The Crisis of Crowding

Quant Copycats, Ugly Models, and the New Crash Normal

Ludwig B. Chincarini

• Publisher: Bloomberg Press
• Year: 2012
• Print ISBN: 9781118250020
• eBook ISBN: 9781118282717
• Edition: 1
• Topic: Business
• Subtopic: Finance
• Index: Business

A rare analytical look at the financial crisis using simple analysis

The economic crisis that began in 2008 revealed the numerous problems in our financial system, from the way mortgage loans were produced to the way Wall Street banks leveraged themselves. Curiously enough, however, most of the reasons for the banking collapse are very similar to the reasons that Long-Term Capital Management (LTCM), the largest hedge fund to date, collapsed in 1998. The Crisis of Crowding looks at LTCM in greater detail, with new information, for a more accurate perspective, examining how the subsequent hedge funds started by Meriwether and former partners were destroyed again by the lapse of judgement in allowing Lehman Brothers to fail.

Covering the lessons that were ignored during LTCM's collapse but eventually connected to the financial crisis of 2008, the book presents a series of lessons for hedge funds and financial markets, including touching upon the circle of greed from homeowners to real estate agents to politicians to Wall Street.

• Guides the reader through the real story of Long-Term Capital Management with accurate descriptions, previously unpublished data, and interviews
• Describes the lessons that hedge funds, as well as the market, should have learned from LTCM's collapse
• Explores how the financial crisis and LTCM are a global phenomena rooted in failures to account for risk in crowded spaces with leverage
• Explains why quantitative finance is essential for every financial institution from risk management to valuation modeling to algorithmic trading
• Is filled with simple quantitative analysis about the financial crisis, from the Quant Crisis of 2007 to the failure of Lehman Brothers to the Flash Crash of 2010

A unique blend of storytelling and sound quantitative analysis, The Crisis of Crowding is one of the first books to offer an analytical look at the financial crisis rather than just an account of what happened. Also included are a layman's guide to the Dodd-Frank rules and what it means for the future, as well as an evaluation of the Fed's reaction to the crisis, QE1, QE2, and QE3.

CHAPTER 1

Introduction

Since the beginning, it was just the same. The only difference, the crowds are bigger now.

—Elvis Presley

Humans are a social species, grouping together to do everything from waging war to attending cocktail parties. This desire to be with others also occurs in the financial markets. Some portfolio managers follow whatever investments are in vogue, mostly because bucking the trend doesn’t pay. If the trend continues and they haven’t followed it, they could look like idiots. If they follow it and it doesn’t continue, they simply join the herd of other wrong-headed investors.

At other times, portfolio managers create an innovation. This innovation usually makes abnormally large returns, so others desperately want to copy the strategy. These copycats eventually learn the ropes and begin trading money in the same fashion. At first this leads to even more profits for the early innovators, because others buy more and more of their trades.

The copycats create a side effect, however: They crowd the space. The strategy’s future returns depend increasingly on the copycat’s behavior.

Oftentimes copycat investors make their trades on borrowed money, which amplifies both their positions and their risks. Modern risk-measurement models generally ignore the presence of copycats and the resulting crowded spaces, which often leads to underestimations of risk. A shock to the system can lead to sudden, sometimes large asset price moves, which can cause panic and failure among the institutions involved in that investment space.

In the past 20 years, globalization, technology, and increased leverage have made the effects of overcrowding more apparent and dramatic. In fact, market crashes are happening more regularly than in the past, with nearly every crisis labeled a 100-year event.

The stock market crash of 1987 was likely the first crisis caused by modern-day crowding. The financial industry had popularized dynamic portfolio insurance, which involved protecting investors from losing money on their portfolios. Many institutions offered this protection by selling the market when it went down and buying the market when it went up. This practice can work quite well if only a small portion of the market pursues these strategies.

But if that proportion grows too large, crowding the space, the market may destabilize. As the market falls, the large group sells its positions, pushing prices further and further down and sometimes leading to a crash. In 1987 there were too many copycats, too much crowding, and too many models that didn’t adequately account for this crowding.

The next big crisis came in 1998, eleven years later. It involved Russian markets and the failure of Long-Term Capital Management (LTCM), a well-known hedge fund. In 1994, Long-Term Capital was one of the largest hedge funds. Managers used technological and quantitative techniques learned at Salomon Brothers to sublime perfection. They were the new financial juggernauts, and everyone wanted a piece of their amazing performance.

Soon other institutions, including the proprietary trading desks of Goldman Sachs, Morgan Stanley, Lehman Brothers, and multiple new hedge funds, began to reverse engineer LTCM’s strategies, all of which involved leverage. The lucrative relative-value bond arbitrage investment area became very crowded. Quantitative copycats saturated the space. Risk models were no longer accurate, because they didn’t capture this crowding and its potential effects. Heavily leveraged positions meant that small moves could destroy an entire firm in a short period of time.

In July 1998, one of the large institutions, Salomon Brothers, began closing its copycat positions. In August 1998, the Russian government defaulted on its bonds. The shock occurred as the relative value funds were scrambling to survive. LTCM was on the brink of bankruptcy; many feared that this would shatter the financial system, just as with Lehman Brothers in 2008. The Federal Reserve stepped in and coordinated a private solution to prevent chaos.

In 2000, Internet stocks traded at ridiculous multiples. The crowd rushed in and the bubble formed. By April 2000, the bubble began to crash. The NASDAQ dropped by 70%. Yet despite investors’ dramatic losses, the aftereffects were comparatively mild, mostly because of the limited amount of leverage in Internet stocks. This put some brakes on the crash.

From 2000 to 2008, every aspect of the U.S. economy got more and more involved in a massively leveraged trade: real estate investing. Instead of involving just traders, as most crowding does, the subprime lending bubble featured politicians, greedy home buyers, mortgage brokers, real estate agents, banks, investment banks, and quasi-government organizations Freddie Mac and Fannie Mae.

Investment banks took outright positions in real estate and also created, sold, and traded derivatives based on housing values. Hedge funds also took various bets on real estate market segments. Insurance companies joined the space by offering insurance to the crowded investors. Rating companies joined the greed train and issued AAA ratings as fast as they could write the three letters and cash the checks. Even the media pushed us forward with talk of rising home ownership, rising stock markets, and good times.

Like the Internet bubble of 2000, this bubble kept growing. Almost everyone was crowding this trade and using unprecedented leverage. Some home owners took leveraged investing to new heights by putting zero money down and enjoying a leverage ratio of infinity.

Risk models were glaringly inadequate. They used historical data, which didn’t include the enormous amount of crowding and overvaluation that existed by 2008. It was only a matter of time before we saw the worst crash since the stock market crash of 1929: the 2008 financial crisis. The massive exposure to a collapsing bubble combined with leverage and short-term borrowing created an unprecedented shock to quantitative hedge funds. Known as the Quant Crisis, this destroyed Goldman Sachs’s star hedge fund.

The crisis gave us a spectacular show: the historic collapse and rescue of Bear Stearns, a government rescue for Freddie Mac and Fannie Mae, hundreds of bank failures, Lehman Brothers’ bankruptcy, a market-wide lending freeze, the failure of a whole host of hedge funds (including John Meriwether’s new fund, JWMP), and unprecedented marketplace interventions from the U.S. government and Federal Reserve.

Three years and a depression later, the markets had slightly recovered. On May 6, 2010, between 2:42 p.m. and 2:47 p.m., the Dow Jones dropped by 600 points, then rose 600 points by 3:07 p.m., events known as the Flash Crash. Procter & Gamble stock dropped by 37% in that short period. What happened? Was a leveraged crowded space wreaking havoc again?

From 2001 to 2008, banks around the world lent money to Greece, assigning it a risk level very similar to that of countries with more discipline and higher productivity, such as Germany. The crowded space kept Greek interest rates at unrealistically low levels, and the Greeks were happy to borrow to fund consumption—until the crowd realized that Greece was a mess.

This is the story of the crisis of crowding. The story begins in 1998 with Long-Term Capital Management’s fascinating collapse and tries to explain the ways in which crowds and leverage demolished one of the most successful hedge funds in history. The failure of LTCM had many lessons for the financial community and for society at large, but no one paid much attention—perhaps because disaster was ultimately averted. Ignored lessons formed a large part of the basis for 2008’s financial disasters, only this time with more leverage, more participants, and a series of policy mishaps.

