By Alan Anderson, David Semmelroth

One area of the finance industry that has been dramatically affected by big data is the trading activities of banks and other financial institutions. An example is high-frequency trading (HFT), a relatively new mode of trading that depends on the ability to execute massive volumes of trades in extremely short time intervals. HFT traders make money by executing a huge number of trades, each of which earns a miniscule profit.

Unlike traditional traders, HFT traders don’t attempt to hold positions for any great length of time and don’t base their trades on fundamental factors such as interest rates, exchange rates, commodity prices, and so forth. The success of HFT trades depends critically on the speed of execution, as they are based on rapid fluctuations in market prices.

As more and more resources have been dedicated to HFT trading in the last couple of years, leading to an “arms race” in progressively faster hardware and software, the profitability of high-frequency trading has declined. As the speed of transactions has increased, the ability to make money based on speed alone has diminished. Further increases in speed are now bringing steadily diminishing returns — the profit per transaction has plunged. As a result, successful trading now depends less and less on hardware and more on software in the form of sophisticated trading algorithms.

An algorithm is a set of instructions used to carry out a procedure, kind of like a recipe. Algorithms are heavily used by computer scientists to instruct computers on how to perform various tasks, such as carrying out mathematical operations.

The use of advanced algorithms for trading strategies carries several potential advantages, such as the ability to test ideas on historical data before risking any money. With HFT trading, there’s no time to test any potential trading strategies, because they must be implemented immediately.

Another advantage to using trading algorithms is that they can be based on fundamental variables, such as interest rates and exchange rates, instead of simply searching through trades to look for temporary price changes. As a result, algorithms can be developed to find ever more complex relationships among securities prices and use this information to earn trading profits. Big data enhances algorithmic trading by providing the ability to search through enormous volumes of data looking for patterns that might not be detectable with smaller amounts of data or slower processing speeds.

With shrinking profits from HFT, algorithmic trading appears to have a bright future, as the increasing availability of data and computer speed enable more and more sophisticated algorithms to be developed.