Channel Breakout Strategy Including Shares According To Volatility

Page 137 of Book Notes - Quantitative Trading Strategies talks about scaling the size of the trades in the equities to adjust for volatility. The book gives the formula of dividing $10,000 by the standard deviation of the stock's past 100 daily price changes. Price change is never rigorously spelled out: Is it the high low differences, is it the difference between successive day closes? We don't know. On page 138 figure 7.9 is a graph charting GE Stock Price and Shares to Trade showing as GE stock price increased, the number of shares traded decreased. This is in some ways a fluke, since it is not the actual price that is considered in the formula, but the amount that it changes each day. A small change in price could have us trading a really huge amount of shares.

I used this formula:

Unfortunately, I have not been able to replicate the numbers in the chart on page 138. It shows GE at $10 a share on 5/25/95, and a trade size of just under 100000 shares. The O/H/L/C of GE that day was 9.46/9.54/9.44/9.52

When GE tock is down around 10 dollars a share, the standard deviation is only 0.06, so this has us trading 170,553 shares! Also tradestation trading converts the shares to trade to a two byte unsigned in, so it allows a maximum of 65K shares to trade.... Correction: The formatting for strategies has a properties for all that specifies the maximum number of shares traded, which default is 65K. I raised the number to 250K. Still no help at reproducing the numbers in the book. I then raised it to 1,000,000 shares, the maxmimum, and my numbers get even farther from what is in the book. How about $22K in profit over 10 years of trading, with gross profit of $2.418M and a gross loss of $2.395M?

The final numbers I come up with are different than shown in the book, and wildly so, no doubt because of the unknown number of shares being traded.

I don't know why at least my number of trades and profitable percent didn't match. A difference of one profitable trade probably isn't enough to account for it. It appears book had 25 profitable trades, and mine had 23 profitable trades.

BTW, I adjusted my chart to include some days in advance of 1/1/1990 so that the maximum bars lookback would have data to work with:
// Trading days before 1/1/1990
// 40 - 11/02/89
// 50 - 10/19/89
// 60 - 10/5/89
// 70 - 9/21/89
// 80 - 9/7/89
// 90 - 8/23/89
// 100 - 8/9/89
// 120 - 7/12/89
// 150 - 5/29/89
// 180 - 4/17/89

Here is the strategy: