Price-weighted construction distorts breadth, support, and trend

The average is not the market

A commonly cited industrial average uses thirty large-capitalization names chosen to stand for major U.S. business lines. Those firms' sales were described as about 20 percent of listed-company sales.

Market breadth asks a different question: how widely participation is distributed across listed stocks, as distinct from the path of a narrow headline average. The official print can move while most listed names do not confirm it.

How the official print is built

The average is a price-weighted average rather than a capitalization-weighted index. Equal percentage moves in higher-priced components shift it more than the same percentage moves in lower-priced components.

Component prices are summed and divided by an adjustable divisor so that component splits and related adjustments do not create a discontinuous official print. A component split lowers a stock's price. In this average it also shrinks the divisor, which magnifies later price changes and abruptly cuts that split stock's influence.

Fractional breaks of support and resistance

Some technicians treat a fractional print through prior support or resistance on this average as a timing signal. That practice is poorly matched to how the average is constructed.

Narrow composition and the ability of a few expensive names to be pushed at support and resistance mean a small group can force the official print through a remembered level. The average's arithmetic then records a break that broader gauges of overall and component performance need not show.

The same peak can imply opposite trends

After a mid-1987 close, nine components split over the next thirty months and the divisor fell from 0.783 to 0.555. That change altered how later official prints compared with the 1987 peak.

Recalculating the August 1987 official high with later split-reflected prices and the January 1990 divisor can place that 1987 level above subsequent official highs. One construction reading affirms a later new high. Another places the high in 1987, with a downtrend afterward.

Trend-following that reads successive highs and lows of the official print can therefore imply opposite continuations or reversals once splits and the divisor are restated. The prices are the same history. The construction is not.

Broader gauges of participation

Narrow composition, the ability of a few expensive names to be pushed at support and resistance, and the average's arithmetic are offered as reasons to reduce its role as a market barometer. Broader gauges of overall and component performance, and market-breadth readings of how widely participation is distributed, keep the question on the listed market rather than on one constructed print.