Friday, May 27, 2011

Fri. 5/27

1:40pm CDT - Arghhhh. Gave back 4 days of gains. And I've basically been breakeven for the past 3 weeks now. Maybe the long weekend will give me some clarity on how to avoid days like this. For the week, net gain of $109.

Net breakdown (contracts traded):
ZS -$2722(60)
RESULTS FOR DAY
Contracts:60
Net $P/L:-2722
Wins:3
Losses:7
Win%:30
Avg$Win:404
Avg$Loss:-562

4 comments:

  1. I guess that your max daily loss should not exceed (or only marginally exceed) your daily goal of $750. If you are limiting the upside, you should cut the downside as well. The good thing is that you have you shown that you can make money on a regular basis. It is just a question of risk management now. Have a good w-e.

    ReplyDelete
  2. I don't believe in fixed limits, like "-$750". If your trade (T) has what you believe to be a positive expectation (E[T] > 0) per trade (your "edge"), you should execute every legitimate opportunity that arises. This is the only way you can make money over the long run. To put it in a poker analogy, if you're down $750 and the next pre-flop lands you pocket aces, are you going to fold? Hell no, bet the hell out of it and make other players pay up.

    You may find that E[T | Some Bad Scenario] is negative, but you need to statistically what exactly "Some Bad Scenario" is, and whether it's some mental quirk or some core aspect of your trading strategy.

    Putting fixed limits on losses is a viable strategy if and only if you are certain that you are mentally incapable of trading correctly after a bad run and need some recovery time to execute your edge correctly. To relate this back to poker, consider that you were dealt AA, but you are 95% certain the other guy flopped something that beats you, but you are so burned on having been in the hole -750, that you can't fold your hand because it's too painful -- then you may want to step back and take a break.

    Automated traders don't tend to have problems violating their rules. Models fail on occasion and we'll turn off models, but generally only when we think we're seeing some well quantified, statistically abnormal behavior. Bad losing days are inevitable, it's just a matter of truly understanding why and whether the bad day made sense in the context of what you've observed when you designed your strategy.

    ReplyDelete
  3. Thanks for comments. I agree with both you and have flipped back and forth many times in my trading years between taking every trade and having limits when to stop.

    ReplyDelete
  4. Isn't "arghhhh" a little piratey? How about using "augh" instead? Just a thought.

    -AT

    ReplyDelete