The most quoted index in the world and the least defensible. Thirty companies, weighted by share price, chosen by committee — a nineteenth-century instrument that outlived every argument against it.
Charles Dow, co-founder of The Wall Street Journal, needed a quick way to tell readers whether the stock market had gone up or down. In 1896 he took twelve industrial companies, added their share prices, and divided by twelve. That was the entire methodology, and in substance it is still the methodology today.
Every serious criticism of the Dow is correct. It is too small, its weighting scheme is arbitrary, and its selection is discretionary. It remains the number on the evening news.
In a price-weighted index, a company's influence is determined by its share price and nothing else. Market value is irrelevant. A trillion-dollar company trading at $40 moves the index less than a much smaller company trading at $400.
The consequence people find hardest to accept: a stock split changes a company's index influence without changing anything about the company. Split four-for-one and your share price quarters, so your Dow influence quarters too. Apple did exactly this in 2020, immediately reducing its sway over the average despite being the same business the following morning.
The Dow is no longer the simple average Charles Dow computed. Every split, substitution and spin-off would create an artificial jump in the index, so the sum of prices is divided by a continuously adjusted figure — the Dow divisor — rather than by 30. The divisor has been revised downward so many times over 130 years that it is now well below one, meaning a one-dollar move in any constituent shifts the index by considerably more than one point.
The index famous for continuity has almost nothing in common with its founding self. None of the original twelve industrials remains. General Electric lasted longest — more than a century with interruptions — before removal in 2018.
Changes continue at pace. In late 2025 Nvidia replaced Intel and Sherwin-Williams replaced Dow Inc. In mid-2026 Alphabet was added and Verizon removed. Published counts of how many times the roster has changed since 1896 differ between sources, so treat any specific figure with caution — what is not in dispute is that the churn has been constant.
The pattern in those swaps is legible: chipmakers and platform companies in, telecoms and legacy industrials out. The committee's stated aim is that the average should represent the American economy, and the composition is a running record of what they think that economy is.
Recognition and continuity. It has been published since 1896 and the public knows it. For measurement purposes the S&P 500 is better in every respect, which is why professionals use it.
Loosely at best. Thirty large companies, weighted by share price, cannot represent an economy where most employment sits in smaller and private firms.
Only the Nikkei 225. Both survive on the same logic — too old and too famous to change without breaking the historical series.