Stacked channel, trendline, and moving-average warnings in 1987

A sequencing drill, not a single pattern

Editorial frame: read the 1987 top as an order of operations. First establish that time, valuation, and a bond-equity split have aged the advance. Then require three chart conditions to fire together: contact with the upper rail of a price-channel, a break of a multi-tested trendline, and a violation of a long moving average. Use elliott-wave alternation only to reject a copycat-correction story.

Age the advance first

By August 1987, after a five-year advance, historically high price-to-earnings ratios and very low dividend yields were treated as evidence that the longer bull phase was unusually extended and overvalued.

A turn toward monetary restraint early in 1987 coincided with peaking bond and Treasury-bill prices, while equities kept rising into August. The archive described that split as typically resolving with stocks following bonds lower. Editorial label: this is intermarket-divergence, and it belongs in the aging step, not in the trigger stack.

Breadth failed to confirm price on both intermediate and short-term horizons. The NYSE advance-decline line lagged new industrial highs, the OTC advance-decline line lagged its composite, and the industrial close at 2722 was unconfirmed by the transportation average. An advance-decline line is a running total of advancing minus declining issues used to test whether an index high is broadly confirmed.

Upper-channel contact and distribution-volume

The first three strong advances after the September 1985 low each lasted about 12 weeks. The fourth advance reached that same duration in the second week of August as the industrials near 2700 approached the upper parallel of an upward channel.

A price-channel is a pair of parallel support and resistance rails around an advance. Editorial reading: contact with the upper rail after that measured run is a completion test. It does not, by itself, complete the stack.

Volume on August 11-13 exceeded any earlier three-day stretch and was read as distribution after an already extended rally, completing the setup for a correction. That reading is distribution-volume: a surge in activity after an extended advance, taken as supply overwhelming demand rather than fresh sponsorship.

Weight the trendline by prior turns

After August, the industrials broke trendlines in early September and October. The October break of a four-point support line near 2500 was treated as the stronger warning because that line had already turned price four times.

A trendline is a straight boundary across successive turning points. Each additional rejection raises the weight given to a later penetration. The same trendline logic was illustrated by the August 1984 advance, which followed an S&P 500 break of a downward resistance line that had turned price seven times.

The long moving-average violation

In the second week of October 1987, a sharp drop through the 200-day moving average preceded the collapse by several days, after the industrials had generally held above that average for the previous three years. A sequence of that kind was also noted before 1929.

A moving average is a smoothed baseline of past closes over a fixed lookback. Editorial reading: a sharp break after years spent above it is a regime-change warning. In this drill it is the third required fire, not a substitute for the aging step or the channel and trendline tests.

Alternation against a copycat dip

By early October many participants treated a 91-point industrial decline as a routine bull-market dip and expected a replay of the March-May reaction that had been followed by a 500-point advance.

Wave-structure alternation was used to challenge that copycat view. Two consecutive corrections inside the same bull market were judged unlikely to trace the same pattern. Editorial restatement: elliott-wave alternation says consecutive corrections inside one advance are unlikely to take the same shape, so the March-May template was not accepted as the working map.