Why the Internet Makes It Impossible to Stop Giant Wall Street Losses

Information technology has allowed banks to make huge bets that can go bust in the blink of an eye, and more regulation won't help.


In 1984, Yale sociologist Charles Perrow published his classic book, Normal Accidents: Living with High-Risk Technologies. The odd term, "normal accident," Perrow wrote, is meant to signal that, given a system's characteristics, multiple and unexpected interactions of failure are inevitable. Perrow studied "normal accidents" that occur in nuclear power and petrochemical plants, at sea where ships collide in open water for no good reason, in air travel, and in defense systems.

Perrow's work is also applicable to banks, economic systems, and financial networks. When Jamie Dimon, the chief executive of JPMorgan Chase, was at Harvard Business School, Perrow's book should have been required reading. Had Dimon absorbed the book's lessons, instead of blaming the bank's recent embarrassing $2 billion dollar trading loss on errors, sloppiness, and poor judgment, he might have said,  "We should expect accidents like this to occur. No reasonable level of controls can prevent them from happening in tightly coupled, very active financial systems. We can handle it."

In fact, if Jamie Dimon, Christopher Dodd, and Barney Frank had all read Perrow's book, they would have discovered that for a large number of complex systems, neither better organization nor technology will make them less prone to accidents, and that frequently they become even more vulnerable to certain types of accidents. 

The recent history of the financial world reads like a textbook account of normal accidents.  Some big failures of note: Lehman Brothers and AIG during the 2008 financial crisis; the May 6, 2010 Flash Crash when the Dow Jones Averages dropped 600 in five minutes, only to bounce back twenty minutes later; the implosion of the Icelandic economy in 2008 that resulted in a 50 percent drop in the value of that nation's currency, and the unchecked growth in the over-the-counter derivatives market -- from a notional value of $60 trillion in 2000 to over $700 trillion today.

The financial world is more accident prone than ever -- thanks in large part to the Internet, the mother of all interconnections. Not only has the Internet supercharged financial innovation and created high rates of growth, but it lies at the heart of many financial normal accidents in the 21st century. In the case of Iceland, the country was welcomed into the European Economic Area, which enabled Icelandic banks to operate throughout the continent as long as they had deposit insurance.  At the time, Iceland had a gross domestic product of less than $20 billion.  Yet its banks mushroomed in size. By 2008 they had over $100 billion in assets as investors raced to capitalize on high interest rates and the rising value of the Icelandic kronur.  Iceland's online banks sucked in $6 billion in deposits from consumers in Britain and the Netherlands in a few years.

But when investors lost confidence in the Icelandic economy, it triggered a normal accident. Money that had flowed in over electronic networks, fled at Internet speeds--an electronic run on the banks. There was no way the Icelandic Deposit Compensation Fund could meet its commitments to depositors. Iceland had created banks that were too big to fail in a country that was too small to save them. The kronur went into free fall, and lost half its value. 

Over-the-counter derivatives live in a dark and non-transparent world.  High levels of connectivity have made it possible for the market for these financial products to grow to massive size. The Internet provides the information backbone for the transactions. The continuous flow of information allows them to evaluate portfolio risk, keep track of trading positions, and make trades. These information-age tools allowed Lehman Brothers to assemble and manage a portfolio that contained 930,000 derivative transactions at the time of its bankruptcy. As one former Lehman employee told me, positions of this size "could never have happened without the Internet."

If you are curious as to how the notional value of over-the-counter derivatives could top $700 trillion when the total output of goods and services of all the world's economies is around $60 trillion, think of it this way: A bank buys a derivative to protect against loss.  If the market moves against that derivative, the bank sells a similar derivative to compensate. If the market reverses again, the bank makes another offsetting transaction. If you do that often enough, you end up with $700 trillion in notional value. In the process you create a volatile accident-prone financial market just waiting for an aggressive trader to make a mistake.

Enter Bruno Iksil, JPMorgan's "London Whale." JPMorgan played a game similar to the one above. Unfortunately, Iksil, a trader based in JPMorgan's London office, had taken such big positions that when the market flipped, he could not unwind them fast enough. When others discovered the exposure they threw financial gas on the fire; placed winning bets against JPMorgan's positions that increased the bank's losses. BlueMountain Capital and BlueCrest Capital Management reportedly each made $30 million capitalizing on JPMorgan's pain -- which is how Iksil, who made his massive bets in a thinly traded market, came to be nicknamed The London Whale. In the end, the normal accident cost JPMorgan $2 billion, and that amount could  easily double.

This is a different kind of normal accident from the one Perrow studied. Nobody tried to make the meltdown at the Three Mile Island nuclear power plant worse. 

I agree with Jamie Dimon's assertion that  Dodd-Frank, the legislation that is supposed to save us, is seriously flawed.  But our reasons differ. Dimon believes it will stifle growth, while I believe that it will become so filled with loopholes that banks will exploit it in ways that make them, in Perrow's words, "more prone to certain kinds of accidents."

The drafters of the bill hope that it will make financial normal accidents smaller and less frequent. Maybe it will. But I think there is a better way: make banks that are too big to fail a lot smaller. That way, when they become victims of normal accidents, it won't hurt as much.