Financial Market Chaos Explained

"It was a million-to-one shot, Doc. Million-to-one!"

The current financial crisis, like most that have come before it, has resulted largely from the excessive use by banks of something called “leverage.” If you have an MBA, you know what this word means. Evidently, however, an MBA does not qualify you to use it with any degree of prudence, as the near-collapse of our financial system has just shown.

Most people who buy stocks are accustomed to putting up all of the cash that is necessary to purchase it. The stock costs $5 a share, and you want to buy 100 shares, so you yourself fork over the $500 in cash to pay for it. But it does not have to be done this way. If you open a margin account with your broker, you only have to put up half the cash necessary to buy the stock. The portion that you deposit yourself is called “margin.” Your broker can put up the rest of the money for the stock purchase in the form of a loan to you.

Why do this? It enhances the potential return on your investment. Instead of buying 100 shares with my $500, I can now buy 200 shares. If I only bought 100 shares and the stock went up by $1 per share, I’d make a profit of 100 * $1, or $100. But if I were using leverage, and I bought 200 shares instead, I’d make a profit of 200 * $1, or $200. My potential return has now doubled; instead of 20 percent ($100/$500), it is now 40 percent ($200/$500). Or, perhaps I’ll use $250 to purchase 100 shares and the remaining $250 of my money to buy something else. Either way, the potential profit on my investment doubles, to 40 percent.

However, so does my risk. If the stock drops by $1, I lose 40 percent using leverage and only 20 percent using all cash. And if it goes down by more than $2.50, I’m broke and could even owe money to my broker. An SEC regulation, imposed after the 1929 stock market crash, prevents stock market participants from using more than 2:1 leverage; in other words, I am restricted to borrowing only double the amount that I put up myself, as in the above example. For other assets, these restrictions do not apply.

Losing your shirt, and then some

In a world where markets hardly ever exhibit extreme movements, using lots of leverage would not be particularly hazardous. Alas, we do not inhabit such a world. Most days, markets move up and down by small amounts. But quite frequently, markets can go wild, moving multiples of their normal volatility in a day, or even several days in a row. What’s worse, sometimes all markets go haywire in unison.

This is a statistical property that quant-types refer to as an “infinite variance.” What it means is that, in the financial markets, things that have never happened before happen all the time. Out there lurking somewhere is a much larger price movement than we have ever seen previously. The French mathematician Benoit Mandelbrot was the first to point this out. Nassim Nicholas Taleb has brilliantly expounded on it, as have others. Apparently, the last people to realize this were the legions of PhDs employed to "measure risk" in the trading departments of the major Wall Street banks.

How can a bunch of quants with advanced graduate degrees commit such an error of judgment, especially when so many others have long recognized the wild nature of markets? To use the fancy statistical methods they so enjoy requires the analyst to assume that market volatility never strays very far beyond the norm. Otherwise, the results these models produce are meaningless. And what the hell did I go to graduate school for if I can't even use these complicated statistical techniques? So they make the assumption that markets are generally placid and hope everything turns out alright.

A couple of years pass and a market crisis hits. In a few days, the investment bank manages to lose several multiples of all the trading profits it has accumulated over the last five years. A risk manager with a PhD from Berkeley is quoted in the press expressing his surprise that such an event has come to pass. For, according to his models, a market occurrence of this magnitude should happen only once every 10,000 years. He is summarily fired, along with the managers who hired him in the first place.

Financial markets eventually return to normal, and a new batch of PhDs are brought in. Employing the same fancy risk models, they once again place huge market bets using absurd leverage. After a few years, the next market blowup arrives. Once again, all the bank's hard-won trading profits since the last crisis are gone. Another risk manager with a Harvard PhD is quoted in the press expressing astonishment at this turn of events; his models predict such an extreme abberation to take place once every 25,000 years. He and his bosses are swiftly removed, and the proces begins anew.

One is reminded of Frank Costanza's explanation to the proctologist as to how a fusilli stauette of Jerry Seinfeld came to be lodged in his rectum: "Million-to-one shot doc, million to one!" If you need some evidence of this phenomenon from the credit crisis, see the following Financial Times article from August 2007. In it, Goldman Sachs utters bewilderment at a market event its models predict to occur once in 100,000 years.

