High Frequency Trading

The drastically reduced cost of electronic trading has made a huge variety of new and different arbitrage strategies feasible. Led by Jim Simons' Medallion Fund, the high frequency hedge fund category is now the most profitable on Wall Street. One industry watcher predicts this type of trading will reach 50% of total stock exchange volume by 2010. 

High frequency hedge funds differ from Wall Street traditions in two principal ways. At Medallion 25% of the payroll is said to be mathematics, physics, and computer science PhDs, rather than MBAs. Instead of adding borrowed funds to invested capital as leverage to increase portfolio holdings, Medallion is said to be de-levered, leaving three quarters of its investors' funds sitting in cash. 

Margins and trade frequency are also far from the norm. Anecdotal evidence indicates Medallion's trading volume is "several percent" of NASDAQ's annual $10 trillion total, implying something like a half trillion, and a recent year's profit before fees was 80% on its $5 billion in assets, or about $4 billion. Dividing profit by volume shows a tiny trading margin of the order of three quarters of one percent. Dividing volume by its $1 billion or so portfolio shows annual turnover of 400 times, or about 2 times a day.


Analysis was done of a sample strategy based on using put option "insurance" purchases to predict institution-sized IBM stock purchases. It appears the options, being less liquid than the stock, are typically bought first by an institution planning to acquire a stock, to minimize risk. During a recent day, eight significant volume and price spikes were noticed in IBM's nearby out-of-the-money puts, the options favored for insurance.  

Tracking the IBM stock at one minute intervals after each of these option spikes showed that if stock were purchased within 30 seconds and sold ten minutes later, these trades would be profitable 80% of the time, with profit averaging 1/4 %. While this was but a third of Medallion's tiny take, it is still twice expenses, mostly commission and spread. Multiplying these eight profits by a 200 day trading year, one gets 1600 times the net margin of one-eighth percent , or 200% gain on the portfolio, which would be 50% on total assets.  

To start a fund utilizing this strategy, a one employee shop would need to raise $15 million, based on scaling down Medallion's $5 billion size and 300 person payroll. Portfolio size would be about $4 million and gross profit $7.5 million. Hedge fund operating expenses and management fee of a few percent each would about equal and offset interest earned on the 75% of capital held as cash. Management incentive fee, typically 20% of profits, or $1.6 million, would leave a return to investors of 5.9 million, or 39% on their 15 million at stake. This is 2% better that the 37% paid out by Medallion, which is not only closed to new investors, but recently paid off (kicked out) existing ones as well. 

Bad news is that 2010 is fast approaching, and rapidly increasing volume implies rapidly increasing competition. To counteract this, Medallion's PhDs are reportedly continuously searching for new strategies.  

Anyone contemplating action in this form of investing may wish do some reading as a first step. Barton Biggs' autobiographical tale "hedgehog" gives a day to day snapshot of the life of hedge fund management; LeFevre's "Reminiscences of a Stock Operator" details the classic trading ploys of the last century (which still work today); Taleb's "Fooled by Randomness" explores by statistical reasoning and computer simulation the hidden risks by which famous traders and fund managers "blow up"; and Swenson's "pioneering" details the nuts and bolts needed to effectively manage money. All are very readable, a rarity in finance.