Back to Community
short term vs. long term risk in the new model

Just a quick question on time horizon of the new risk model, maybe someone knows... my understanding of Fama-French type "factor" risk models is that they forecast only long-term results (e.g Investopedia suggests 15 year time horizon perhaps? https://www.investopedia.com/terms/f/famaandfrenchthreefactormodel.asp), and the QT model was built from 2 year daily data regressions (if I remember correctly from the webinar last Monday). But most Quantopian models are encouraged to trade fairly often I believe (e.g. "day" order of magnitude). It seems like for positions held for hours or small N days, you would want your risk model regressions built from intraday data. E.g. for ordinary beta to SPY, you would want to know how the stock is correlated to the reference SPY during the day. Or you would want to know how your stock correlates to the associated sector ETF during the day. Like an "intra day" risk factor model...otherwise your trading system may be penalized for long term risk when you are not holding anything long term. Comments?

1 response

Good question. Yes, the Fama-French type factor model works better over long horizons in general. I think, if using monthly data, 15 years could be a good choice. The Q risk model (not Fama-French type) is not designed to capture the influence of these factors in long-term horizon.
About frequency of tradings, let us talk about two relevant cases:

  • Intraday trading strategies. The Q risk model is built with daily data, if a strategy is purely intraday trading, I do not think analyzing it with an (EOD) position based performance attribution is a good idea. According to the ideas of strategies, I may try to decompose the daily algo returns to the risk factors directly, then use return-based performance attribution to check. We will have a lecture related to return-based performance attribution soon.
  • Daily rebalanced strategies. Rebalancing daily does not mean it is not possible to hold the same assets in one month. Suppose I have 100 assets in my portfolio on day 1, after rebalancing on day 2, I may still have 90 same assets. Even if all of the 100 assets are replaced on day 2, it is ok, since we have the daily factor exposures and factor returns are computed in advance. The performance attribution is just combining them based on the assets’ weights in your portfolio for each day.

https://www.quantopian.com/lectures/risk-constrained-portfolio-optimization and https://www.quantopian.com/posts/introduction-to-the-quantopian-risk-model-in-research are good references about the position based performance attribution and how to access assets’ daily factor exposures in Q research platform.

Disclaimer

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.