Rarely have I seen a practice as universally misunderstood as high frequency trading. People seem to misuse the term high frequency trading in much the same way that the term hedge fund was misused in the past. Hedge funds were often referred to as an investment style rather than a legal structure for unregulated investment vehicles. While hedge funds can be broken up into various style buckets such as long/short equity and convertible arbitrage the style buckets often share little resemblance with one another and the funds within the style buckets can often vary widely in their investment approach. For example; one long/short equity fund may focus on small cap stocks while another focuses on technology stocks.
Hedge funds and high frequency trading firms remind me of snowflakes in that no two are perfectly identical. While snowflakes are each unique they all share a common core structure. For example; snowflakes are hexagonally arranged frozen water molecules with perfectly straight sides (facets) which are angled at 120° to one another. Each snowflake builds upon it’s common core structure and develops differently depending on its unique environment and interactions with its surroundings.
While no two are exactly alike, high frequency trading firms also all share common traits. To be considered a high frequency trading firm you must deploy fully automated trading strategies, across one or more asset classes, which identify and profit from short-term pricing inefficiencies. These inefficiencies may last from milliseconds to hours. Most high frequency trading firms close their positions at the end of every trading session and hold cash (or remain market neutral) until the resumption of trading. High frequency trading strategies try to make small amounts of money on each trade. The profits from these trades are amplified by high volume.
Estimates suggest that HFT accounted for about 15% of US equity share volume as recently as 2005. The Tabb Group estimates that high frequency trading is now responsible for 61% of US equity share volume. They also estimate that 83% of HFT firms trade equities, 67% trade futures, 58% trade options, 36% trade bonds and 26% trade FX.
It is largely believed that there are three broad investment strategies that comprise high frequency trading including automated market makers (the most common), predictive traders and arbitrage traders. These players are typically traditional broker dealers, hedge funds or proprietary trading firms using private capital. The Tabb Group estimates that there are between 10 and 20 broker dealer prop desks and fewer than 20 active hedge funds employing HFT techniques. The independent proprietary trading firms are believed to be between 100 and 300 in number.
Automated market makers provide liquidity to market participants and generate revenue from rebate trading. Rebates are earned when traders buy or short stock over and ECN and receive a rebate or payment from that ECN for “adding liquidity”. The ECN charges those who “take liquidity” and use a portion of that fee to compensate liquidity providers. This practice is typically employed in high volume stocks because the HFT firm has to be assured that they can trade the stock without moving the price (since price movements do not account for their strategy). This has lead to asymmetrical volume in the markets where active stocks benefit from even more liquidity while low volume stocks are ignored.
Arbitrage traders typically look to benefit from short term divergences from price correlations between stocks. There are also volatility traders who look to profit from differences in implied volatility and future forecasts for realized volatility with stocks underlying options contracts. Pairs traders try to exploit relative price discrepancies between closely related companies (think Home Depot and Lowe’s). Stocks which benefit from the same macroeconomic factors are often highly correlated. When a price correlation breaks down the stocks are paired against each other in a bet that their price relationship will revert to the mean. Short term stat arb resembles pairs trading but may pair various securities or baskets against each other.
The most controversial strategy is liquidity detection. High frequency traders that employ liquidity detection strategies have developed programs which look to identify large institutional orders sitting in dark pools or other liquidity venues. They do this by sending small orders on recon missions. The small orders interact with large orders by being filled very quickly. When this happens repeatedly or when orders are executed in larger amounts than the displayed size of the market hidden liquidity is present. When large orders are identified the HFT firms trade ahead of that order in the belief that large orders will move the market which they will benefit from.
There are certainly pros and cons with high frequency trading. On the positive side, high frequency traders add liquidity, narrow spreads and provide greater book depth. They ensure greater price discovery and provide for more orderly markets. On the negative side, liquidity detection programs may cost investors money by making it more difficult to trade large quantities of stock at good prices. High frequency traders who act as virtual market makers also profit at the expense of designated market makers. Designated market makers are appointed by exchanges which require them to be on both sides of the market. They are required to make markets in both highly liquid and illiquid stocks. It is more difficult to make money in illiquid stocks (despite their wider spreads) but it is looked at as the cost of doing business. Virtual market makers are not bound by these same rules and can cherry pick the liquid stocks. This makes it more difficult for the designated market makers to earn a profit. If HFT traders are allowed to act as market makers without having to adhere to the requirements of market makers, it could undermine the trading of less liquid stocks and further divide the market along market capitalization lines.
High frequency trading is on the top of the SEC's regulatory agenda for 2010 as a result, in part, of the pressure being applied by Senator Edward Kaufman who was very recently quoted as saying that he was concerned that the amount of money being allocated to high frequency trading strategies could quintuple over the next year.
Turning back the technological clock is impossible but one can rest assured that high frequency trading as we know it today will most likely look different in 2010.
Posts
