- Blackrock, one of the largest hedge-funds, has a computerized trading system called "Alladin." It costs several billion dollars and consists of more than 5,000 interconnected computers. Six-hundred information technology employees interact with the system to manage $12 trillion in assets, according to "Larry Fink's $12 Trillion Shadow."
- Alladdin grinds through a $30 billion portfolio in a few seconds to relate the portfolio's holding to thousands of external events like interest-rate shifts and stock and other asset prices on markets worldwide. It then executes buy and sell orders to align the portfolio's designated risk profile with the analyzed events. Larry Fink, Blackrock's CEO, sees a need for this kind of nearly continuous risk-management to achieve the profits and protection against losses the increasingly volatile stock markets require.
- Fink and others have achieved remarkable profitability with this kind of IT-based trading. They may also have created the very volatility they try to protect against. On May 6, 2010, the Dow (an index of 30 of the largest U. S. stocks) plunged nearly 1,000 points, about 10 percent of its value. As the plunge began, IT operators, alarmed by what they saw, simply walked over to their computers and entered the code for "sell everything and shut down." Within seconds, hundreds of thousands of pending sell-orders had no place to go. Prices accelerated downward. At one point a stock worth more than $40 a share moments before found buyers when the market value reached 1 cent.
- A related IT phenomenon, high-frequency trading, has also profoundly changed the market environment. These trades take place from computer to computer with little human intervention. Manoj Narang, owner of one of the many high-frequency trading offices around Manhattan, noted an ongoing sale of NETL, one of many thousands of stocks trading at that moment on one of his computer networks, and shrugged, "Never heard of it," "Speedy New Traders Make Waves Far From Wall St." reports.
- High-frequency trading takes place so rapidly that humans can't intervene--they're too slow. A trade execution takes about 20 microseconds (20 thousandths of a millisecond) from start to finish. Many of these high-frequency trades match small discrepancies between trading prices on different markets. The trading program holds the stock just long enough to take advantage of the discrepancy before selling it again. One electronic trading house estimates that on average, it holds an asset about 11 minutes. This has little similarity with what most investors think of as investing and reveals a very imperfect relationship between trading at human speed and algorhythmic high-speed trading. "Algorhythmic" trading means trading conducted under the control of abstract mathematical formulas. Two weeks after the May 6 "meltdown," analysts still don't know what caused it or how to prevent it from happening again, according to "Surge of Computer Selling After Apparent Glitch Sends Stocks Plunging."
Blackrock's "Alladdin"
Abacadabra
But Who Creates the Volatility?
High-Frequency Trading
The Imperfect Relation Between Humans and Computers
SHARE