PART I

The 1998 LTCM Crisis

All human beings are interconnected, one with all other elements in creation.

—Henry Read

The 2008 financial crisis really began 10 years earlier, with the collapse of the famous hedge fund Long-Term Capital Management (LTCM). LTCM was not an ordinary hedge fund. It was a large financial intermediary with a vast amount of technology and a lot of very experienced, intelligent managers.

LTCM’s enormous, consistent success seemed evidence that it was possible to tame the financial markets with sophisticated experience and quantitative tools. Some people were jealous of LTCM’s success. Others were inspired, as it showed them that traders could understand and manage the financial markets.

Suddenly, in just two months in 1998, LTCM stood on the brink of bankruptcy. The firm would have failed without an emergency cash infusion and rescue from a consortium of investment banks. The rescue brought cheers from those who envied LTCM and cries from those who wanted to become LTCM, plus a lot of stories that weren’t even true.

Many of LTCM’s trades were clever. LTCM’s experience in the financial markets was second to none. Its risk-management framework was on a par with state-of-the-art systems, but the firm underestimated the danger posed by crowds. Lured by LTCM’s success, other investors had entered the firm’s investment space. LTCM’s risk management failed to measure the ways that these crowds changed investments’ return and risk. With leverage and quant copycats running for the exits, LTCM found itself trapped in the fire.

CHAPTER 2

Meriwether’s Magic Money Tree

We’re sucking up nickels from all over the world.

—Myron Scholes

The Birth of Bond Arbitrage

In 1974, John Meriwether, having just received his MBA from the University of Chicago, went to work as a government bond trader at Salomon Brothers. In those days, bond trading was not a quantitative endeavor. Traders bought or sold bonds they thought looked good or bad. John Meriwether realized that bond pricing was highly quantitative and saw that, if he could tap into this quantitative pricing, he could not only outperform his industry peers but make lots of money as well.

He slowly began recruiting top talent, hiring both highly trained quantitative and old-fashioned traders. He went to MIT, Harvard, and other places to find experts in economics, finance, and other sciences. He planned to teach them the basics of bond trading and then tap their intelligence to find ways to mathematically model various fixed-income products, find inherent mispricings, and make money on them. Some of his hires included Larry Hilibrand, who had just finished a master's degree in mathematics from MIT and was hired in 1980. Dick Leahy, with a BS from Boston State College, worked at Merrill Lynch and then joined Meriwether in 1986.

Eric Rosenfeld, a PhD in economics from MIT, was working at the time as a professor at Harvard Business School. Meriwether called to ask if Rosenfeld had a bright student who would like to come to Salomon. Rosenfeld was sick of teaching case studies and grading exams and asked Meriwether if he could try the job. He left Harvard 10 days later and never went back.

Rosenfeld’s interest in using quantitative techniques to exploit profitable opportunities started when he was an MIT undergraduate in the early 1970s. He enrolled in a statistics class with the famous econometrician Jerry Hausman. Hausman needed a summer research assistant to help build predictive models for NFL football games. For the previous 12 years Rosenfeld had collected NFL football game data, including the game’s day of the week, which team was home or away, the Las Vegas betting spread, whether the game was played on turf or grass, and each team’s winning percentage at the time of any given match. The pair used this data to build an econometric model to predict the winning margin on NFL games, then bet on games. The model worked during the two-year project; then both researchers moved on to other things.

Many recruits had interesting backgrounds independent of their financial experience. Eric Rosenfeld and fellow MIT student Mitch Kapor created a regression program at MIT to help Eric with his PhD dissertation. They eventually called the program VisiPlot and sold it alongside the first spreadsheet program, VisiCalc. They sold the rights to a software company for about $1.2 million.

Rosenfeld went on to teach at Harvard Business School, while Kapor went on to work for that software company. Realizing that one product could combine all ...

Table of contents

• Cover
• Additional Praise
• Series
• Title Page
• Copyright
• Dedication
• Foreword
• Preface
• Cast of Characters
• Chapter 1: Introduction Introduction
• Part I: The 1998 LTCM Crisis
• Part II: The Financial Crisis of 2008
• Part III: The Aftermath
• Appendices
• Glossary
• Bibliography
• About the Author
• Index
• End User License Agreement