercial real estate boom before Japan’s “Lost Decade,” to the 1998 financial crisis spurred by the collapse of Long-Term Capital Management. It represents one of the most systematic and large-scale explorations of bubbles and financial crises ever written.
The Science of Bubbles
Even though Kindleberger didn’t set out to prove any one theory, his close examination of the historical data led him to strong conclusions. First, he noticed that the main driver of asset-price bubbles was almost always an expansion in credit supply, that is, an increased willingness by creditors to lend to borrowers with no discernible improvement in income growth. What does an expansion in credit supply look like in the context of housing? Imagine a renter walking into a bank and asking for a mortgage to purchase a new home. Normally, the banker will immediately inquire about the renter’s income. If the banker deems his income too low to support a large mortgage, she will restrict the size of his mortgage to some fraction of the overall value of the home. In many cases, the restriction will prevent the renter from buying the home he had in mind. Now imagine that for some reason completely unrelated to the renter’s income level, the bank decides to give him a much larger mortgage at an even cheaper initial interest rate than what he would normally receive. For the exact same income, the bank is suddenly willing to provide more credit. This will likely affect the renter’s demand for housing; he may, for example, choose to buy an even bigger home. If this happens on a wide scale, the increased willingness of lenders to provide credit inflates house prices. For the same level of risk, bankers are willing to supply more credit, and this leads to house-price growth.Kindleberger noticed this strong pattern in many episodes, so much so that he established the axiom that “asset price bubbles depend on the growth of credit.” He gave numerous examples. The tulip bubble in the seventeenth-century Netherlands was sparked by a form of vendor financing, which is debt between buyers and sellers of tulips. The canal mania in eighteenth-century Great Britain was fed by loans from newly established country banks to the entrepreneurs developing the canals. This is what was happening in Detroit. A vast expansion of mortgage credit to borrowers otherwise unable to buy a house fueled an enormous house-price bubble in neighborhoods with many such borrowers. House prices in these areas of Detroit rose by 80 percent in the decade before the housing crash. When it popped, prices collapsed by 60 percent.
What Is a Bubble?
What should be the price of an asset such as a stock or a house? Standard asset pricing theory suggests that it should equal the sum of expected payoffs from the asset. For a stock, this is simply the expected sum of dividends one receives from holding the stock, appropriately discounted for time and risk. For a house, it is the same calculation with implicit rental income, or the income one could earn from renting the house.Do bubbles exist? There have been many instances of rapid price growth—like the 2002–2006 housing boom or the 1997–2000 Internet boom. These episodes ended in spectacular busts, and it is tempting to call them bubbles after the fact. But what if the price booms were legitimate and based on economic prospects at the time? How can one prove the existence of bubbles without a doubt?
In 1988 future Nobel laureate Vernon Smith and his coauthors, Gerry Suchanek and Arlington Williams, published a seminal paper on the existence of bubbles. The authors conducted an experiment where participants were each given an initial allotment of cash and stocks that they could trade with one another. The experiment had fifteen trading periods. At the end of each trading period, the owner of a stock received a dividend payment that could have one of four values with equal probability—0, 8, 28, and 60—for an expected value of 24 cents. Standard asset pricing theory provides an exact value for the price of a stock in this example. At any point in time, it should equal the expected future dividends from the stock. Therefore, the stock price at the beginning of period 1 should be 24 × 15 = $3.60, and it should decline by 24 cents in every subsequent round. Smith and his coauthors made every one of their participants aware of this calculation to make sure they understood the security they were trading.
The environment in Smith’s experiment was a caricature of the true world. It had none of its complexities and uncertainties. There was no uncertainty about when the stock would stop paying dividends, or about the maximum or average payment that one could receive from the stock. There was no political uncertainty, nor any concern about mismanagement of the stock’s cash flows. If there were ever a market where the stock price should equal the present value of expected payments, it was Smith’s lab experiment. Yet the authors found something remarkable—an outcome that has been repeated many times by various researchers since. Stock prices in Smith’s experiment fluctuated wildly, with prices at times deviating two to three times away from their fundamental value. Of the twenty-two experiments conducted, fourteen saw a stock market “characterized by a price bubble measured relative to dividend value.”
The results bore an uncanny resemblance to the “excess volatility” phenomena first documented by Robert Shiller in 1981 for the U.S. stock market. In his seminal paper that led to the creation of the field of behavioral finance, Shiller showed that stock prices moved too much to be justified by the subsequent movement in their dividends. This phenomenon was later succinctly summarized by Jeffrey Pontiff in 1997 when he demonstrated that closed-end mutual funds were significantly more volatile than the market value of the underlying securities. Closed-end mutual funds hold stocks and bonds like regular “open-ended” mutual funds. But the key difference is that closed-end mutual funds are traded separately—independent of the underlying securities—and have their own independent price. Theoretically, the price of a closed-end mutual fund should mimic the total value of its underlying securities. But Pontiff found that this was frequently not the case. Prices of closed-end funds deviated from the value of the underlying securities.
