Algorithmic trading

Algorithmic Trading is a technical name for the high speed trading of stocks with the help of fast computers, ultra-fast networking and sophisticated algorithms. High-Frequency Trading (HFT) is another term, which might better explain the concept. High-frequency trading systems buy and sell stocks in the tiniest fraction of a second. It takes humans several seconds to make decisions or to receive orders to buy or sell stocks. Considering also the time to type the stock name, the amount, the price and the conditions of the transaction, one can imagine that trading by humans cannot be faster than a few seconds per transaction. A computer programmed to make decisions about a purchase or sale and performing the mundane tasks for the transaction will be at least a thousand (in some cases, a million) times faster than a human. HFT probably started sometime in 1999. It is a little known subject outside the financial world. On Wall Street the methods are kept secret. Apparently The New York Times published the very first article on it in 2009. Most HFT seems to create shortterm positions; due to the advantage of speed, the computer algorithms buy or sell stocks, and a few milliseconds later reverse their positions. It is obvious that trading companies noticed and capitalized on properties and benefits thereof very early on. Before I proceed I should add this: According to a study done in 2009, more than half of U.S. equity volume is estimated to be processed by HFT firms. This should give you a wakeup call. We are talking about billions of dollars of U.S. securities bought, sold, leased, borrowed and shorted in a day using computerized techniques, not involving humans at all, except as programmers who design the algorithms and produce the software. Is it legal? Good question. For now, I could say that it is not illegal. Unless a judge, armed by law, orders the practice to stop, it will go on.

But the bells started ringing. Several European markets pr oposed to curtail or ban HFT. The U.S. Federal Bureau of Investigation announced an investigation of HFT in April 2014. HFT is not just about the speed of their networks and computers against the slowness of humans. The decisions are made and executed within microseconds. Human speed or dexterity is no match here; it is usually one set of computers (and algorithms) against another, and it contains significant risks. HFT is known to contribute to volatility; for example, HFT is seen as one of the reasons for the May 6, 2010 crash. Moreover, HFT is creating an unfair advantage against investors by trading (a few milliseconds) ahead of investors on index funds for a fraction of a cent in profits in every trade. The critics say that HFT practice has created a two-tiered market in which certain classes of traders can unfairly exploit others. In addition to the fastest computers and most sophisticated decision making algorithms, HFT firms moved closer to market buildings and secretly laid out new fiber cables underground to gain extra micro seconds!

The strongest critics say the market is rigged by HFT traders who "front run" orders placed by investors. Front running is an illegal practice; it refers to a stockbroker executing orders for his own benefit, knowing in advance the market orders of his own customers. One can see how HFT would be a perfect tool for this; within a millisecond, the broker will buy and a sell, leaving less profit to his own customers.
One of these critics is Michael Lewis, a financial journalist who has been covering market practices since the early 1990s in his articles and books. His newest book, "Flash Boys," tells the stories of several people involved in HFT algorithms, including the Russian chess player and programmer Sergey Aleynikov, who was working on developing algorithmic trading techniques at Goldman Sachs.

Currently, there are several investigations of HFT in New York and elsewhere. The least one can say is that the legality of many HFT practices is seriously being questioned. Algorithmic trading is an area of research and application overlapping both the investment world and computer science; it is an intriguing subject. The participants are generally secretive about the algorithms and methods they use or discover. It involves many strategies. The programmers need to understand the rules of the markets well and write programs that mimic human decision-making, except reacting to events within microseconds. It is yet another area where we are able to inject automated reasoning that was initially reserved for humans only.