High Frequency Trading and the Use of Algorithms
When people say time is of the essence, what do you think of? A day? An hour? Maybe even half an hour? What about 3 milliseconds? This is the time saved by Spread Networks’ fiber optic cable, newly installed within a nearly straight tunnel through mountains and all between the Chicago exchange and the New York exchange based in New Jersey. Information can now travel from Chicago to New Jersey and back in a mere 13 milliseconds. Although this seems like nothing to an average person, companies were willing to spend millions to use it. In fact, the first 200 companies that signed up to use the tunnel’s information paid $2.8 billion combined.
This was so companies could process large transactions faster than the market could react. This is known as High Frequency Trading (HFT). Having a faster speed gives companies the ability to buy a company’s stock before the price increases or sell it before the price has a chance to decrease in a process known as front running. Although it is illegal for brokers to buy shares of a company before placing a large order for a client that would influence the price, it is legal for traders to process transactions faster as it is considered public knowledge.
To achieve these speeds companies use fiber optic cables, microwave connections, and co-locating. Fiber optic cables provide the fastest signal over a long distance. For shorter distances, however, microwaves are faster. This is why Spread Networks used fiber optics as the fastest way to transfer information from New York to Chicago. Fiber optics also has the ability to transfer more information at one time. The only downside is that fiber optics requires a greater initial start-up cost due to the fact boring holes through mountains and other requirements along with the cables necessary to be as efficient as possible are all very costly. Spread Networks spent hundreds of millions of dollars to build the fastest route possible. Companies have also begun co-locating their servers. This is the process of placing computers near an exchange, in order to cut down on latency, again just milliseconds at a time. With a mix of these three operational strategies, companies are able to make transactions faster than anyone else.
For these computers to work and actually save time, they must run algorithms to execute decisions. The formulas, created by humans, allow computers to process a buy or sell transaction whenever the parameters are met near instantaneously. Thus, processing and transmitting speed are vitally important, especially when these companies compete against each other to complete large transactions, where the price could fluctuate based on the size of the trade and the speed of who gets there first. It is estimated that somewhere between 70% and 90% of trading in the US today is done by algorithms.
Many people take issue with this, claiming that the speed of corporations hurts the average trader. To help solve this, Rep. Peter DeFazio is introducing a bill called the Wall Street Tax Act of 2019, which “could have a particular negative effect on high-frequency traders” by taxing securities transactions. This would help raise nearly $800 million over the next 10 years while discouraging high frequency trading. This is a very similar bill that was introduced by Sen. Brian Schatz in the Senate this past March. There are 25 countries, including major markets such as the UK, Hong Kong, and India, that currently have a financial transactions tax and several more considering imposing one. A study done in Italy found that a securities transaction tax “reduced stock market liquidity and increased volatility”.
Economic Nobel Peace Prize Winner Joseph Stiglitz claims that it “is likely to increase the overall efficiency of the economy and may actually enhance the efficiency with which the stock market performs its most important roles.” In his paper, he says “There are four circumstances under which governments frequently resort to selective taxes: (1) the commodity being taxed has a highly inelastic demand, so that the tax has little distortionary effect; (2) the commodity being taxed is a luxury good, consumed largely by the very rich, and not much weight is accordingly attached to the reduction in its consumption (perfume falls into this category); (3) the commodity being taxed is associated with certain benefits provided by the government (a benefit tax)—gasoline and airport taxes fall within this category; and (4) the commodity being taxed has some socially undesirable characteristics, such as giving rise to a negative externality.” The stock market meets all of these criteria, except number 3 which could be up for debate. One of the main reasons that economists are generally opposed to selective taxes is because they cause distortions in the market. With all these conditions present for the stock market, however, there will be a minimal effect. No matter what, the pie (amount of returns in the market) will always be the same. The only difference is how it is allocated between everyone. He goes on to say that there may actually be less volatility due to the fact that there will be thinner markets and fewer people selling faster once they start to lose money. This is one of the main reasons the government today wants to implement a financial transaction tax and prevent flash crashes.
High frequency trading is a major component of all trading volume done in the US. Although it is not illegal, some people believe that it is another way for corporations to beat the average American. Today, it is just as important for financial institutions to have people that can program trading algorithms as it is for them to have people that can trade based on the fundamentals of a security. Whether a tax bill is passed will likely not make much of a difference in the overall volume of trades completed daily due to the markets inelasticity. On Apil 4, the Senate will vote on Schatz’s bill and we will have a better understanding of the future for high frequency trading in the United States.