Forum Replies Created
- Posts
Is there a statistical way to measure objectively the degree of curve fitting of any model? Can one grade from 0-10, objectively, how curve fitted a model is? I understand that the more historical data the better; the more variables introduced into the model, the more curve-fitted. But I have yet to read about an objective statistical way to measure curve fitting. In the case of all your models, which show great simulated historical results, to what degree can we know that those results will hold in the future?
Thanks,
MichelThank you for your prompt response. I would like to try utilizing futures instead of ETF’s as you suggest. However, I am having trouble understanding how many contracts to buy or sell.
Assuming that I want to trade $1MM on this strategy. Furthermore assuming 30% ZIV / 70% EDV and the June VIX contract at 19.50 and the March UB contract at 172, how many contracts should I short/buy of each?
Should I match the expiration of the UB contract with the VIX contract and buy June UB instead of March UB?
Should I switch contracts for the VIX futures once a month in order to keep a constant 5 month period to expiration?
Lastly, the back testing of this strategy does not extend the the “acid test” period of 2008 when the VIX reached 90. Do you have any idea of how the strategy would have performed then and what the drawdown would have been?My concern about this strategy is that it cannot be traded with large amounts of money. The daily volume of the ZIV, VXZ and EDV is quite low. The bid-ask spread is very high at most times. This results in execution costs that are unacceptably high if one trades, say, $1MM or more on this strategy (TMV is better as far as volume)
Have you tried studies which utilize ETF’s that have higher volume characteristics such as for instance the SVXY instead of the ZIV for instance. I know that SXXY is more volatile than ZIV but that could be compensated through the allocation.
My goal is to be able to eventually trade $3-$5MM per strategy. Do you feel that that is realistic? What are the strategies in your arsenal that are best suited to larger amounts of money?
