|
PureBytes Links
Trading Reference Links
|
Richard,
> I concur that the results do not agree between AB and MS
>computations.
yes they are #
> I would certainly expect Tomasz to use the conventional TR. I
>looked at the results on !RUT and !SPX - which have almost NO gaps,
>i.e. the range always includes the close, hence the TR = H-L almost
>everywhere and the results also are inconsistent.
I think its TR is
TR=Max(Max(H-Ref(C,-1),Ref(C,-1)-L),H-L);
Graph2 = ATR(1)-TR;
> Tis a puzzlement. At any rate, the random walk stuff is simply a
comparison of price vs. sqrt(n)*delta where all the play comes from
the choice of the random step size, delta.
>
yes and Tomasz don't use the atr(1) or TR above
Stephane
|