Does algorithmic trading work?

There has been a lot of discussion (such as on Reddit, Quora, Hacker News, and other communities) about quantitative/algorithmic trading:

Does algorithmic trading really work for individual traders? Isn’t it just another seemingly sophisticated method like technical analysis?

Does algorithmic trading really work? What are the potential gains and limitations?

What kind of return can an average algorithmic trading firm achieve today?

Algorithmic Trading: The Play-at-Home Version

Algorithmic Trading: A Brief Introduction

Does algorithmic trading work? I don’t know for sure, but I think a lot money is made in market making (Citadel Capital comes to mind), which tends to full under the umbrella of algorithmic trading – the two are closely related.

Algorithmic trading is often a full-time job. It involves a lot of trading and paying constant attention to order books, as well as tons of backtesting and data analysis. Because your competitors are constantly evolving, so must you, and staying ahead of the competition is very time consuming. Just because something is automated or algorithmic doesn’t mean you can be complacent.

Major firms such as Goldman will have vastly more resources (computing power, access to top talent) than a typical DIY quant trader. However a major advantage for small traders is that they don’t have the same liquidity constraints as large traders. Large traders move the market (and this may be undesirable and must be taken into account when placing trades), but smaller traders can slip through without having to worry about complications such as ‘market impact‘.

Being bigger comes with many perks. People believe that Warren Buffett, for example, trades the same stocks everyone else does (except he is really good at it), but this wrong to some degree. Buffett has access to special deals (such as preferred shares on Bank of America and Goldman Sachs that pay large dividends) that ordinary investors would never have access to.

Slippage and commission fees can adversely effect performance. It isn’t too uncommon for a system that is profitable without fees and slippage to be rendered unprofitable ones those two factors are taken into account.

Then there are a plethora of statistical biases that plague traders: look-ahead bias, data dredging/snooping biases, curve fitting, and so on.

Here is an example of how slippage can negatively affect the equity curve of a strategy:

Success may still be elusive. The literature shows that most (about 90-95%) traders fail. Although Renaissance (James Simons), Citadel (Kenneth C. Griffin), and Quantum (Soros) have had huge success, they are outliers and have access to special propitiatory ‘tools’ and markets that ordinary traders don’t have access to. It’s not like these quant firms are using a mom and pop discount brokerage firm as everyone else does – rather they are the firms, meaning that they often make money from ‘making markets’, not trading them with a directional bias (long or short) as most people do. By making markets and using direction-neutral trades, these firms can make money in pretty much all market conditions.

Most quant firms and traders use leverage, which is necessary to generate large returns from small underlying movements, but this can backfire in a big way as the 1998 implosion of Long Term Capital Management showed.

Overall, I don’t think quantitative trading is as glamorous as many think it is, and I’m not sure if the returns are worth the effort. The historical returns for the S&P 500 average about 7-8% a year (exuding dividends), which many funds, despite access to top talent and top trading tools, fail to beat. There are simpler methods, based on mathematics such as the ETF decay, that an also generate very good returns and don’t require full-time trading. A lot of caution must be taken in any strategy, and risk management is crucial.