Growth-Value Switching

I varied the allocations between Large Cap Growth stocks and Large Cap Value stocks in a portfolio in accordance with P/E10. I maintained a 25% allocation in T-Bills and a combined 75% allocation in stocks.

I used my Gummy 04A01 (January 28, 2005) version of the Deluxe Calculator V1.1A08.

In earlier investigations, I looked at Large Cap Growth stocks or Large Cap Value stocks separately. The non-stock investments have been either commercial paper or TIPS.

Here are my findings:
1) Switching can improve performance.
2) When switching in accordance with P/E10, you should treat Large Cap Value stocks as the traditional stock component and Large Cap Growth stocks as the traditional non-stock (i.e., commercial paper or TIPS) component. That is, buy Value Stocks heavily when P/E10 is low. Reduce your Value Stock allocation when P/E10 is high.
3) There are three regions of interest:
a) P/E10 thresholds of 12 to 16 produce 1-Failure rates of 4.9% and 5.0%
b) P/E10 thresholds of 19 to 21 produce 5-Failure rates of 5.7% and 5.8%
c) a P/E10 threshold of 24 produces a 10-Failure rate of 6.6%
4) I estimate the potential of Growth-Value switching as being (approximately) that all of these levels of performance would occur at the same time. That is, I estimate the potential of Growth-Value switching as producing the following:
a) a 1-Failure rate of 5.0%,
b) a 5-Failure rate of 5.8%,
c) a 10-Failure rate of 6.6%.
5) Even if Growth-Value switching were to produce these results, it would be substantially inferior to Switching with Large Cap Value and commercial paper. Switching with Large Cap Value and commercial paper produced the following:
a) a 1-Failure rate of 6.9%,
b) a 5-Failure rate of 7.5%,
c) a 10-Failure rate of 7.8%.
6) Growth-Value switching produces mixed results when compared to Switching with Large Cap Growth and commercial paper and with Switching with S&P500 index and commercial paper. It is likely to remain inferior at the 1-Failure rate, but superior at the 5-Failure rate and at the 10-Failure rate.
7) Normal optimization would make the 1-Failure rate the most important.

I decided not to pursue Growth-Value Switching any further. Its potential appears to be limited.

Growth-Value Switching Data

Have fun.

John Walter Russell
August 29, 2005