Four Big Risks of Algorithmic High Frequency Trading
Algorithmic trading (or "algo" trading) refers to the use of computer algorithms (basically a set of rules or instructions to make a computer perform a given task) for trading large blocks of stocks or other financial assets while minimizing the market impact of such trades. Algorithmic trading involves placing trades based on defined criteria and carving up these trades into smaller lots so that the price of the stock or asset isn't impacted significantly.
The benefits of algorithmic trading are obvious: it ensures "best execution" of trades because it minimizes the human element, and it can be used to trade multiple markets and assets far more efficiently than a flesh-and-bones trader could hope to do.
What is Algorithmic High-Frequency Trading?
High-frequency trading (HFT) takes algorithmic trading to a different level altogether -- think of it as algo trading on steroids. As the term implies, high-frequency trading involves placing thousands of orders at blindingly fast speeds. The goal is to make tiny profits on each trade, often by capitalizing on price discrepancies for the same stock or asset in different markets. HFT is diametrically opposite from traditional long-term, buy-and-hold investing, since the arbitrage and market-making activities that are HFT's bread-and-butter generally occur within a very small time window, before the price discrepancies or mismatches disappear.
Algorithmic trading and HFT have become an integral part of the financial markets due to the convergence of several factors. These include the growing role of technology in present-day markets, the increasing complexity of financial instruments and products, and the ceaseless drive towards greater efficiency in trade execution and lower transaction costs. While algorithmic trading and HFT arguably have improved market liquidity and asset pricing consistency, their growing use also has given rise to certain risks that can't be ignored, as discussed below.
The Biggest Risk: Amplification of Systemic Risk
One of the biggest risks of algorithmic HFT is the one it poses to the financial system. A July 2011 report by the International Organization of Securities Commissions (IOSCO) Technical Committee noted that because of the strong inter-linkages between financial markets, such as those in the U.S., algorithms operating across markets can transmit shocks rapidly from one market to the next, thus amplifying systemic risk. The report pointed to the Flash Crash of May 2010 as a prime example of this risk.
The Flash Crash refers to the 5%-6% plunge and rebound in major U.S. equity indices within the span of a few minutes on the afternoon of May 6, 2010. The Dow Jones plunged almost 1,000 points on an intraday basis, which at that time was its largest points drop on record. As the IOSCO report notes, numerous stocks and exchange-traded funds (ETFs) went haywire that day, tumbling by between 5% and 15% before recovering most of their losses. Over 20,000 trades in 300 securities were done at prices as much as 60% away from their values mere moments earlier, with some trades executed at absurd prices, from as low as a penny or as high as $100,000. This unusually erratic trading action rattled investors, especially because it occurred just over a year after the markets had rebounded from their biggest declines in more than six decades.
Did "Spoofing" Contribute To the Flash Crash?
What caused this bizarre behavior? In a joint report released in September 2010, the SEC and the Commodity Futures Trading Commission pinned the blame on a single $4.1-billion program trade by a trader at a Kansas-based mutual fund company. But in April 2015, U.S. authorities charged a London-based day trader, Navinder Singh Sarao, with market manipulation that contributed to the crash. The charges led to Sarao's arrest and possible extradition to the U.S.
Sarao allegedly used a tactic called "spoofing," which involves placing large volumes of fake orders in an asset or derivative (Sarao used the E-mini S&P 500 contract on the day of the Flash Crash) that get cancelled before they are filled. When such large-scale bogus orders show up in the order book, they give other traders the impression that there's greater buying or selling interest than there is in reality, which could influence their own trading decisions.
For example, a spoofer may offer to sell a large number of shares in stock ABC at a price that's a little away from the current price. When other sellers jump in on the action and the price goes lower, the spoofer quickly cancels his sell orders in ABC and buys the stock instead. Then the spoofer puts in a large number of buy orders to drive up the price of ABC. And after this occurs, the spoofer sells his holdings of ABC, pocketing a tidy profit, and cancels the spurious buy orders. Rinse and repeat.
Many market-watchers have been skeptical of the claim that one day trader could have single-handedly caused a crash that wiped out close to a trillion dollars of market value for U.S. stocks within minutes. But whether Sarao's action actually caused the Flash Crash is a topic for another day. Meanwhile, there are some valid reasons why algorithmic HFT magnifies systemic risks.