Those who really command on Wall Street

Algorithms are silently leading Wall Street

Algorithmic engineers are the first choises of investment companies. Hedge Funds like Renaissance Technologies, Citadel and Tgs Management are fighting to get the best ones working on algorithms in order to automatically trade the markets. These people can set up rules aimed at enhancing calculation and resolution of problems, in order to quickly analyze data and decide what to buy and sell in the investment world, with minimum human involvement.

On Wall Street, algorithmic and quantitative trading, based on sophisticated statistical models to structure interesting operations, are conquering everyone. In many trading floors quantitative methods are spreading, and investment companies are competing to hire mathematicians. On the other hand, traditional strategies are losing ground, involving the study of budgets and exchanges of ideas with corporate clients.

The Growth of Quant Funds

«A decade ago, the brightest graduates wanted to become traders in Wall Street investment banks, now they make fake cards to enter the quant funds», reports Anthony Lawler, head of Gam's quantitative investments. Last year the Swiss manager acquired the British quantum firm Cantab Capital Partners, which uses sophisticated mathematical models for trading strategies, for 217 million dollars. Guggenheim Partners built a $ 1 million «supercomputer cluster» at the Lawrence Berkeley National Laboratory in California, reports senior managing director Marcos Lopez de Prado. Algorithmic trading has existed for a long time. In a 1974 article, pioneer Ed Thorp peered into the columns of the Wall Street Journal. In 1988 the Wsj reported that a Chicago-based options trading company had a secret information system. In 2010, journalist Scott Patterson wrote a besteller on the rise of quantitative methods. The forerunners imagined that traders following algorithms rather than instinct would one day become the kings of Wall Street.

That day has come. Quantitative funds are now responsible for 27% of all stock exchanges in the US, compared to 14% in 2013. And they have almost reached small investors, which represent 29% (Tabb Group data). At the end of the first quarter, these hedge funds held $ 932 billion of investments, more than 30% of total assets in hedge funds (Hfr data). In 2009, quant funds amounted to $ 408 billion, 25% of hedge fund assets. In the first quarter the new net investments reached 4.6 billion dollars.

Machine vs manual trading

Computers are doing better than humans. In the last five years, quant funds have gained an average of 5.1% per year. During the same period, the hedge sector improved by an average of 4.3%. And in the first quarter the quants have increased by 3%, against 2.5% of the average hedge fund. The phenomenon is helped by two forces. The controls have made it difficult for investors to get an advantage by prodding top managers for information and investors are now having an ocean of data at their fingertips.

The next frontier? Intercepting data with drones and other avant-garde means to understand the trend of companies and the economy in real time. Quantities differ from high-frequency traders, who tend to focus on operations lasting just a few milliseconds, and have experienced a decline in volatility and increased competition. ETFs also use algorithms, but they are more oriented towards investors who prefer to expose themselves to specific sectors. Quantitative trade can last from a few minutes to a few months. Some analysts fear that businesses and investors tempted by the new trend remain disappointed. The most successful quantos have been operating for years. And hiring research doctors is not a guarantee of profits.

Increased competition could jeopardize returns and give a deceptive feeling of security on stability. In 2007 what is remembered today as quant meltdown was caused by the similarity of the strategies, which induced several quants to sell all together. William Byers, who wrote How Mathematicians Think in 2010, warns that translating the world into numbers can give the misleading belief that predictions from computers are more reliable than they really are. The more investors resort to complicated models, the more likely they will look like one, fuelling serious market disruptions. So far nothing has stopped the quantitative arms race, which is creating a large number of jobs in the financial industry.

In August, Balyasny Asset Management hired researcher Gilbert Haddad, a former Schlumberger and General Electric, who studied nanoparticles at the University of Wisconsin and has a Ph.D. in engineering. The billionaire investment company of Steven A. Cohen, Point72 Asset Management, is converting half of the managers to a "more man-machine" approach. The old school teams work alongside data experts and financial analysts take evening classes to learn the fundamentals of big data science. Point72 has invested tens of millions of dollars to analyze data avalanches, including credit card receipts and pedestrian traffic. According to rumors, last year the group lost most of the traditional strategies, while the quantum amounted to about 500 million dollars. Chief market-intelligence officer Matthew Granade urged London School of Economics students to study programming languages to be more competitive.

Even big investors fall against algorithms

One of the most famous investors in history, Paul Tudor Jones, who predicted the collapse of Wall Street in 1987, made huge profits with rapid movements and an average annual earnings of over 17%, yet he hardly produced anything in 2014 and 2015. According to people close to him, until last year Jones felt the pressure of the growing success of quantum. In October he chose Dario Villani, an Italian Ph. D. in theoretical physics, hired in 2015 to give a shock to Tudor Investment. In Greenwich, Connecticut, Villani develops programs that replicate the positions of managers through tools that increase the risk to improve returns without threatening the stability of the company.

Man has always sought to have advantages in information. Baron Rothschild in 1815 organized a network of field agents and travelers pigeons to know in advance the outcome of the Battle of Waterloo. Today, quants hope to act faster than traditional investors. At the end of the 90s an algorithm could try to ride the price of a stock, buying at a certain level and selling at a given time. Today they update forecasts based on the analysis of past and present data, bombarded by hundreds of inputs in real time. Sooner or later they will affect the need for staff in large investment companies.