Fool.com: Trouncing the Dow, Part 2 (Foolish Four) December 15, 1998

<FOOLISH FOUR PORTFOLIO>

Trouncing the Dow, Part 2
The Numbers

Houston, TX (December 15, 1998) -- Last week, we looked at Kenneth Lee's new book Trouncing the Dow. Today I will report on a backtest I performed using his picks and our independent database.

In the book, the backtest extends from 1973-1996. Ken gave me the picks for 1997 -- AT&T <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE: T)") else Response.Write("(NYSE: T)") end if %> and Caterpillar <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE: CAT)") else Response.Write("(NYSE: CAT)") end if %> -- and using these, I've brought our backtest up-to-date though the end of 1997.

Checking the numbers

The data I used is based on January starting prices. Ken's data is based on December ending prices. His returns came up mostly a bit higher than what I found. I checked his numbers in several of the cases where we differed most, and the prices were always exactly correct. In some cases, he understated the dividends. (I'm not sure whose database he used, but that seems to be a common problem in the databases I've seen.) This would tend to lower his returns, not raise them. So the fact that his returns were higher than what I found confirms what we mentioned last week: The end of December is probably a better starting date than the first trading day in January.

Buying and selling rules for the backtest

I used the picks Ken gave for the end of the year as the picks at the start of the next year. These were held for a year, sold, and the next year's picks were bought. This is the same set of rules as were used in Ken's backtest. The only differences are the one-day difference in time and the fact that we carried the test one year further than Ken was able to do in the book.

In the table below, "BI" is the Benchmark Investing return. "BI ct" is the count of stocks the Benchmark strategy selected in the given year. (Benchmark Investing is what Ken calls his strategy -- classier than Trouncing the Dow but, I suspect, not as saleable.)

YEAR

Dow 30

S&P 500

HY10

Fool 4

RP 4

BI

BI ct

1973 -10.86% -14.66% 3.89% 25.74% 17.28% -3.93% 14

1974 -15.64% -26.47% 1.04% 5.25% 20.00% -5.86% 14

1975 44.25% 37.20% 52.17% 68.71% 68.71% 73.54% 5

1976 29.36% 23.84% 33.24% 36.92% 37.93% 32.74% 19

1977 -12.58% -7.18% 1.17% 5.32% -2.96% -4.56% 14

1978 2.53% 6.56% 2.44% 9.89% 9.89% 9.03% 16

1979 11.34% 18.44% 9.69% 2.17% 17.70% 14.79% 20

1980 25.29% 32.42% 32.95% 48.18% 24.20% 40.04% 15

1981 -3.30% -4.91% 4.87% -4.63% 9.66% 7.38% 9

1982 19.80% 21.41% 20.87% 41.58% 56.88% 58.31% 5

1983 35.35% 22.51% 38.43% 41.74% 36.72% 35.83% 7

1984 -0.12% 6.27% 7.45% 10.24% 10.30% 17.00% 3

1985 30.98% 32.16% 30.61% 22.85% 49.82% 51.73% 1

1986 21.87% 18.47% 29.43% 27.30% 29.67% 71.94% 1

1987 15.72% 5.23% 8.56% 18.75% 17.89% 50.09% 1

1988 13.78% 16.81% 17.96% 13.62% 22.08% 16.18% 6

1989 31.95% 31.49% 29.64% 15.28% 47.35% 23.85% 5

1990 -9.14% -3.17% -10.01% -17.61% -17.61% 9.97% 2

1991 30.36% 30.55% 35.24% 81.61% 34.81% 27.03% 4

1992 11.00% 7.67% 6.35% 29.94% 29.94% 15.25% 3

1993 17.91% 9.99% 23.54% 26.22% 30.26% 37.00% 4

1994 3.73% 1.31% 2.43% 4.72% 7.59% 8.53% 2

1995 36.69% 37.43% 37.10% 30.42% 47.05% 40.57% 8

1996 24.32% 23.07% 27.47% 24.34% 26.56% 46.14% 2

1997 22.33% 33.36% 20.39% 22.31% 19.49% 40.16% 2

25 Year Compound average returns (1973 - 1997)

Dow30 S&P500 HY10 Fool4 RP4 BI

13.83% 13.06% 17.69% 21.86% 24.62% 26.62%

25 Year Total Returns --$10,000 became

Dow30 S&P500 HY10 Fool4 RP4 BI

$254,756 $215,208 $586,583 $1,401,316 $2,454,361 $3,655,824

25 Year Performance Measures (1973 - 1997)

Dow30 S&P500 HY10 Fool4 RP4 BI

Avg. 15.08% 14.39% 18.68% 23.63% 26.05% 28.51%

StDev 16.93% 17.12% 15.64% 21.88% 19.13% 22.59%

Sharpe 0.4697 0.4309 0.7412 0.7492 1.0045 0.9561

So, Benchmark Investing averaged two percentage points better than the next best strategy, RP4. This 2% difference over a period of 25 years is huge, as you can see by the "$10,000 became" line -- almost 50% more money. BI does have a higher standard deviation (possibly because of the large variation in the number of stocks) and a slightly lower Sharpe ratio than RP4, but the Sharpe ratio still looks good compared to our other strategies.