All of this suggests that bubbles do exist and that they can make prices deviate substantially from their long-run fundamental value. But our question is more specific: Is there a connection between debt and bubbles? Why are real-world examples of bubbles often accompanied by a run-up in debt, as Kindleberger so comprehensively demonstrated? The idea that the price of an asset should equal the total revenue one expects to receive from it is intuitive and straightforward. Debt plays no role in this calculation. But if the people buying the assets are borrowing money to finance their purchases, as in the episodes that Kindleberger uncovered, is there any reason that the price of a stock or a house should be higher? The price of an asset should depend only on the return one expects from holding the asset, regardless of how the buyer finances the asset purchase. The Kindleberger insight on the importance of debt in bubbles is difficult to reconcile with standard asset pricing theory.
If we want to introduce a role for debt in determining asset prices, we must move away from standard asset pricing theory. We need to consider a world where prices may periodically deviate away from the sum total of their future dividend stream—a world in which bubbles may exist. Perhaps in such a world, debt matters. This brings us back to Vernon Smith. Smith augmented his initial experiment by allowing his lab traders to buy the stock on margin, meaning they could borrow money to purchase stocks. The ability to borrow money should have no impact on the price of an asset under standard asset pricing theory, but Smith and his coauthor David Porter found that the availability of debt indeed made bubbles even larger.
Why Does Debt Fuel Bubbles?
Traders buy and sell assets to make money. If buyers know they are buying into a bubble that is about to burst, they won’t buy the asset. And if there are no buyers for the asset, then the bubble would not exist. Logic dictates that a bubble can only exist if the buyers are “optimists” (a gentle word for those with “irrational exuberance”) or if the buyers believe there will be a “greater fool” to buy the asset in the future when prices are even higher.We can now start to build a theory of how debt stokes bubbles. John Geanakoplos has investigated how debt enhances the buying capacity of optimists, or those who are convinced that asset prices will continue to rise. By enhancing optimists’ buying power in the future, debt increases the probability that a greater fool will indeed be waiting tomorrow.
Imagine a world with 100 identical houses for sale. Two types of people populate this world: optimists and pessimists. Pessimists believe that a house is only worth $100,000. Optimists, on the other hand, believe that the value of a house is 25 percent higher, at $125,000. So optimists are willing to buy a house for any price that is equal to or below this amount. This simple model of the world assumes “differences in beliefs” about asset prices, which, as anyone who has discussed house prices with a friend or family member knows, is a pretty realistic assumption.
So what will the actual price of a house in this world be? It depends on the number of optimists versus pessimists. If there are enough optimists to buy 100 homes, then the sale price of all homes will be $125,000. But if there aren’t enough optimists and some houses must be bought by pessimists, then all houses must sell for $100,000. The reason for this is that competition implies that all identical houses must sell for the same price. As a result, the market price is equal to the lowest price that clears the market, the price that guarantees there will be at least 100 buyers.Suppose our world begins with no debt. Optimists have to pay cash to buy a house. Moreover, let’s say the total wealth of all optimists combined is limited to $2.5 million. As a result, they can buy no more than 20 houses at $125,000 a piece. The optimists can’t buy all the houses, and the price of houses must therefore drop to $100,000 to attract pessimists to buy. Without debt, the price of all houses is $100,000, and optimists buy 25 houses while pessimists buy the remaining 75 houses.
How does the introduction of debt financing affect the price of houses? Suppose we now allow optimists to borrow 80 percent of the value of a house. In other words, for any home purchase, they only need to put 20 percent down in cash before getting the loan. The ability to borrow dramatically expands the buying power of the optimists. For every $1 of cash they put in, they can borrow $4 of debt. They can now leverage their cash of $2.5 million five times to buy houses worth up to $12.5 million. In fact, with the enhanced purchasing power that debt affords, optimists can buy all 100 houses in the market. When we introduce debt, the price of a house will be determined by the optimists’ willingness to pay. House prices immediately jump to $125,000 each when debt is introduced.
In the world with debt, optimists buy all the houses. They put down 20 percent, or $25,000, for each house, and they borrow the rest. But who is willing to lend to the optimists? Nobody is willing to part with their hard-earned money unless they are sure they will get their money back without loss. Since there are only two types of people in our world, pessimists must be willing to lend in order for optimists to borrow.
Will the pessimists lend? The pessimist thinks that a house is worth no more than $100,000, so he believes the optimist is overpaying. But he is perfectly willing to lend $100,000 to the optimist to buy the house for $125,000. Why? The pessimist has the house as collateral. In the pessimist’s judgment, the overly confident optimist will be forced to part with his down payment once the bubble bursts and prices return to their true valuation of $100,000. But the pessimist understands that his money is protected. He made a loan of $100,000, and the house is ultimately worth $100,000.
