100% Cash and Out of Sample Returns

People email me every day asking how the timing model has held up during this market dump (short answer: great). The model is currently 100% in cash/bonds (more on that below).

When I do a blog re-design I will include a tab with historical results and updates if enough people are interested (and I also plan on doing an update to the paper at the end of the year). All returns are gross of management and transaction fees.

Here are out-of-sample results (meaning after I published the paper), from 2006-9/2008. (They would look better if you included October, of course.) I think we are setting up for a monster Nov/Dec rally, but what do I know?

The model is doing what it was designed to do, namely, protecting capital as markets decline and achieving equity like returns with bond-like volatility. (Chart fixed thanks to reader Sea – had to rescale the Y-axis the S&P was down so much!)


I was going to title this post 10 things to do when you are sitting on 100% cash, but I didn’t think that was a good idea because 1) when you are beaming about how smart you are the market has a way of humbling you, and 2) there are a lot of people losing a lot of money, and that is nothing to gloat at. (PS – If the S&P closed here, it would be the third worst year ever since 1926...)
A reader emailed in asking how many times it occurred when the model was 100% cash (out of five positions). It turns out that has only happened 4 times since 1972-2008:
While this is a tiny statistical sample, I thought it would be interested to see what the resulting returns were for the buy and hold portfolio over the following 12 months. The theory of course goes back to the old “buy when there is blood on the streets”. The 12-month performance for the four dates was :

6/30/1974, 11.34%

4/30/1980, 18.39%

9/30/1981, 14.11%

10/31/1981, 16.04%

That works out to an average of 14.97%, or about 3.5% higher than average. Thus, it’s usually a good time to be investing. What about when only one position is on (20% invested)?

9/30/1973, -19.43%

4/30/1974, 2.94%

5/31/1974, 8.54%

7/31/1974, 16.04%

8/31/1974, 8.68%

9/30/1974, 20.43%

10/31/1974, 17.07%

11/30/1981, 14.47%

3/31/1982, 32.98%

7/31/1984, 25.95%

8/31/1984, 19.00%

11/30/1987, 19.23%

12/31/1987, 18.46%

5/31/1990, 14.24%

12/30/1994, 22.74%

1/31/1995, 24.91%

9/30/1998, 11.35%

10/30/1998, 9.58%

Similar results at 14.84%, with about 3.4% abnormal returns.

With September being one of the worst months on record for stocks, the strategy of waiting a month then investing for 2 months could be a great time November 1st. For equities back to 1900, when stocks have declined by over -9%, the resulting “wait a month then invest for two months” offers good returns on average, but the range is large. The worst two month return would have been around -24%, and the best around 91%…That lines up nicely with the traditional year end rally.