Hi,

I have a question about the so called "survivorship" bias.

When backtesting strategies I backtest only symbols which exist at the moment. So I do not backtest symbols which where delisted in the past. Does anybody knows how to get stockdata of delisted symbols?

There is a similar problem with for example the sp100. Which symbols where in the SP100 in 1999?

How do you solve these problems with wealthlab?

Thanks for your answers

Best Regars

Andres

Andres,

Pretty much hits come up for me when searching the forums for "delisted" or "survivorship""

Extensive historic stock data( listed and delisted) that works with WLpro?
Delisted stock historical data from www.csidata.com
Is there data that maintains older issues for delisted, bankrupt companies?
WARNING: S&P 100 Survivorship Bias

And here's a programmatic solution from Community Components:

Backtesting with Dynamic (Evolving) WatchLists

Hello Eugene,

thank you for your answers.


Andres

The Norgate Provider solves survivorship bias for Wealth-Lab backtests!

Here's an example of the 3x2 strategy with the 2-line code modification required to make it a historical test using Norgates S&P 100 past and present components DataSet. If you run this with Raw Profit mode to get all the trades, something interesting and unexpected happens that we're going have to write an article about.

Although only a 2-line modification is required, there's an extra routine to sell old positions when its series ends (merger, chapter 11, or whatever reason).

CODE:
Please log in to see this code.