Final comments

These numbers (and detailed looks I did in some cases of difference) verify that the price and dividend information, and hence the total return information, in Trouncing the Dow is accurate. This system has produced higher returns than any other Dow system I know of, with a reasonable standard deviation and a good Sharpe ratio.

I only have two minor complaints with the system as described in the book. One is the selling rules. I like the selling rules used in the backtest -- simply hold for a year and sell. That's really the system that both Ken and I have backtested. He does suggest some other selling rules that lead to longer holding periods, but these haven't been backtested as far as I know. While they make sense, I'll probably stick to the calendar year approach.

The other quibble I have with the system is that the number of stocks held varies so widely. It's a natural consequence of the rules -- only buying the stocks that are strictly undervalued. But I'd like to see further research done on a concept of using the prices as a ranking system -- always buying the 4 "cheapest" stocks by Trouncing the Dow standards, for instance. There have already been some threads on the Dow web boards discussing this, and I expect that those discussions will be worth watching.

Even with those reservations, as it stands, it's the best performing Dow system I know of.

Questions?

Unfortunately, we are not able to answer most questions about Benchmark Investing. However, Ken Lee has been participating in the Dow board on the web and has been answering questions there. His name on the boards is dowbuys.

Fool on!

<% =headlines %>

Get the Fool's new book - The Foolish Four

Current Dow Order | 1998 Dow Returns

12/15/98 Close

Stock Change Last -------------------- UK + 7/16 41.00 IP + 5/8 42.19 MO +1 13/16 53.56 EK + 13/16 71.69

Day Month Year FOOL-4 +1.81% -4.04% 9.53% DJIA +1.47% -3.22% 11.57% S&P 500 +1.89% -0.07% 19.82% NASDAQ +2.32% 3.23% 28.16% Rec'd # Security In At Now Change 12/31/97 276 Philip Mor 45.25 53.56 18.37% 12/31/97 206 Eastman Ko 60.56 71.69 18.37% 12/31/97 289 Int'l Pape 43.13 42.19 -2.17% 12/31/97 291 Union Carb 42.94 41.00 -4.51% Rec'd # Security In At Value Change 12/31/97 276 Philip Mor 12489.00 14783.25 $2294.25 12/31/97 206 Eastman Ko 12475.88 14767.63 $2291.75 12/31/97 289 Int'l Pape 12463.13 12192.19 -$270.94 12/31/97 291 Union Carb 12494.81 11931.00 -$563.81 Dividends Paid YTD $1092.81 TOTAL $54766.87

</FOOLISH FOUR PORTFOLIO>

YEAR	Dow 30	S&P 500	HY10	Fool 4	RP 4	BI	BI ct
1973	-10.86%	-14.66%	3.89%	25.74%	17.28%	-3.93%	14
1974	-15.64%	-26.47%	1.04%	5.25%	20.00%	-5.86%	14
1975	44.25%	37.20%	52.17%	68.71%	68.71%	73.54%	5
1976	29.36%	23.84%	33.24%	36.92%	37.93%	32.74%	19
1977	-12.58%	-7.18%	1.17%	5.32%	-2.96%	-4.56%	14
1978	2.53%	6.56%	2.44%	9.89%	9.89%	9.03%	16
1979	11.34%	18.44%	9.69%	2.17%	17.70%	14.79%	20
1980	25.29%	32.42%	32.95%	48.18%	24.20%	40.04%	15
1981	-3.30%	-4.91%	4.87%	-4.63%	9.66%	7.38%	9
1982	19.80%	21.41%	20.87%	41.58%	56.88%	58.31%	5
1983	35.35%	22.51%	38.43%	41.74%	36.72%	35.83%	7
1984	-0.12%	6.27%	7.45%	10.24%	10.30%	17.00%	3
1985	30.98%	32.16%	30.61%	22.85%	49.82%	51.73%	1
1986	21.87%	18.47%	29.43%	27.30%	29.67%	71.94%	1
1987	15.72%	5.23%	8.56%	18.75%	17.89%	50.09%	1
1988	13.78%	16.81%	17.96%	13.62%	22.08%	16.18%	6
1989	31.95%	31.49%	29.64%	15.28%	47.35%	23.85%	5
1990	-9.14%	-3.17%	-10.01%	-17.61%	-17.61%	9.97%	2
1991	30.36%	30.55%	35.24%	81.61%	34.81%	27.03%	4
1992	11.00%	7.67%	6.35%	29.94%	29.94%	15.25%	3
1993	17.91%	9.99%	23.54%	26.22%	30.26%	37.00%	4
1994	3.73%	1.31%	2.43%	4.72%	7.59%	8.53%	2
1995	36.69%	37.43%	37.10%	30.42%	47.05%	40.57%	8
1996	24.32%	23.07%	27.47%	24.34%	26.56%	46.14%	2
1997	22.33%	33.36%	20.39%	22.31%	19.49%	40.16%	2

25 Year Compound average returns (1973 - 1997)
Dow30	S&P500	HY10	Fool4	RP4	BI
13.83%	13.06%	17.69%	21.86%	24.62%	26.62%