Discussion in 'Data Sets and Feeds' started by pkoufalas, May 3, 2007.

  1. pkoufalas


    Anyone using datastream as their data provider and willing to provide me with S&P500 and SPX daily prices? I can pay.

    I'm backtesting an options stat arb trading strategy and can't replicate someone else's backtest results. I'm wondering if the option price data I'm using is the culprit.

    I bought options data from CBOE's MDX and use Yahoo for the index, whereas the comparison results used Datastream. There is either something wrong with my filtering of the daily options chains or the OHLC prices really differ, as the average prices and standard deviations of the filtered data from MDX don't match those from the filtered data from Datastream. So, the inputs don't match and hence it's not surprising that the outputs differ.

    If you can help, please send me a message. The differences in results are driving me crazy!!! Especially since mine show a loss after fees.

  2. pkoufalas



    Perhaps I shouldn't be surprised at the lack of it reasonable to say that Datastream terminals wouldn't be much used outside of financial institutions, universities and the like?

  3. BENG


    Didn't you answer your own question? The data sets are different...
  4. pkoufalas


    I filtered the CBOE/MDX supplied daily options chains to extract the options of interest to the stat arb strategy. Assuming I did that "correctly", I found that the number of options traded (backtested) and their closing price statistics differ from those that were obtained by the other party using the Datastream supplied data set.

    I'm a beginner, so I may have misunderstood the other party's option selection (filtering) criteria, or the CBOE/MDX and Datastream datasets might really differ. I don't know which. I've tried several different approaches in order to see if I can match the comparison results for number of options traded and closing price statistics, but to no avail. Eg using SPX last bid- last ask midpoint rather than last sale, time-synchronous closing S&P500 index value (as recorded in the options data) rather than index close from Yahoo, variations on minimum traded call and put option volumes (and, or, sum).

    So I'd like to examine the Datastream-supplied options data -- I suspect it will give me important insights into the proposed stat arb strategy. I'll also try to find some Datastream documentation on their options data and the data fields in it; perhaps a description of the fields will give some clues. Eg their call/put close prices are not last sale prices but 5 min prices after the exchange, or averages during the last 15 min of trading, or ... Surely such details are important to understanding the viability and implementation of the strategy?

    Thanks for your reply.