In an environment where extreme price changes are frequent, excessive leverage is dangerous. Most people should never use more than 3:1 or 4:1 leverage trading anything. The financial institutions that just collapsed, or came near to collapse, were sometimes using leverage of 30:1. That’s ridiculous. With that kind of risk, a few positions that move against you at the same time can lead to huge losses. And that’s exactly what happened. Since real estate prices have started falling, big leverage has led to outsized losses on the mortgage-backed assets held by major financial institutions.

It remains to be seen what kind of long-term regulatory response the US government will come up with in response to this crisis. Whatever it is, it may very well include limits on the amount of leverage that any financial institution can take on. History shows this may not be as simple as it seems.

(Thus ends the short version of the crisis. Readers seeking a little more detail are invited to read on.)

Money and credit: More slippery than we think

Every asset price bubble is fueled by the expansion of credit, often beyond the limits central banks would like to impose. This was the thesis of Hyman Minsky, reproduced by late MIT economist Charles Kindleberger in his book, Manias, Panics, and Crashes, regarded as the classic text on the subject. Financial history, Kindleberger shows, is in large part a struggle between monetary authorities, who seek to limit the supply of money and credit, and private financial actors, who are continuously devising new and innovative ways of creating substitutes for cash. These cash substitutes are used to buy more assets and can occasionally bid their prices up to unseen heights, only to be followed by a crash once the dentists have finally decided to get in on the riches.

In the early 19th century, for example, financial institutions around the world began exchanging sterling bills of exchange instead of shipping the actual silver to one another. So when America started sending these bills to China as payment for goods, the silver itself could remain in the US. More silver enabled greater lending by banks, which led to an effective expansion in the money supply. This extra money had to go somewhere, so it contributed to a boom in asset prices.

In the mid-19th century, banks collectively established clearing houses to guarantee payments to each another. The clearing houses would issue clearing house certificates, which the banks could use in lieu of cash to settle payments to one another. The actual cash could then be diverted elsewhere, and in the 1850s it fueled a global boom in economic activity and asset prices. The excesses culminated in the Panic of 1857.

In the early 20th century, financial entrepreneurs established "trust companies," which were banks in all but name - except that they were not subject to traditional banking regulations and could engage in riskier forms of speculation. The trusts represented new ways of extending more credit, which spawned excessive commercial lending and stock market speculation. The failure of these trust companies precipitated the great Panic of 1907.

In the 1920s, the big innovation was "call money,” or funds that could be borrowed on margin to buy stocks. Under this system, you would only have to put up 10% of the value of your stock purchase. Your broker, borrowing call money from a bank, would put up the rest. As stock prices rose, brokerages had to attract more and more funds to support the continuing use of call money to buy stocks. To do so, the banks lending the call money to the brokerages raised the interest rates offered to depositors. By 1928-29, these interest rates had become so attractive that they began diverting massive amounts of funds away from normal consumption and investment. This, argues Kindleberger, was likely a major cause of the Great Depression.

The upshot of all this is that it may be a lot harder than we think for the US government to regulate future crises out of existence. People will always find new ways of engaging in excessive risk-taking. The cycle of booms followed by meltdowns will thus continue.

Derivatives

In recent times we've seen yet another innovative substitute for money and credit: derivatives. Derivatives allow you to place bets on the movement of asset prices without actually buying and selling the assets. Some derivatives are traded on established exchanges. An example is commodity futures. It would be somewhat impractical to bring live cattle onto the floor of the Chicago Board of Trade. To resolve this problem, a derivative of live cattle is traded instead. That is what a futures contract is; instead of trading the actual cow, you’re only buying or selling the right to take possession of a certain quantity of live cattle at a set price on some future date. Other commodities, like oil, gold, and soybeans, are traded in the same manner.

More recently we've seen the massive expansion of the market for over-the-counter (OTC) derivatives, which are traded not on exchanges but rather directly among major financial institutions and their clients. This market is worth many trillions of dollars and dwarfs the market for traditional exchange-traded derivatives. There are a number of good books out there that explain OTC derivatives in relatively clear language. The best I’ve seen is Traders, Guns, and Money by Satyajit Das. It has done much for my understanding of these arcane financial time-bombs.

