OptionsSimulate.c produces different .t8 than archive download

I'm making good progress with an options trading algo based on the Financial Hacker scripts.

However, my results are very different between the downloaded SPYa.t8 file and the SPY_SimOptions.t8 file that was generated by the OptionsSimulate.c script. I thought the downloaded file was also generated by the same script.

Any thoughts? jcl?

Thanks.

Are both files from the same years?

Thanks for the response.

The download description on: http://zorro-trader.com/download.php, says this.

"D1 SPY Options 2011..2017 - Artificial option chain history (.t8 format, 30 MB)"

So, I assume it contains options data from 2011-2017, inclusive. Based on simulations results, it appears there are options chain data when requested by contractUpdate().

The OptionsSimulate.c script has these variables:
StartDate = 2011;
EndDate = 2017;

So, I assume it will generate options data from 2011-2017, as described in the Financial Hacker blog. The only difference as I understand it is that jcl provided a SPY.t8 download as an alternative to generating the .t8 using the script, which is time consuming.

http://www.financial-hacker.com/algorithmic-options-trading/#more-2198

You can use the History.c script for comparing the files and checking the date range. The article was from a year ago, so the SPYa end date is certainly different to your data.

Thanks for jumping in, jcl.

Where do I get History.c? it's not in my Zorro/Strategy folder, or any of the script archives at FH.

I just changed the StartDate and EndDate to [2012-2015] in the Zorro OptionsSimple.c script and the results were different between the downloaded SPYa and the newly generated SPY_SimOptions. The [2012-2015] is in the middle of the time range, so the date should be the same - right? Unless the SPY.t6 that was downloaded now is some how different than when you created SPYa last year?

Or, could it be that I'm using a newer version of R (3.4.3) than when you generated SPYa a year ago?

Thanks!

Without digging in, I will point out that dividend-adjusted history changes for every time that there is a dividend payout.

Check this out:
https://www.nasdaq.com/symbol/spy/dividend-history

Your dividend-adjusted history will be completely different every three months!

Assuming the original script used dividend adjusted data, it should be a bit tricky to recreate the synthetic data exactly.

Finally, it is not impossible for different data vendors to have different formulas for calculating dividend-adjusted history.

assetHistory is called with the UNADJUSTED flag, as pointed by the FH blog, when playing with options.

Aha, it is unadjusted.

So which service provider is your data source for unadjusted data? And is this data identical to the Yahoo data that is no longer available? Do we even have a way to confirm this?

By the way, the history script is under:
Strategy/History.c

It's available in the latest beta.

Finally, check out the Z history editor, which lets you investigate t6 files. (Last time I checked, not t8.) Get it here:
http://zorro-project.com/download.php

I heard Yahoo data was shutdown,so I changed the script to use STOOQ;

assetHistory("SPY.US",FROM_STOOQ|UNADJUSTED);

I used CVStoHistory.c on the 2 .t6 files, copy and pasted in Excel, aligned date rows and saw that the datasets were pretty much the same - off by 0.01 here and there.

Wish there was a way to inspect the .t8 files.