Quote from bwolinsky:
Is there something wrong with showing a profitable backtest? You have the program, and yet you doubt I'm not getting those results?
Quote from Index Maximus:
I'm questioning how you can incorporate data which does not exist, i.e., volume data on something which does not trade and therefore generates no volume. If it is a key component of your backtest and it is faulty, then the validity of your test is suspect. Don't you agree?
Thanks for that.Quote from bwolinsky:
Listen to me, quit using E-signal as your data provider. Please switch to Fidelity's data. Create another datasource with just .SPX 15 minute Fidelity data, k?
LOL ... you make way to many assumptions.Quote from bwolinsky:
On your point, I've received several thank yous for this piece, so I can only assume that they were able to duplicate it.
Thanks for asking the pointed questions Index Maximus.Quote from Index Maximus:
I see your approach to this analysis is truly rigorous.

Quote from Wayne Gibbous:
There certainly could be data differences. But using an untradable index isn't getting us anywhere. There is volume in our SPX data, but what is it? Is it option contracts, or just the total volume from all SP500 stocks? I could find out, but it really doesn't matter. If you care to validate your results, sort of a "peer review", we can do it. It's easy:
- Post a test on a real security.
- Post just a few trades from anywhere in the last few months.
- Post a couple days of data you are using from the same period.
I did all of that in this thread already and it only took a few minutes.
Quote from MandelbrotSet:
...
Let's get something straight.Quote from bwolinsky:
Don't be such an ass, I did 99% of the work and you're expecting me to change what? My data? My data is what it is.
And I've also built programs in WL way longer than anybody here, save for Spydertrader, and a couple others I have worked extensively with on wl4.wealth-lab.com.
I don't use Fido. ! have been using Wealth-Lab for 7 years. Way before the Fido buyout.. Unlike you, I am not a newbie at this. You are hiding something. I don't know why and I don't care.Quote from bwolinsky:
I'm doing this as a service, Wayne, and I don't know owe you anything for it. I've posted the program which is 99% of the process. I posted what results I'm getting out of the program.
I've not denied there is not enough data, but if you're not going to use Fidelity data, unless you have a hacked copy which I suspect here or you would have done it by now, then you're not going to be able to verify the results.
You're also doubting my ability to perform a backtest on data that I didn't create, but Fidelity did. Anyway, I'm sure somebody out there has duplicated it by now.