I wrote a post over at the Dough Roller about how the Dow Jones index was calculated, and the Finance Nerd left a great comment with a link to a Stanford Paper written by John B. Shoven and Clemens Sialm, which busts some myths about the Dow Jones Industrial Average (DJIA).
But, first a little background on how the index is calculated.
Calculation of Dow Jones Industrial Average
Unlike other major indices, the Dow Jones index is based on simple price averages and doesn’t take into account market capitalizations.
This means that the index is calculated by adding up the price of its 30 stocks and then dividing it by 30.
You are probably thinking that if you just add up the price of 30 Dow stocks and divide it by 30, the resulting number can’t possibly be the 8,300 or so, which the Dow is currently trading at.
You are right, and the reason it has reached so high is because it is divided by a number called the “Divisor”, which takes into account — stock splits and dividends. The divisor was invented to make sure that dividend and bonuses don’t impact the index. Discussion about the divisor can get a little involved, and if you are interested in the Divisor, head over to this post.
The other curious thing about the DJIA is that it is composed of 30 stocks selected by the editors of the WSJ. It is not an index of the 30 largest companies, as a lot of people think it is. The editors at WSJ have the freedom to pick and choose between stocks, and they select the best options (in their mind) to represent the overall economy.
Given these facts, let’s head back to the research paper.
There are three things that the paper examines about DJIA and much like Mythbusters, run their tests to see whether they are busted or not.
- It is too simplistic, so it must be flawed: The index is based on price weights, instead of the other superior methods like market capitalizations. That means a stock which has a higher price has more weight in the index, than a stock which has a lower price (regardless of the size). This should render the index quite flawed.
To test this myth – the researchers constructed a Dow Jones index that was value weighted (based on market capitalization) and compared it with the original index. The results are quite astonishing.
The actual DJIA stood at 239.43 in October 1928 and closed at 9,181.43 on December 1998. The value weighted index that was equal to 239.43 in October 1928 would have closed at 9,842.37 points in December 1998. They compare the monthly returns between the two indices and while the value weighted index outperforms the DJIA in 422 months out of 843, DJIA outperforms it in 421 months. The correlation between the two return series is 0.9778.
As far as I can see, there would have been very little difference between the value weights and price weight Dow, and this myth is busted.
The stocks are selected subjectively, so it must be flawed: The stocks that go to make the index are selected by the editors of Dow Jones and there are no set rules on the entry to this index. The editors decide which stocks will best represent the economy and then include them in the index. You can see that this is quite arbitrary and it leads you to believe that the index must have flawed results.
The researchers construct these hypothetical indices:
- Value Weighted Dow: Same thing we talked about earlier
- Value Weighted Big 30: Index of 30 largest publicly traded companies in the US from 1928 – 1998.
- Value Weighted Total Market Index: Index of all US Headquartered securities, which had over 8,000 companies in December 1998.
- And S&P 500
(They also use something called as Equal Weighted Index for calculation, but I am not including it here because it is pretty flawed itself, and doesn’t add much by way of comparison)
The researchers normalized these indices to start at a value of 239.43 on October 1928, and this how the indices fare:
- DJIA: 9,181.43
- VW Dow: 9,842.37
- VW Big 30: 11,211.32
- S&P 500: 13,776.55
- VW – Total Market Index: 12,141.95
As you see, the gap widens up, and looks like the Dow has actually underperformed the broader indices. This one presents no clear answers, so I leave it to you to decide whether this flaw makes any difference or not.
Myth # 3
This is the greatest part of the paper. Since stock market returns are used to gauge returns to investors, leaving out dividend payments doesn’t truly tell you how much investors made from the markets. So this is a serious flaw, which all other major indices share with DJIA.
The paper constructs a value weighted Dow Jones index that includes Dow Dividends and guess where it would stand in December 1998?
So, if you were to include dividend payouts, the index would have been 25 times of what it was!
Don’t even need to say who won this round.
This is a great paper, which examines some commonly held notions, tests them against data, and then comes up with some unexpected answers. At 16 pages, it is a relatively quick read and if you liked this post, you should definitely give it a try. You can download it at this link.
Photo Credit: kwc