imported post
A review of this bookI borrowed from amazon. The author is credited.
Gaetan Lion (see more about me) from Mill Valley, CA USA - The authors maintain you can use a 15 year moving average of various valuation indicators as a buy signal to invest in the S&P 500. The indicators include: Price level, P/E ratio, Price/Book Value, Price/Cash Flow, Price/Sales. Whenever the S&P 500 trades under its 15 year moving average on these indicators, it is a good time to buy.
They show that between 1977 and 2001 an investor using any of these indicators (P, P/E, P/B, P/CF, P/S) to invest in the S&P 500 whenever it traded lower than its 15 year moving average would have beaten an investor doing dollar cost averaging with monthly investments over the same period. One condition is that the market timer would invest $200 every month he had a buy signal, and not invest anything when he did not. Meanwhile, the dollar-cost-averager would invest $100 on a monthly basis. Over the entire period, the market timer gets to invest only $20,000 in the S&P 500 ($200 times 100 months). Meanwhile, the dollar cost averager invests $30,000 in the S&P 500 ($100 times 300 months). In each case, the market timer comes out ahead and ends up with more money in 2001.
The authors attempt to make a case that the market timer superior results (regardless of the indicator used) is due to buying into the market when it is low. But, the success of the market timer is due to his accelerated equity investment schedule. By early 1985, the market timer has made his full investment of $20,000 in equities. By the same date, the dollar cost averager has invested only $10,000. The dollar cost averager much slower investment schedule will never allow him to catch up to the market timer. The dollar-cost-averager average holding period of his stock portfolio is only 12.5 years compared to 21.5 years for the market timer. This is why the market timer wins.
The above example is repeated five times (once for each indicator) with the exact same flaw. The accelerated equity investment schedule follows an identical pattern regardless of the indicator used. What these guys did is called backtesting. The authors looked at various moving averages such as 5, 10, and 15 years. And, devised different investment rules until they came up with a combination of a moving average and a rule that would beat income averaging. They did it, but ran into a methodological flaw (the frontloading of the investment) they were not even aware off.
Additionally, this investment strategy is most unrealistic. Who could stomach investing twice as much as his regular investment schedule just when the market is experiencing a severe Bear market (that is what it takes for the market to go south of its 15 year moving average).
This strategy is not market timing. Following this strategy you would have been locked out of making any additional equity investments since 1985. Also, the authors do not have a "sell signal" spelled out because it would kill their strategy. The stock market only rarely goes south of its 15 year moving average (regardless of the indicatory used) and quickly goes back up north of it. Thus, with a sell signal you would never hold your S&P 500 holdings for long. Yet, the authors stress that their strategy does not show superior returns until 15 to 20 years out. The lack of a sell signal would have left you fully exposed on your S&P 500 holdings invested prior to 1985 to the crash of 1987, the downturn of 1989, and the severe Bear Market of 1999 to 2002.
That's a cool website though.