Quantopian's community platform is shutting down. Please read this post for more information and download your code.
Back to Community
MaxLossExceededError

I have been trying to test an alpha factor using alphalens (sentiment), but I always get a MaxLossExceededError of over 90.1%
but if i try the same code but with the factor being SMA of prices then it has no problem.

Date of pipeline output = 2016-01-01 to 2017-01-01
Date of pricing data = 2015-12-01 to 2018-01-01

Why is this happening?

2 responses

Hi Aaranya, this looks like a perfect use case for passing quantiles=None and bins=[whatever number of bins you want] arguments to get_clean_factor_and_forward_returns().

There are two reasons for MaxLossExceededError:

  1. The pricing data doesn't stretch far enough into the future.

  2. You are losing data due to "binning." What this means is that there
    are so many similar (or exactly the same) values that its impossible
    to fit them into evenly sized quantiles.

It seems like you are aware of the first cause, so the issue is probably a result of the second cause. For example, try to fit the following values into 3 Quantiles: [0, 0, 0, 2, 1].

It is impossible to evenly fit those numbers into 3 quantiles, so we must use bins instead of quantiles. Bins act mostly the same as quantiles, with the notable exception that they can be different sizes (quantiles need to be roughly the same size). For example, those numbers might fit into 3 bins like this: [0, 0, 0], [1], [2]. They might also fit into 2 bins like this: [0, 0, 0], [1, 2].

It seems like we need to change our documentation to do a better job revealing the cause of that error. Thanks for the feedback.

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.

Hey thanks Cal, yup I figured that its mostly the lack of variation in the factor values that is causing this problem.
The sentiment_signal factor only ranges from -3 to 6 - 10 integer values.

Got around the problem by converting them to zscores and by adding a dummy factors z score to it.
Thanks