Frequently Asked Questions
Q: You mentioned in recent emails that part of the switching process is contextual. If this is the case, how did you come up with your backdated results? Was historic contextual data somehow factored in when compiling the backdated results?
A: The backtested results (prior to 1991 when I created the model) were done using NASDAQ Composite, S&P 500, and leading stocks of the day. The leading stocks list was not fully dynamic so results prior to 1991 could have been improved had the list been dynamically updated on a daily basis, as it has been done in real-time since 1991. Also, context was not taken into account so the results could also have been improved on this basis. That said, the markets prior to 2000 were relatively clean. The 1970s, 1980s, and 1990s were clean trends relative to the shorter lived trends and grinding rallies we have seen in this decade, save for 2001 which was a strong downtrending market and 2003 which was a strong uptrending market. 2002 was downtrending but sloppy so whipsaw was its middle name. Thus, while the results from 1974 to 1990 were not optimized due to being unable to take context into account and have a list of leaders continuously updated, the results were strong nevertheless. Incidentally, most timing models have fallen short of being able to even outperform the major indices since 2004.