CDOs

One example of an OTC derivative featuring prominently in the current crisis is the Collateralized Debt Obligation (CDO). The CDO enables lenders to offload onto somebody else the risk that borrowers may not pay their money back. Let's say a local mortgage lender extends a home loan to a "subprime" borrower whom the lender expects to default on his mortgage. The lender doesn't care, because upon finalizing the loan he sells the right to collect interest and principal payments to, say, Lehman Brothers. Lehman Brothers then takes a bunch of these mortgages, or corporate loans, or any other kind of asset, and packages them together into a CDO, which it then sells to customers.

Officially, however, Lehman Brothers itself does not perform this task. Rather, it sets up a Special Purpose Vehicle (SPV) to manage the restructuring process and subsequent sale to clients. What's the point of the SPV? First, the SPV is registered in an offshore tax haven. Second, much like the trust companies in 1907, it is a bank in everything but name and is completely unregulated. Third, by selling the rights to collect payments on these mortgages to the SPV, Lehman Brothers generates immediate cash flow. This is considered preferable to the long-term trickle of cash it would otherwise receive if it waited for the various mortgages to be paid off over time.

It hardly needs mentioning that this immediate cash flow translates directly into current profits for Lehman Brothers and, thus, bigger bonuses for its traders and executives. This commentator is not expecting them to give their bonuses back to shareholders now that the company has gone belly-up.

The SPV is basically structured as a bank, with assets on one side and liabilities on the other. The assets consist of the right to collect interest and principal payments on the mortgages. Let's say the CDO is worth $1bn and is composed of 100 loans of $10m each. (Recall that the SPV doesn't actually "own" the loans; it just owns the right to collect on these loans. This is what makes the CDO a "derivative.") Each of these loans has been granted to a different borrower, and the borrowers vary widely in their (perceived) ability to pay interest and principal on the loans.

The SPV raises the cash to "buy" this debt by selling the right to collect on these loans to customers - usually big financial institutions and wealthy individuals. This debt is divided into three categories. $920m of the $1bn total might consist of “super-senior” AAA-rated debt, considered very safe from the threat of default. Another $50m consists of "mezzanine" debt, which is subordinated debt, the principal on which can be collected only once the holders of the AAA-rated debt have been paid. The last $30m is equity. If any three of the 100 borrowers at the end of the debt chain default on their payments, the owners of the equity in the CDO do not get paid.

One might think that the holders of the equity and mezzanine debt are taking on all of the risk here. One would be mistaken. The current crisis has seen declines in the value of the “super-senior,” AAA-rated debt tranches as well. Remember: in the markets, things that have never happened before happen all the time. And because the financial institutions that trade these assets have taken on so much leverage, on the order of 30:1, the price of the assets has only to decline by a relatively small amount to produce massive losses for these firms. Goodbye, Lehman Brothers.

Conduits

In many cases, the large Wall Street banks do not even bother sell the assets on to long-term investors. They rather stash them in vehicles they have established called “conduits.” Conduits are similar to SPVs, in that they are located in offshore tax havens and buy the mortgages or other assets from the banks. But instead of selling these assets to long-term investors, the conduits raise the money to buy the assets by selling short-term “commercial paper." Commercial paper represents a promise by the conduit to pay the holder of the paper a stream of interest payments until the term, typically lasting a few weeks, expires. At this point the conduit gives the commercial paper holder, usually another big bank or money market fund, its money back. If the conduit is to keep holding onto the mortgage assets it owns, it has to issue more short-term commercial paper, and the process begins anew.

The sole point of conduits, so far as I can tell, is to generate immediate income for the bank and bonuses for its executives while hiding from investors the true extent of the risk the bank has taken on. For if the conduit fails to keep finding new short-term financing every few weeks, the bank must take the loss back onto its own books.

For years, the scam worked. Then came August 2007, when the market for short-term commercial paper seized up. Increasing defaults on subprime mortgage loans had spooked the holders of the commercial paper. So when the terms on the paper expired, as they routinely do, the banks could not find any more buyers. The same thing has happened several other times since then, including in the last week, causing very severe and very immediate problems for the major Wall Street banks. You see, the amount of money controlled by conduits amounts to $3 trillion. That’s trillion. A failure to find more financing for these assets could spell catastrophe. This has forced central banks to inject hundreds of billions of dollars into the financial system to help maintain an orderly market. They did the same thing back in March. But even this wasn’t enough to save Bear Stearns.

