Quantopian's community platform is shutting down. Please read this post for more information and download your code.
Back to Community
Quandl VIX spike of 7200 on 2016-03-18

Looks like there is a huge spike in yahoo_index_vix close data series on 2016-03-18, its value is 7200:

from quantopian.interactive.data.quandl import yahoo_index_vix as vix  
data = np.array(map(np.array, vix[['asof_date', 'close']].sort('asof_date', ascending=True)))  
data[6605]

array([datetime.datetime(2016, 3, 18, 0, 0), 7200.0], dtype=object)  

Is this a mistake?

4 responses

Hi Viktor,
It looks like this was a bad data point from Yahoo. This data point was later corrected in the Quandl set. I've attached a notebook demonstrating this.

For all of our data sets from 3rd party partners (i.e. those listed at quantopian.com/data), we monitor for corrections and revisions to previous data points and store those revisions in separate table. You can access that data in interactive mode by appending _deltas to any dataset. So in this case import yahoo_index_vix_deltas to see the two revisions we've processed for this data set since we started processing it.

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.

Josh.

There were no bugs in Yahoo database

Date Open High Low Close Volume Adj Close*
Mar 21, 2016 14.57 14.73 13.79 13.79 0 13.79
Mar 18, 2016 14.05 14.36 13.75 14.02 0 14.02
Mar 17, 2016 15.34 15.38 13.82 14.44 0 14.44

Mar 16, 2016 15.96 16.33 14.89 14.99 0 14.99
Mar 15, 2016 17.60 17.85 16.84 16.84 0 16.84
Mar 14, 2016 17.01 17.67 16.69 16.92 0 16.92
Mar 11, 2016 17.09 17.27 16.28 16.50 0 16.50
Mar 10, 2016 18.17 19.59 17.06 18.05 0 18.05
Mar 9, 2016 18.56 19.11 18.31 18.34 0 18.34
Mar 8, 2016 18.38 18.89 17.82 18.67 0 18.67
Mar 7, 2016 17.98 18.04 16.87 17.35 0 17.35
Mar 4, 2016 16.48 17.35 16.05 16.86 0 16.86
Mar 3, 2016 17.25 17.56 16.32 16.70 0 16.70
Mar 2, 2016 17.98 18.41 16.78 17.09 0 17.09
Mar 1, 2016 19.84 20.17 17.66 17.70 0 17.70

So just to be clear, yahoo HAD a bad data point, which they subsequently corrected, but which is kept in the pipeline results because had you tried to use it as-of that day, you would have been burned by it, so it's best to burn backtests too.

Correct. But we track the revisions for use in lookback windows (once the correction is known). We track when we know about the correction and only make that data available to the simulation when the correction would be known.