No. of Recommendations: 8
An update the GTR1 backtest of using BCC for a QQQ signal.
The first step requires BCC = 7 this incorrectly leaves out several months of gains at the start.
Next, although BCC = 0 which requires all three signals to be Bull before going to cash works
a more cautious approach is to only require two of the three signals to be bullish BCC = 0!1!2!4.
The 19990310 to 20230310 results:
CAGR 10.76 SAWR 8.2% MDD -36.8 UI 11.7 Annual Turnover 1.25
For QQQ buy and hold same period:
CAGR 8.17 SAWR 3.1% MDD -83 UI 43.7 Annual Turnover 0.0
https://gtr1.net/2013/?s19990310h5::BCC:lt8:cashif...Fantastic Results!
But just a little caution with depending on a backtest like this tuned and retuned after the last few market downturns. And I just retuned the results to use 2 of 3 indicators rather than all of them. The other big problem with market timing is that there just aren't enough cycles which last for more than a year in markets to give good statistical evidence of their accuracy. Remember BCC was developed after the 2008 downturn. Before that there were 100's of posts on the MI board comparing the three major timing services Timing Cube, Highlight and Intelli-Timer. For a long time before those three Fund Advice which had a publicly posted timing system since 1983 went silent in June 2008. To BCC's advantage two of the three indicators the 200 day returns and the 100 day no new highs are two of the oldest around.
I'm an old guy who can't help but looking for that magic market system that will protect me from the next big correction. But I've been around long enough and seen so many timing systems look like a sure thing only to crash that I wouldn't trust any system with all my marbles no matter how solid they look. One famous example from way back in the '60's Joe Granville in a weekly market newsletter became famous for predicting significant market moves. So many people started believing in his predictions that in April 1980 he went from 100% short to buy and the market went significantly up. His newsletter had over 20,000 subscribers at $250/year and 1,000+ subscribers at $500 /year would get telegraphic notices of market calls. He closed down his prediction service after he made a series of bad predictions. Later a study of his overall accuracy after concluded that his overall prediction accuracy was significantly better than random. As a side note during his peak, a small company I cofounded was developing an automated phone notification system (Pre internet) for his prime subscribers when he closed up shop.
RAMc
