Wednesday, August 14, 2013

Multi-targets



Optimizing exits is something I've been working on for a while and its one of the hardest things to do in trading. There is a conflict between a clean exit and catching a large run.

For example, in today's first trade, I had a MFE of 1.2 and the second had an MFE of 1.5 for three contracts. Catching the top tick is magical and no sane trader should aspire to do it every time. However, if I had chosen to exit all three contracts on both trades today, my take would be .20*6 = 1.2

Another potential optimization is to exit on the close or beyond the first bar that does not have a strong close. In today's example, that would net me .90 + 1.8 = 2.7

Translated to per contract, it would take my winning from just under .30/contract to about .90/contract almost 3x better.

However, this is a reasonable thing to do only when you expect a trading range type behavior. On days when a trend is obvious -- which is quite hard before it sets up -- it may be better to hold on for the large move.

In the coming days you will see me move to a hybrid exit instead of fixed exits for 2nd and 3rd contract.

3 comments:

  1. Hi, Cad-
    hope your CL work is going well--seems to be.
    can you elaborate a bit on when to take profits from a second (runner) contract? I recall you talking of exit at second attempt against the position. Is that a second lower Low, and L2 or simple a second bar of the other color? Thanks.

    ReplyDelete
    Replies
    1. Exiting on the push of the second leg is a sensible thing to do on a trading range type day. However, on trend moves, you can usually hold for more.

      Exiting on the 3rd push on the other hand, is often a good exit since it typically means a prolonged pullback or trend termination. For example, if you bought b28 on this day, b36 would be a third push and its a good exit if it went far enough (i.e. within recent range of day)

      One of the hard things about exits is that you need to adjust expectations according to recent range, current trend strength, amount of move already underway, etc. This takes a bit of experience, but I'll try to break it down into simple rules where possible

      Delete