Credit-default swaps

Another important derivative involved in the current crisis is the credit-default swap. Let's say a bank makes a loan to a business but for some reason becomes concerned about the customer's ability to pay it back. Traditionally, the bank would have to call in the loan or sell it to another bank. This might antagonize the customer and jeopardize the bank's future relationship with him. Eventually, someone had the insight to offer to pay the bank a sum equal to the amount of the loan in the event that the customer defaulted. In return the bank would pay this person a premium. That way, the bank can dump the risk of default onto somebody else without actually selling the loan - or letting the customer know what it is doing.

This basically amounts to an insurance policy, but the person selling the insurance isn't licensed by the government to do so. So they call it a "credit-default swap" instead of insurance. When you read in the newspaper back in March that "the cost of insuring against a default by Bear Stearns has risen to $2.7m for each $10m of debt," the reference here is to the premium demanded by the issuers of credit-default swaps in return for insuring against a default by Bear Stearns.

“Too big to fail:” Avoiding financial Armageddon

Incidentally, Bear Stearns was the issuer of many credit default swaps. Its downfall threw into question the ability of banks to get the money they had been promised by Bear in the event that their customers default. If this were allowed to happen, then everybody who bought credit-default swaps from Bear Sterns would have panicked and sold the assets Bear was insuring, since they no longer had protection against default. With everyone a seller and nobody a buyer, this would have caused the prices of assets to spiral downward to zero.

If you’re a company and the prices of the assets you own start to plummet, this makes banks less willing to lend you the money you need to operate on a daily basis. You are put in a difficult position; you either raise enough cash to continue on, or you go bankrupt, and your assets are put on the market in a fire-sale. But when lots of companies start selling the same assets at the same time, the prices of these assets decline even further, causing more bankruptcies, more asset fire-sales, and yet more bankruptcies.

We have not seen this happen in the current crisis…yet. But history does offer up an example of what happens when this process is allowed to unfold. It was called the Great Depression, and it is precisely the scenario the Fed has been trying to avoid. When everybody wants to sell and nobody wants to buy, and the prices of assets are plummeting as a result, it is the Fed’s job to step in and lend money to solvent companies who are short of cash. That is what it means to be a “lender of last resort,” which is the Fed’s formal job description. However, insolvent companies – those whose debt obligations outstrip the value of their assets – should be allowed to fail.

In theory, at least. Some institutions have managed to make themselves “too big to fail.” That is, regardless of their solvency, they have to be bailed out, or the whole financial system goes to hell. Bear Stearns was considered too big to fail, and so the Fed bailed it out (actually, Bernanke prodded JP Morgan to buy Bear Stearns and and then guaranteed the investment. That’s basically a bail-out).

This was also the case with AIG, the world’s largest insurance company, which the Fed just saved from chaotic extinction with an $80 billion loan. The sheer size of this loan suggests that AIG’s problems may have gone beyond a mere need for short-term financing and extended to its very solvency. But AIG was too big to fail. It had become one of the market's biggest issuers of credit-default swaps, explained above. Since CDS are similar to insurance policies, AIG thought it wise to expand into these derivatives. Evidently, it got in over its head. Merrill Lynch was also too big to fail, which is why the Fed pushed Bank of America to absorb it. Lehman Brothers, on the other hand, had not quite managed to make itself too big to fail. So the Fed allowed it to collapse entirely.

A global solution?

After a few days, Fed and Treasury officials decided to stop bailing out institutions on a case-by-case basis and create a system-wide solution. The goal is to stop the prices of assets – mortgages and stocks – from falling way below the value they would obtain in an orderly market. So, under the new plan, the government will buy these assets, thereby putting a “floor” under their price. It will then invite private partners who specialize in distressed debt to restructure the assets and sell them. If all goes according to plan, the assets the government would now own will be sold at a profit, which will make up for the $700 billion the government has to shell out to buy the assets in the first place. $700 billion is a lot of money. Let’s hope it works.

The problem is, there are many other asset classes besides mortgages that have not yet entered the crisis. The major financial institutions continue to deal in CDOs and credit default swaps backed by corporate debt, auto-loans, credit card debt, and so on. There are untold billions of dollars in assets remaining that may yet go bad. If that happens, well, we're screwed.