There is a common belief that very old data may not show similar pattern as of today's. In a quant's world is not very uncommon to say that algorithm which used to perform very well two years back are useless today. This in a way tells not to back test on too old data.
Whereas there is another school of thought which tells to backtest as much back as possible so as to capture as many uncertainties like say the May 6, 2010 flash crash. What do you guys feel ?
How far back is ideal for backtesting?
I have been back-testing as far back as the Quantopian data will support: early 2002. Unless you have reason to think that your strategy only applies to recent data, I'd recommend longer backtests. If anything, it should reduce the risk of over-fitting. My two cents...
To my mind backtest should be done on full market cycles.
Number of bull and bear legs should be equal.
As Quantopian data starts only from 2002 you have only two choices;
Bottom (March 2003) to Bottom (March 2009)
Top (Oct 2007) to now(July 2016).
Аs we do not know if the market pass the top in current cycle I start all my backtests slightly early -from Jan 2007.
As far back as possible but we have to keep in mind that the market environment of today is unlike anything seen in history. In particular, algorithmic trading has changed how the market operates. Now we have a bunch of algos with the memory of a gold fish driving up market prices because they just look at technical trends, and they are easily manipulated because of that.
I think you just have to accept that no algo is all-weather. My VIX Bot tries but I know there are holes in it that I can only hope doesn't happen.
Please sign in or
join Quantopian to
post a reply.
Need help?
Contact support.
Company
Learn & Support
Notices
Follow Us
The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by Quantopian.
In addition, the material offers no opinion with respect to the suitability of any security or specific investment. No information contained herein should be regarded as a suggestion to engage in or refrain from any investment-related course of action as none of Quantopian nor any of its affiliates is undertaking to provide investment advice, act as an adviser to any plan or entity subject to the Employee Retirement Income Security Act of 1974, as amended, individual retirement account or individual retirement annuity, or give advice in a fiduciary capacity with respect to the materials presented herein. If you are an individual retirement or other investor, contact your financial advisor or other fiduciary unrelated to Quantopian about whether any given investment idea, strategy, product or service described herein may be appropriate for your circumstances. All investments involve risk, including loss of principal. Quantopian makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances.