Stops... in your trading systems...

Discussion in 'Risk Management' started by FlyingSoros, Jun 20, 2010.

  1. Hi all,

    I've read some threads about stops in trading systems...

    What stops do you favor? Trailing stop based on percentage off the peak, profit target stop, fixed stop, etc.?

    I am talking about systematic trading... Thank you!
     
  2. If you're planning on purely system trading, why not back-test all of the above? It's not like computer horsepower is exactly expensive these days.
     
  3. +2. There are a lot of variables, and testing will best fit the trader's habits (at least, in the past).
     
  4. Totally agreed. Test, test, test. And think creatively about how to structure stops mathematically. Don't fall into the lazy mental trap of thinking only of the kinds of stops brokers offer.
     
  5. Stops are bad.

    1) Stops are about you, not the market. The market doesn't care where you got involved. Therefore what you're stopping is yourself. You may find utility in that, but there by definition cannot be any inherent value in loss-cutting based on arbitrary entries.

    2) Let's say you backtest a certain way of stopping out positions and it improves your historical results. If what you've done is actually valid, you've actually just found a new system, with your stop as an entry and your "unstopped" exit as an exit.

    3) Stops don't work at all--they literally do not accomplish what they're intended for--in the worst of cases. Ask the people who were "smart" and had stop losses in on May 6.

    4) If you want to alter the distribution of a given trade's expectation *prospectively*, the only way is to use contingent claims, also known as options (for the purposes of this discussion). Think they're too expensive? You get what you pay for.
     
  6. Interesting to hear from someone else that thinks stops are bad. I trade quantified strategies and generally stops hurt my results, so I look to other methods to reduce my risk. It can be quite painful to trade this way but in the past I have found that by using stops I take more frequent losses. Now occasionally I take a large loss (along with less small losses) but overall I'm more profitable.

    I'm not against stops, they just don't work well with my strategies. for other strategies I can see why they would be essential.
     
  7. "Hurts my results" is relative. The occasional large loss also probably increases risk of catastrophic failure, such as on the flash crash previously. If you are not leveraged, no big deal. But if you are, your strategy just rearranges the risk continuum around your trading methods, it does not necessarily make it better. I think everyone should at least consider entering a hard disaster stop.
     
  8. Stops are about preservation.

    When you enter a long, you do so because "you believe the market will rise from your entry"... and your stop should be "the amount I'm willing to risk that my play is correct".

    Where to place a stop is both a guess and an art. Suggest stops be placed at some chart point that makes sense to you.

    Trading without stops is suicide in the making.
     
  9. I care about risk adjusted returns over the long term and I generally don't use leverage. In fact I generally have a significant allocation to cash as well as long and short positions (not enough shorts the week of the flash crash though).
     
  10. Fixed that for you.

    What I said about stops is fact, not opinion or conjecture.

    Look at it this way: a stop is an entry in the opposite direction of a given trade. If it helps you make money, or lose less, it has a positive expectancy. Why not just trade it as a separate, uncorrelated system?
     
    #10     Jun 23, 2010