How Small Traders Win in the Era of Algos

While Renaissance Technologies makes most other hedge funds look foolish, even the independent trader can copy their discipline   NFLX, PANW   Long-time followers of mine in this bull market know that every quarter I go over the holdings of hedge fund Renaissance Technologies because they were one of the early quant houses that made algo trading so successful and popular.   The founder, Jim Simons, was a mathematics professor in the 1970s who never thought about the markets much. He sort of stumbled into testing some theories on stocks, and the rest is history as he was pulled headlong into the markets and created a powerhouse with over $50 billion AUM (assets under management).   And Simons made a point of not hiring MBAs, traders, or anyone with a background in finance. He only wanted physicists, engineers and other quantitative problem-solvers to come work for him and mine the data of markets to find unique correlations, patterns and new edges.   What kind of data patterns and correlations are they after? Well, with 90 Ph.Ds. on staff, mostly math, science, and engineering types, we can only guess that they are sifting through mountains of fundamental, price, economic, weather, and consumer patterns, looking for those small anomalies between individual stocks and industries and other asset classes.   Mining Data Others Ignore   If it’s a popular, well-known correlation, they don't want anything to do with it. They hunt in the noise of tons of data for things that others can't see, or are not even looking for.   This week, Matt Levine at Bloomberg View wrote briefly about Renaissance after colleague Katherine Burton published a full story on the company and its funds and practices for the December/January issue of Bloomberg Markets magazine. Here's how Levin opens his piece, quoting data from Burton's story...   The big problem with Renaissance Technologies, the Long Island-based "pinnacle of quant investing" founded by Jim Simons, is that its Medallion fund makes too much money.   Medallion was up 21 percent for the first six months of 2016. It was up 35.6 percent last year, 39.2 percent the year before, 46.9 percent the year before that. This keeps going. The last down year was 1989. The fund had a rough few days in August 2007, but ended the year up 85.9 percent. It has returned about 40 percent per year, on average, net of fees, since it started in 1988.   Of course it can't keep compounding returns that way because of the size factor. What you can do well with a 5 billion dollars you can't necessarily as well, much less better, with 50 billion. And that's why they are forced to simply return profits to investors, who are primarily employees now since the fund was closed to new investors in 2005.   Matt Levine's piece on Bloomberg View can be found here and the full Burton story is linked above.   Can a Human Trader Copy Black Box Success?   While the computer programs that work for Renaissance are still a big mystery -- like we don't know how many strategies just trade intra-day to make money -- it's safe to say that they create a lot of turnover in stocks, exploiting new patterns or "edges" in thousands of stocks.   But doesn't that mean a lot of extra risk?   What most people miss about the success of the algos and black box trading systems is that they run through markets with big size in thousands of stocks and instruments because the risk control is automated too.   It’s not like you or I trading 100 stocks at once and going crazy trying to keep track of the risk and profits. They program the computer models to seek and destroy profit opportunities and to manage the risks in real-time too so that they are never destroyed.   So speed and continuous, instant access make the difference too, especially if the model is wrong about an opportunity. In that way, they take human emotion completely out of the decision-making equation.   With that I want to introduce you to our guest today, who is like David to the Goliath Renaissance. Jeremy Mullin is a colleague of mine at Zacks where he starts with a simple quant model built on earnings momentum – the Zacks Rank -- and then overlays his own suite of technical trading filters and what I will broadly call “behavioral analysis” because he pays attention to extreme moves in stocks that are often driven by algo trading that is exploiting investor fear and greed in the markets, which therefore sets up new opportunities for him.   Jeremy has spent the last 13 years as an equity, futures and options trader. His main focus when trading stocks is high beta equities and earnings moves. He uses technical tools when entering and exiting trading setups, but also watches order flow to get a “feel” for market direction.   Check out the Mind Over Money podcast, episode 4, to hear my interview with Jeremy.   Kevin Cook is a Senior Stock Strategist with Zacks Investment Research where he runs the Tactical Trader service.