In his Trading Strategies/Alpha Power post, Roland describes a spreadsheet which simulates 1000 weeks of trading for 50 synthetic "stocks" -- all of which are entirely generated by random numbers. That is, there is absolutely no trading advantage that can be gained through analysis of market sentiment, momentum, earnings, etc. This is an ideal data set to isolate and study position-management as an investment strategy in and of itself. His results -- though shrouded in secrecy! -- are compelling.

I assume he does his analysis of this "market data" in WLP by saving the data as ASCII files, and then creating a data set within WL that reads them.

I haven't been able to find any resource in the WL universe that explains the ASCII file structure of a static data source. Any help?

It strikes me that one way to accomplish something akin to this would be to export random 1000-bar data sets from actual stocks, picked at random -- and then mixmaster the values in a spreadsheet to randomize the data stream for each stock.

Is there a way to export a data set for an individual stock?

QUOTE:
... the ASCII file structure of a static data source. Any help?

Check the User Guide; ASCII topic.

QUOTE:
Is there a way to export a data set for an individual stock?

Search for "export and ascii". For sure you'll find dozens of topics with code.

Also in WL5Wiki: Exporting to ASCII comma-separated files (CSV)

Thanks Eugene, I forgot about that nice Wiki of ours ;)

Last comment about ASCII format that perhaps shouldn't go without saying:
The format (date, decimal characters, field delimiter, etc.) can be just about anything, but you have to properly specify these when creating the ASCII DataSet in Wealth-Lab.

Eugene's wiki href was on the money. Thanks, guys.