Nova last night was a repeat of a 2000 show on the spectacular failure of Long Term Capital Management, a hedge fund setup to use the Black-Scholes and similar models (Scholes himself was one of the principals) to make hugely leveraged investments in virtually all the world’s financial markets. LTCM collapsed and threatened to take the markets down with it, as this book (When Genius Failed: The Rise and Fall of Long-Term Capital Management) explains:
LTCM began trading in 1994, after completing a road show that, despite the Ph.D.-touting partners’ lack of social skills and their disdainful condescension of potential investors who couldn’t rise to their intellectual level, netted a whopping $1.25 billion. The fund would seek to earn a tiny spread on thousands of trades, “as if it were vacuuming nickels that others couldn’t see,” in the words of one of its Nobel laureate partners, Myron Scholes. And nickels it found. In its first two years, LTCM earned $1.6 billion, profits that exceeded 40 percent even after the partners’ hefty cuts. By the spring of 1996, it was holding $140 billion in assets. But the end was soon in sight, and Lowenstein’s detailed account of each successively worse month of 1998, culminating in a disastrous August and the partners’ subsequent panicked moves, is riveting.
Models like Black-Scholes are inherently unreliable, as this collapse shows.
Utter silliness. Black-Scholes, to be valid, depends on the very assumption that one cannot guarantee oneself money by using it or any other model. The key assumption is that arbitrage opportunities don’t exist over the long run.
How stupid for anyone to claim that they could beat the system using a model that claims that no one can beat the system.