Equity Spread Trading Hedge Ratios

Not True. For two stocks A and B

Hedging Ratio = Beta(A,B) = Cov(A,B)/Var(B)

Correlation = Cov(A,B)/Std(A)Std(B)

If Correlation = 1, then

Cov(A,B) = Std(A)Std(B)

Then, substitute this back into the hedging ratio calculation, you get

Beta(A,B) = Std(A)Std(B)/Var(B) = Std(A)/Std(B) Since Var(B) = Std(B)*Std(B).

So, under 100% correlation, the correlate hedging ratio is the volatility ratio.

The correct hedging ratio for 0% correction... is 0! If the pair isn't correlated, it's not a pair trade.
More...

I am not a 100% sure but that doesn't seem right. The hedging ratio is the ratio of the price differences within a time period; it is not beta i.e correlation. Those price differences are in turn determined by anything from volatility to correlation to an asteroid attack etc.

Hence, why sjfan and rodmike9 are talking two different languages.
 
Why neutralize well behaved , repetitively mean reverting spread with a volatility or dollar hedge ratio
that causes the spread to diverge, or, dampens the variability of the excusions about the mean?
The chart of the neutralized spread does not look good, test or measure good. The non neutral does.

One step further ...what if a curve fit hedge ratio has been behaving best for MR spread trades for a period of time you deem significant (similar to trend acceptance filter) ? No garauntees of futher performance but
the curve fit value has been providing the best cointegration. Why not use it.
Seems the simple retail amateur like me will look at a raw, simple, well behaved spread and then do something to fix it even though it isnt broke.
 
You might notice that sjfan's post was made over 5 years ago. And the Like feature only became available with the recent software upgrade. That's OK, this is still a timely and interesting topic! Post on!
 
Last edited:
You must log in or register to reply here.
Back
Top