Why not? All good ideas. But if you have the software to test the model, that is the way to go. Otherwise, models are often beautiful and make sense, but they don't always correspond to reality.
Yes, one of the problems that i've found recurrent in spreads is the difference of liquidity between the legs. Sometimes, I find good results based on close prices but it is only because one of the leg is more liquid and close price is more accurate.
Otherwise, I find that hybrid systems are more unstable than the simple original method( mean reversion ). The more you create conditions, the worst it is. Testing volatility adjusted spread systems may imply more numerical variables and i don't like it at all.
If i find something interesting, i DON'T post it
( just a joke... )
Yes, one of the problems that i've found recurrent in spreads is the difference of liquidity between the legs. Sometimes, I find good results based on close prices but it is only because one of the leg is more liquid and close price is more accurate.
Otherwise, I find that hybrid systems are more unstable than the simple original method( mean reversion ). The more you create conditions, the worst it is. Testing volatility adjusted spread systems may imply more numerical variables and i don't like it at all.
If i find something interesting, i DON'T post it
( just a joke... )