the Dow Jones Industrial (DJI) average

Despite its media ubiquity and the fact it has survived all these years, the DJI is a weird index:

–it contains only 30 names out of the thousands of publicly traded companies

–although the index owners have tried to make it more relevant in recent years by adding Apple, Microsoft and Nike, it still by and large represents the big-cap names of America’s yesterday

–the weighting of a given name is a function of its per share stock price, not the size of the company.  As a result, Microsoft counts less than Visa, despite being 3x V’s size.  MSFT is about 1200% the size of IBM, but has only a third more weight (stock price of 150 vs 110).  While the ease of calculation this brings might have been important in the pre-computer age, it’s an anachronism now

–because of all this, using the DJI is a convenient signal to the listener that a speaker knows very little about stocks.  Odd that it should be used by media stock “experts”   …or maybe not.


Pre-APPL, MSFT, NKE, the DJI did have one important use.  When it started to outperform, that meant that a rally was near its end (and portfolio managers were buying the least interesting, but cheapest stocks) or that pms were seeking the safety of large, mature companies.  The additions above have lessened that appeal.

However, in the current climate, the DJI is an interesting collection of coronavirus losers.

Year-to-date, as of the close on Monday, the Dow was down by 35%, the S&P 500 by 30%, NASDAQ by 24%.  Since then, the DJI has been by far the best performer.  Interestingly, the Russell 2000, which measures mid-cap US, was down by 40% on Monday and has bounced by about 8% since, tying it with the NASDAQ for smallest bounce.

Two days isn’t much to go on, but one read is the market thinks the bailout will mostly benefit the large old-guard industrials.  A caveat:  the 57% rise in Boeing over the past two days accounts for two percentage points of the DJI rise.




housecleaning at the Dow Industrials

Standard and Poors, the part of the McGraw Hill empire that controls the Dow Jones averages, announced today that the Dow Industrials would be bouncing out three of its thirty components, effective September 20th.

The companies to be shown the door are :  Alcoa (AA), Bank of America (BAC) and Hewlett-Packard (HPQ).  Their offense?   …stock prices that are too low, and not enough diversification appeal.

They’re being replaced by:  Goldman Sachs (GS), Visa (V) and Nike (NKE), all of which have higher stock prices and supposedly give the Dow more diversification.

We all know the Dow is a weird index.  That’s because it’s calculated using each stock’s per share price (not the total value of the company’s equity) as the weighting factor.  So a stock that sells for $50 a share has twice the potential impact on the index of a stock that sells for $25–even though the latter may be a much bigger company, with a much larger total market value.

No wonder Alcoa, an $8 stock, and Bank of America ($14.60) are gone.  They’re insignificant!

No surprise, either that Apple ($494) and Google ($888) aren’t in.     …too large.

Why calculate an index this way?  The only answer I can come up with is that in the (computer- and calculator-less) late nineteenth century, when the Dow was invented, the math was simple enough for reporters to get done quickly at the end of the day.

Why does it still exist?   …and why do individual investors still pay attention?  At first blush, the answer is that the Dow is almost all the news media talk about.  But the media would ditch it in a second if the Dow weren’t a surprisingly good mimic of the more sensibly constructed S&P 500, which is what investment professionals use as their benchmark.

Think about it for a minute (something I haven’t done until recently).  What are the chances that a small, wacky index can track the S&P 500 so closely?

As I recall, it wasn’t always this way.  Back in the day, investors tracked the Dow vs. the S&P.  If the Dow was doing better, it meant the ghost of Christmas past was in the room.  If the S&P was outperforming, then smaller stocks in less mature industries were on the relative rise.

Not so much anymore.

It seems to me a tremendous amount of brainpower and computer time has recently gone into–and continues to go into–tuning the Dow Industrials so that they’ll keep on tracking the S&P pretty faithfully.  That’s the really interesting thing about the Dow, to my mind.  And it’s the reason AA, BAC and HPQ had to go–not for diversification, but because they weren’t helping the Dow track the S&P.

why professionals use the S&P 500 as a benchmark, not the Dow Industrials

the Dow at an all-time high

Yesterday, the Dow Jones Industrial Average reached an all-time high, signalling a recovery of the index past its previous, pre-Great Recession peak.  This is certainly good news psychologically, since the Dow is a very widely recognized name.  And any report that puts the downturn behind us is welcome.  As it happens, I was listening to Bloomberg Radio when the new high-water mark was first achieved.  A Surveillance commentator played down the fact that the S&P 500 had not yet recovered all its recession losses (although it’s close), saying that the use of the S&P over the Dow is “just” a question of “institutional bias.”

That’s simply incorrect.  …hence this post.

what’s wrong with the Dow

There are three reasons professional investors use the S&P 500 to benchmark their performance rather than the Dow:

1.  wider coverage.  The Dow Industrials consist of 30 American companies; the S&P, as the name suggests, has 500.

2.  greater relevance.  The Dow consists, by and large, of very mature, slow-growing firms.  No Google, no Apple, no Qualcomm, no Amazon.  No biotech.  Home Depot and Wal-Mart are the only retail stocks.

3.  Dow construction is weird.  For reasons best known to himself, Charles Dow decided not to make a stock’s total market value be the measure of its weighting in the index.  He chose the per share stock price instead.  It was probably easier to calculate.

This decision has dire effects on the index.  A 1% change in the price of IBM, whose shares sell for $206 each, counts for over 7x as much in the index as a 1% change in the price of Microsoft, even though the two companies have almost the same total market value.  Why?  It’s solely because MSFT has split its stock in the past to maintain an “affordable” share price, and sells for $28.35 a share.

That’s right.  In the Dow, even a tiny company with a high per-share stock price would outweigh a behemoth with a gazillion shares at a low per-share price.

At least Charles didn’t pick the number of letters in the company name as the weighting factor.

how to game the Dow

What would I do if a pension fund hired me as an equity manager and gave me the Dow as a benchmark?  First of all, I’d have to consider whether I’d want to work for a bunch of nut jobs.  But the strategy for beating the index is clear.  I’d try to figure out whether we’re in an up market or a down market (harder than it sounds).  If the former, I’d buy the ten companies with the highest per share stock prices.  If the latter, I’d pick the ten with the lowest prices.  Unless you were unlucky enough to have epic  clunkers like Bank of America or Hewlett-Packard (both Dow components), nothing else would matter.