Hi,

I have been using Metatrader4 for many years now but want to move to software where I can do Portfolio level testing.

I read this article by Andreas Clenow: https://www.followingthetrend.com/2014/05/why-i-prefer-rightedge-for-strategy-modeling/

Under the heading of PORTFOLIO MODELING he explains how wealthlab does its portfolio testing and why he thinks it is done wrongly. (Compared to the RightEdge method.)

This article was written in 2014. I just want to make sure if you still use the same logic for your pertfolio testing as mentioned in the article, or whether you have updated this in the meantime?

Thank you.

Hi and welcome to the forum.

It has not changed in Version 6 and Wealth-Lab's backtesting engine is not event-driven. It still processes the data symbol by symbol, bar by bar in Raw Profit mode and then applies the position sizing overlay to arrive at a portfolio simulation. (The RE method is what somebody called date-sequenced backtesting.) WL's method is by no means "wrong" but practically speaking, it's a tradeoff good enough to get the job done.

P.S. I always felt that this article is somewhat biased and probably have an idea why (but won't tell as it's private matter). Indeed RE had the advantage of starting many years after WL (which pioneereed the portfolio backtesting for retail investors). But many of the concerns are questionable or a matter of taste. Like getting position sizing in a single line of code vs. DLLs (it'd be funny to see what kind of mess would the resulting code of an advanced position sizing technique look like). Or peeking into the future because when one learns how to code WealthScript (C#) he finds out how to avoid this pitfall pretty quickly. With scalability he may have the point but WL has never been positioned as the paltform of choice for global money managers.