Feedback on White Paper
Feedback on Blackstar Funds white paper posted yesterday:
“Hi Michael! Thank you for your informative blog - I read it every time you place a new entry. I particularly enjoyed your recent post titled “Trend Following on Stocks”. However, at the very end of the publication the author makes a trading slip. It’s what I call the ’stock split bias’. In the article the author ends by mentioning that they filtered stocks that traded below an absolute dollar figure of $15 - to filter out undesirable trades/stocks. While the logic seems sound on face value, in back testing it filters out stocks without taking into consideration the stock splits. As an example, take Microsoft. With its current stock split history it doesn’t trade above $15 until around mid 1997, however, back in 1997 Microsoft was actually trading around $120! Therefore, neglecting MSFT prior to 1997 because of its current stock split history would not accurately reflect the performance of the system back in 1997 (where it would have traded MSFT). Of course if the back-testing incorporated stock split data this problem could be circumvented, however, using absolute dollar figures on price should not be used in back-testing if incorporating stock splits cannot be used. Thanks for your time and I look forward to reading more great trend following material (loved your book too!). From an avid reader and trader, Ryan”
The author Cole Wilcox wrote me back to clarify:
“He is smart to recognize this, however if he re-reads Appendix 4 he will see that we have accounted for this by maintaining 2 sets of data. Our minimum price filter is based on the actual prices that hit the tape, not the split adjusted data.”
