Milktrader

Iterating Until Convergence

Tuesday, February 16, 2010

My Trading Mask

The marriage metaphor is getting a little stale for me when it comes to trading. You don't want to get married to your trade, but you are committed to your trading plan (in a married sort of way). When you deviate from your trade plan, your infidelity is a burden for you to bear alone when you go to bed at night. I'm leaving personal relationship analogies behind for now. Instead, I'm introducing a new style of trading: the Janus style. Janus is a euphemism for two-faced and suggests some unscrupulous character flaw, but remember that many famous people have lived double lives and still contributed to the culture in which they lived. Think Superman. Super hero when duty called, but otherwise a news reporter. Or how about Spiderman ( Peter the photographer) and Batman ( Bruce the philanthropist). I can't remember what Catwoman's day job was, but I'm pretty sure it wasn't running around town in a latex outfit.  Maybe Catwoman is not a good example, but you get the idea. Analogy applied to trading? I am continuing my pursuit of black box system trading by day, but I am assuming my own alter-ego and becoming a discretionary trader by night.

The problem with system trading is that it takes a while to develop properly. And quite frankly I'm running thin on patience. At times, you wonder if you will ever trade again. You see, you cannot deploy a trading strategy until you've properly tested it, and nobody can define what proper testing is since it's literally impossible to accomplish. Besides getting stuck in trading purgatory, it's also easy to get turned off by the mathematical nature of system trading. Trading involves people, not physical constants, so there is a little trouble in reconciling the math of systems trading with the psychology of markets.

I'm not abandoning the system trading voyage by any means. I'm just getting warmed up with intermarket filters to generic market system trading which will be based on Bayesian frameworks. There is much work to be done in the field of machine learning, not the least of which include kernel regression and radial basis functions. Really, I'm just using this handy new metaphor to justify to myself the decision to re-enter the realm of discretionary trading. I'm not going to say that discretionary trading will complement my system trading development and vice versa, because that's a little thin. Clark Kent was not a better reporter because he wore a cape on occasion. No, the two are actually quite antithetical.

Take the role of human emotions as an example. The emotionless system trader executes trades flawlessly, with great precision and confidence, fully aware of his expected winning percentage, profit expectancy and drawdowns. When his signal isn't present, he doesn't trade but instead does crossword puzzles or some such thing. Fear and greed is neutered. Not so for the discretionary trader. I'm expecting sweaty palms and feelings of dread and anxiety. And of course, I fully expect the thrill of overconfidence when I got a winner. I'm adopting no trading rules and my trading plan is to make money. I know that violates many trading sensibilities, but what good is an alter-ego without a little excitement?

My discretionary weapons of choice will include spot forex, futures and options. I'm not going to be a complete idiot and go all in on one bet, but I also intend to mock well-established axioms along the way. I'm going to tweet and document all my trades, which will be conducted with a real money account of $20,000. Yes, small change to some, but this is an experiment and if I fail miserably, I don't want to deal with anything more than not paying the credit card company from which I plan on getting the money (kidding about the credit card).

Janus is not only a two-faced Roman god, but is also identified with gates and doors, beginnings and endings. And though it has never been proven, it is said that he liked to have some cookies with his milk.

4 comments:

  1. I know exactly how you feel! I have been there before - and regretted it. I believe I am a bad/emotional discretioanry trader (one of the main reason I wanted to go and develop automated trading systems).

    All the discretionary trading has done is take my attention away from my development efforts, given me lots of stress. And even if I am net positive, I believe life might have been better without it.

    Now I am full steam on system trading development and that feels great! But good luck with your decision - am just sharing my experience.

    PS: how about Tiger Woods for a real-world "double-life" character: Golf champion during the day / [...fill the blanks...] at night? ;-)
    ReplyDelete
  2. Jez,

    When do you see yourself deploying your first system trade with real money? We both know that you could go out tomorrow and buy a system and make your first trade tomorrow night after you load the software, but we are both building this from the ground up probably because we need to know why it works and why it doesn't, etc.

    I see my first system trade with real money materializing 3Q this year. I think that is a realistic goal.
    ReplyDelete
  3. Hi guys, I'm a little further behind you in the path of system development, but I'm experiencing the same problems. Its great to hear of other people in the same boat; I'm also starting to wonder when the testing will ever finish. Its getting quite hard to keep the day job going and also devote time to development at night / weekend. Starting to feel a bit burned out. And the wife isnt too happy about the situation either!

    But I'm hoping there a pot of gold at the end of this rainbow, and I find the slog reassuring in a way - it probably keep most people from being successful at this game. So there is more opportunity for those willing to put the work in.

    On another happier note, in my day job I often get to evaluate institutional hedge funds, some of which are black / glass box in nature. I'm often surprised by their relative lack of sophistication, and still reasonable results. So it is possible to make money at mechanical trading; the slog will hopefully be worth it in the end.

    Ian
    ReplyDelete
  4. You can take an average trading system and realize decent results with money management. As long as you have positive expectancy, of course. The area you may want to look into is called Optimal f.
    ReplyDelete