Stops, from Duncan Coker
All the testing I have done on stops indicates they work exactly as you would expect. They reduce risk and they also reduce expected return. Sometime they reduce the return to breakeven or negative in which case its better off just not to trade. In other cases they eliminate bankruptcy risk and allow for some return, but not the expected return you get without stops. It seems to fit right in with EMT, CAPM and practical ideas on markets regarding risk and return. If you are selling disaster risk, you expect to be paid for it.
Ralph Vince writes:
Or, expressed in terms of their effect:
Total Return = (A^2 - V)^(Q/2) - 1
Where A = Average return per trade (expressed as a multiple, i.e. 1.0 + return)
V = Variance in the A's
Q = total number of A's
When stops reduce V more than they increase A^2, it pays to use them.. I suggest conducting tests on this, it is far more revealing (in terms of the distribution of the A's) then meets the (rolling) eye(s) when looking at the simple equation.