Also, when looking at the Amazon example, or testing stocks with my spreadsheet, you have to keep in mind the limitation - that we are just seeing how a particular stock or ETF would trade with constant value investing.
Since it is in isolation, it doesn't necessarily show how it would actually affect the portfolio's drawdown or return.
For example:
1. We might not have needed to add all cash to the portfolio because some other stocks in the portfolio may have kicked off cash.
2. In 2004, the AMZN example's cash pool was over 30% - which meant that the excess could have been invested in a new position.
This is actually a drawback of the spreadsheet for comparing the STR system to AIM BTB or B&H for increasing securities. Since STR doesn't increase the constant value automatically, the cash pool builds up and drags the return. But, you have to imagine that, in actual practice, when the cash pool gets beyond 30% of the portfolio value, you are supposed to add new positions.
For example, if you compared STR, AIM BTB, and B&H with Microsoft during the boom years, the STR spreadsheet would show a lower return, with a high % of the portfolio in cash. But, you have to remember that you would have added more positions that would then also be rebalanced. If those excess positions had been Oracle and/or Apple, you could have more than made up ground with the other systems.