Easiest Way to Beat the S&P?

Note: Please read the disclaimer. The author is not providing professional investing advice.

First off, let me just say that I’m posting this as more of a things that make you go hmmm curiosity than anything else. We all are intrigued by trading formulas that have been shown historically to beat the market averages – especially simple ones such as Joel Greenblatt’s Magic Formula. But we’ll never know if future returns will correlate with those of the past.

So after reading Greenblatt’s little book, I was inspired to come up with my own magic algorithm that beat the S&P. And I found one – fairly quickly.

No, it doesn’t outperform Greenblatt’s or perhaps make as much rational sense. Nor has it been as thoroughly analyzed to see if it’s just a quirk of data mining, like the Foolish Four. I will only make the following 3 claims about it:

(1) It’s simpler than the Magic Formula
(2) For the time period tested, which includes the 17 years Greenblatt used plus 2 additional years, it outperformed the S&P 500 average
(3) It achieved the above with a lower standard deviation (a common quantifier of risk)

Ready for the algorithm?

Step 1: Buy Exxon Mobil Corporation (XOM) on January 1st.
Step 2: There is no step 2 – hold XOM forever. But if you have more money to invest later, wait until the next January 1st and then see Step 1.

Want some backtesting results to support using such an algorithm? OK, below in Table 1 are annual results starting with 1988 just like Greenblatt, and stopping with 2006 (whereas he stops in 2004).

For the market average, I’m using the Vanguard 500 index fund (ticker VFINX) because this is probably what you’d buy if you wanted to closely track the S&P 500.

FYI I’m computing the annual ROR as the current year’s close divided by the previous year’s close.

Table 1. Annual Rates of Return
1988 16.3% 21.2%
1989 31.4% 19.6%
1990 -3.3% 8.9%
1991 30.2% 23.2%
1992 8.2% 5.2%
1993 9.1% 8.1%
1994 1.2% 0.8%
1995 37.5% 39.4%
1996 22.9% 25.4%
1997 33.2% 28.5%
1998 28.6% 22.4%
1999 21.1% 12.6%
2000 -9.1% 10.3%
2001 -12.0% -7.6%
2002 -22.2% -8.9%
2003 28.5% 20.6%
2004 10.7% 28.1%
2005 4.8% 11.8%
2006 15.6% 39.1%
Ave 12.0% 15.5%
Std Dev 16.6% 12.3%

Average & standard deviation computations above are geometric, not arithmetic.

How easy was that?! Just by buying XOM you would have averaged 15.5% over the last 19 years, versus the S&P’s measly 12%. This means your money would have doubled every 4.8 years instead of every 6.1 years.

What’s more, you would have achieved this higher return with less yearly fluctuation with this single stock versus the larger annual variation seen by the whole S&P 500.

What would be a good next step? Well for one thing we only know what happens when you put your money into XOM on January 1 of each year. What happens if you use another month and day instead? If this is truly a respectable algorithm, as a minimum it should work in those conditions too.

Finally, I just picked XOM because I was looking for any Dow stock that achieved a higher average ROR with lower variance versus the S&P since 1988. This happens to be the only one in the current Dow 30 but I wouldn’t be surprised if there are others outside of the Dow that have an even higher average ROR with an even lower annual standard deviation (risk).

Perhaps one other item worth mentioning is sustainable growth rates. XOM has done well historically but can this go on forever? Afterall, as postulated in Value Investing for Dummies, can we really expect a company to continue to deliver large earnings growth when their gross profit has reached $158 billion, unless we discover life on other planets (i.e. untapped markets)?

Hint: it’s not impossible if the company uses its cash to acquire other young companies that can provide such growth!