|
PureBytes Links
Trading Reference Links
|
"Continuously test and count or quantify the accuracy " Sounds like a
system to me.
If you simply want to test these indicators as entries without system type
exits then just set up to exit after a certain number of bars that might
correspond to the time frame you have in mind for actual trading. If
barssinceentry = 5 for example. This is how we test our entry methods when we
begin to build a system. Anything less than 55% winners after n bars means
your entry is no better than random.
Don't be concerned about profitability, just evaluate the results based on
percentage of winners. The reason that you have to judge on percent winners
is that the rest of the performance results are entirely dependent on the
quality of your exits. Don't expect your percentage winners to hold up after
logical exits are added. Most trend following exits tend to cut losses and
hold winners and hurt the winning percentage but they escalate the
profitability.
Chuck LeBeau
traderclub.com
|