Trader Terminator

in Investments, Technology

Open the Pod Bay Doors HAL

The lovers of Science Fiction already know the perils of allowing machines to do too much in our world (i.e. Skynet and the Terminator series). Given these concerns, I continue to hear more and more cool ways that automated systems are making our lives better (and will improve overall lifestyle). Beware the concept of the Trader Terminator.

Currently many trading systems have automated heuristics out there to deal with situations on the Equities markets. A lot of these are simple rules like, if BCM (a strong equity) goes below $12/share sell or if BCM trends upwards buy (as bad examples). Thanks to very smart programmers, these rules and systems have (luckily) become much more complex, and with complexity comes opportunities. I have worked on early “artificial intelligence” systems, but the heuristics these days are extraordinary (in my simple understanding of them).

Trader Terminator

Eniac was the greatest piece of technology of its era, now my digital watch has more computing power than this marvel of technology. Technology always evolves.

The interesting question is, even with really advanced “thinking” systems, will it make things better for me (the individual investor), us (the collective that holds a mutual fund) or we (all investors in general)?

What if I design an automated trading system (call it the XROB2013 (or by its nickname the Trader Terminator) ) which uses as many cool and exciting algorithms and ideas as possible? The Trader Terminator could be an exciting and new way of doing things, if (as the designer) I chose the right algorithms and heuristics. It would be easy to test the Trader Terminator against all the old data from the market to see how the trading system might do. The problem is this is only a test environment.

XROB2013 The Next Big Fund ?

I launch my XROB2013 Mutual Fund (with a spiffy ad campaign about the Trader Terminator being the last mutual fund you might need), and it uses these algorithms to buy and sell the best stocks to hold (at the time) and for argument sake, we’ll say the axis we worry about is growth (i.e. we’ll ignore dividends and such). If my XROB2013 is successful, suddenly  many folks want to buy shares in the Mutual Fund and the Fund Maxes out and might be declared “best in show” for a while (is best in show a real investing fund award?).

What happens next? Looking into my technology history books, there is no such thing as a secret and since Patent’ing a trading heuristic isn’t likely, there will be “rip offs” of XROB2013 that might appear (that maybe have lower MERs or such, to make them more attractive (it would be cheaper because there was no R&D, it was simply a rip off)). Suddenly XROB2013 loses its lustre for a lot of investors. Worse, the algorithms now are working against each other creating an orthogonal change in trading ideas, so the XROB2013 “smarts” would have to be changed or there might be further erosion of the value of the Mutual Fund.

Technology for the Sake of Technology ?

All of this to say, that no matter how great an automated trading/investing system anyone puts together, the market ends up finding ways to lower the impact of any new idea in a very short time and going back to a relative level of stasis. Don’t be enticed by claims of “the next big thing” in market trading, unless you are sure it’s going to work.

{ 1 comment }

  • Matt December 18, 2013, 3:48 PM

    Love the example, any system like this would be risk prone because even if you programmed it with all the best known rules you could find there are still unknowns such as natural disasters that it wouldn’t be able to properly factor in without human help.

    The one piece to this idea you brought up is regardless of the systems humans can still beat them (yes I am going to bring up Warren Buffett).

    But if you were to take this line of thinking for an automated system and instead of tyring to get the best returns possible settled in on a more modest but more guarenteed rate of return. I think this is where your XROB could do well – say XROB2013 – 5%.

    Reply

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: