Larry's Market Comments at Critical Points

[market_watcher]

Hey Larry,

I got a system I just developed for shorting the QQQs that's produced 4 signals since last July 5th's closing high. I see you share alot of your findings, so here's hopefully a good contribution. The first 3 signals were winners. The 4th signal came on June 5th at the QQQ close of 30.64

1. Do a linear regression of price of QQQs against time in days since the closing high on July 5th.

2. Calculate the 9-day EMA of the percentage above or below the regression line the QQQ is at the end of the trading day(4pm)

3. If the 9-day EMA is more than 5.5% above the regression line's predicted value for that day, short the QQQ. Cover the short when the 9-day EMA is less than 5.5% above the regression line on a closing or intraday basis. Your choice but intraday would avoid any whipsaw action.

4. Average gain 11.52%. Holding time of 10,8 and 16 days.

5. For a stop loss, use 4%. The intraday high has never exceeded the shorting price by more than 3.8%

Main problem I see: Knowing when the trend changes from bull to bear. From that point, the regression line needs to change as well. For example, nothing this clear cut can be created regressing price on time from before the July high.

I'll try to work on finding something similar for going long. Or, maybe you just stay long the rest of the time?