Nvidia to the Dow.  Yawn…

One of the most widely discussed market topics this morning is that Nvidia (NVDA) will be replacing Intel (INTC) in the Dow Jones Industrial Average (INDU).  My reaction: so what?  We shouldn’t care very much about an anachronism.

Despite its flaws and relative lack of relevance, the Dow Jones Industrial Average remains widely referenced.  It’s understandable why the non-financial media might continue to define stock market movements with this historical benchmark, but it continues to play a key role in financial media reporting.  Faithful readers will notice that I almost never mention the Dow.  Years ago, I wrote a piece entitled “It’s Time to Retire the Dow”, and I stand by that opinion.  (Unfortunately, I can’t link to it – several years ago we revamped our website and older articles were lost.)  My reasoning was and is:

1. INDU not actually an index – it’s an average.  That reflects the technology available to its eponymous creator, Charles Dow, in 1896.  In 1884 he created the Dow Jones Transportation Average to predominantly track railroad stocks, the “high tech” sector of that time, though it included Western Union and a Steamship Company.  Prior to Mr. Dow’s innovation, there was no easy way to define whether the “market” was up or down.  He created a measure that could be calculated with a pencil and paper – an arithmetic average.  Like the Transportation Average, INDU originally had 12 stocks, and the number of components was only raised to 30 in 1928, near the end of the “Roaring Twenties”.    The periodic additions and subtractions to the index are meant to reflect the index committee’s best estimate of a small portfolio that represents the US market, but that means it is both narrow and arbitrary.
2. Because INDU is actually an average, it is therefore price-weighted, not market capitalization weighted. 
   • There is a reason why most major global indices, including the S&P 500 (SPX) and Nasdaq 100 (NDX), use the latter technique.   The computer age allows us to continuously multiply a stock’s price times its shares outstanding and then divide the sum of those calculations to create an index.  The more money invested in a company, the bigger its weight.  It’s fair to argue whether some of these indices have gotten too top-heavy, but that reflects the market’s love for mega-cap stocks, not the index construction. 
   • Because an average is calculated by simply summing the component’s prices and then dividing the sum, it becomes price-weighted.  The higher a stock’s price, the larger its weight in the calculation.  To my mind, that is essentially random-weighted.  UnitedHealth Group (UNH) and Goldman Sachs (GS) have the two highest weights in INDU by virtue of their $500+ stock prices.  That gives them nearly 12x the weight of $41 Verizon (VZ).  This is despite GS and VZ having nearly identical market capitalizations of around $170 billion.  Does that make sense to you?  Apple, which is the largest US company by market capitalization, is currently #12, just behind Travelers (TRV) and ahead of JPMorgan (JPM).  When NVDA and Sherwin-Williams (SHW) are included next week, SHW will have nearly 3x the weight of NVDA despite the semiconductor company having about 36x the market cap of the paint company and seemingly an infinite more mindshare.
3. Follow the money.  The market has voted as to which type of index construction it prefers, and there is FAR more money invested in products that track SPX and NDX than INDU.  The only significant ETF that tracks INDU is DIA, which has assets of about $35bn.  SPY and VOO, which track SPX, each have over $500bn.  And the gulf widens even further when we include mutual and pension funds.  The amount indexed to INDU is a relative pittance. 

It is noteworthy when a well-followed market measure changes its composition, especially when it does that infrequently.  But when it comes to real-world importance for the affected stocks, it is a relatively meaningless decision from an increasingly irrelevant metric.