Industry leadership carryover as a bull-regime test

Early-advance breadth and the regime question

In the opening months of a well-defined advance, most industry groups tend to rise together. That early-advance breadth later gives way to a split among continuing leaders, choppy names, and laggards.

After seven months of 1984 weakness, a three-session August advance of nearly 100 Dow points on record volume reopened a practical question: had a new bull market begun, or had an earlier advance resumed?

Leadership change as a hallmark

A contemporaneous market comment treated a change in industry leadership as a hallmark of a new bull market. It contrasted an energy-led late-1970s advance with a consumer-led early-1980s advance.

Relative-return tables for those two advances showed energy groups absent from the late-1970s top three, while consumer-oriented groups were strongly represented among the early-1980s top ten.

Overlap between those two top-ten lists still exceeded what a random draw from about 70 groups would imply. That left open whether the later advance was a new cycle or a second phase of the earlier one.

The leadership-reset hypothesis

Industry-return histories beginning in 1958 supplied six major bull-market cycles and five leadership transitions, with an average of 61.3 groups surveyed.

Across those five transitions, observed top-ten group overlap supported the leadership-reset hypothesis: a distinct new bull-market regime is more plausible when the top-performing industry groups differ from those that led the prior advance.

A 35-year review of 39 industry groups did not reveal reliable group-level cycles beyond a general tendency to follow the market averages.

Industry carryover inside a double-header cycle

A double-header cycle is a bull-market interval that unfolds in two similar up-phases separated by an interruption, so second-phase leadership is compared with first-phase leadership rather than only with the previous cycle.

In three historically similar two-phase advances, first-to-second-phase top-group overlap exceeded a random baseline. The phase-potency ratio in those cases put the second-phase percentage rise at about seven-eighths of the first-phase rise, over about 13 months.

All 73 reported industry groups rose above the 1982 low, though individual issues could still have failed to do so.