In this simple example, debt facilitates an increase in the price of assets by enabling optimists to increase their influence on the market price. Ironically, it is the pessimists—even though they disagree with the valuation of optimists—who make it happen. Without help from pessimists, the optimists would not be able to raise the price of a house by 25 percent. This is a crucial lesson when we think of assigning blame after a crash. We are more than willing to blame “irresponsible home owners” who stretched to buy houses. But the house-buying binge was only possible given the aggressive lending behavior by banks.Debt raises house prices in the example above, but is this necessarily a bubble? We label people in our economy as either optimists or pessimists. Whether the increase in house prices represents a bubble depends on which of the two is right. If the optimist is right, then house prices will remain at the elevated level, and there will be no bursting of any bubble and no crisis. However, if pessimists are right, then the increase in house prices will be temporary and sometime in the future the bubble will burst.
In addition to facilitating bubbles, debt also helps sustain them—for a while at least—due to its impact on expectations. A relaxation in access to debt means that more optimists can enter today and in the future. This bolsters the belief that there will be a greater fool who will buy the asset at even higher prices. And the party gets even bigger. The expectation of a bubble growing even more entices speculators to enter the market in addition to optimists. Notice that there is an element of animal spirits in our explanation of the housing boom, even when debt plays an important role. The optimists in our framework can be viewed as irrational buyers of homes, willing to pay more and more because they believe house prices will rise forever. In this sense, the debt view and the animal spirits view are not mutually exclusive.
But the big difference is the role of debt. In the debt view, the bubble cannot get out of control unless irrational optimists are able to get debt financing to sustain it. Further, even rational speculators may enter the market if they believe that irrational optimists can still get loans as the bubble expands. Debt plays a crucial role. The distinction is important, because some argue that debt had little to do with the house-price bubble in the United States before the Great Recession.Why Lend into a Bubble?
As the example above illustrates, lenders are willing to lend only because they are convinced that their money is safe. They are sure that the underlying collateral protects them even when house prices inevitably decline. Debt leads to bubbles in part because it gives lenders a sense of security that they will be unaffected if the bubble bursts.But what if lenders are wrong? What if they are actually exposed to this risk? The answer is closely related to a phenomenon that Nicola Gennaioli, Andrei Shleifer, and Robert Vishny call “neglected risks.” They argue that certain unlikely events can materialize that are completely unexpected, because investors neglect the risks that they could happen. In the context of the housing crash, many investors may have neglected the risk of house prices falling more than 10 percent. During the financial crisis, people investing in money-market funds may have believed that no fund could ever “break the buck,” or pay back less than the nominal amount put in the account.
Obviously, such neglect leads investors to make systematic mistakes and exercise poor economic decision-making. But Gennaioli, Shleifer, and Vishny show how the financial sector amplifies this neglect and produces a full-blown financial catastrophe. The key insight is that bankers will create securities that are vulnerable only to these neglected risks. In other words, the securities sold to investors will load heavily on the neglected risk itself. For example, if investors convince themselves that house prices throughout the country cannot fall by 10 percent or more, then bankers will create securities that retain their value in every scenario except when house prices throughout the country fall by 10 percent or more. Because these securities look riskless to investors, they will be produced in abundance. This large expansion in the supply of securities that look riskless will fuel an asset-price bubble by allowing optimists to buy even more expensive homes. When house prices do in fact fall more than 10 percent, the result is financially catastrophic.
What is the best kind of security to sell to investors who neglect certain risks? Debt. Debt has the unique feature that it convinces investors they will be paid back in almost every future scenario. An investor buying debt believes what they are holding is safe, independent of the underlying asset they are financing. The financial sector convinces investors that they are holding “super-safe” debt even in the clear presence of an asset bubble. This helps us to understand why Kindleberger found another common historical pattern: asset-price bubbles were often fueled by debt that looked extremely safe. As he put it, “In many cases the expansion of credit resulted from the development of substitutes for what previously had been the traditional monies.” Creditors were convinced that new debt instruments were as safe as currency backed by precious metal or government guarantees.
There is another lesson behind the neglected-risks framework: Debt instruments lead investors to focus on a very small part of the potential set of outcomes. As a result, they tend to ignore relevant information; they may even miss blatant fraud. Suppose, for example, investors provide a loan to a business. If the investors are convinced that their loan will be repaid even if the business manager steals some money from the cash drawer, then the investors are willing to ignore the stealing. In contrast, if the investors are equity investors, meaning that they share the profits of the business, they will have a strong incentive to detect theft. Debt convinces investors that they don’t have to worry about fraud because their senior claim on the asset protects them.
In a world of neglected risks, financial innovation should be viewed with some degree of skepticism. If investors systematically ignore certain outcomes, financial innovation may just be secret code for bankers trying to fool investors into buying securities that look safe but are actually extremely vulnerable.
* * *
In a cruel twist of irony, Kindleberger passed away in 2003 at the age of ninety-two, just as the mortgage-credit boom was starting. He did an interview with the Wall Street Journal the year before he died. What market concerned him the most? Housing. As the article put it, “The object of his greatest fascination today is the real-estate market. For weeks, Mr. Kindleberger has been cutting out newspaper clippings that hint at a bubble in the housing market, most notably on the West Coast.” He wasn’t yet certain, but he suspected a housing bubble. He saw one telltale sign: “Banks are ready to mortgage more and more and more and more,” he said. “It’s dangerous, I think.”
Source: House of Debt
wonkmonk's notes 