A three-layer audit: regime, breadth, and group RSI

Before a name could stand alone

A trading floor that seated equities, convertibles, corporates, and municipals within earshot was used to argue that a move in one market often changed another. Foreign investor flows into and out of U.S. markets were tabulated from the early 1970s.

That seating was the practical case for intermarket analysis: reading equities against bonds, policy rates, currencies, and cross-border flows so one name sits inside a weeks-to-months market regime. The stock chart was not the first document. It was the last one allowed to speak.

Name the bond-rate-currency regime

At an equity low, an accommodative policy and bond-market setting was treated as required context. Once activity recovered, stronger stocks were not assumed to keep confirming bonds, because credit demand would create friction. The regime could help the low and later stop helping the leaders.

Major foreign indexes were watched daily because capital was said to move faster across borders. A strong dollar was treated as a magnet for foreign buying of U.S. stocks and bonds that would, if extreme and persistent, later pressure export-oriented industrials. The 1987 foreign-flow reversal was attributed to a currency clash rather than to derivatives creating the trend.

Test whether the move is widely shared

Market breadth meant participation measures: advancing versus declining issues and volume, new-high versus new-low differentials, most-actives, and climax percentages. They were used to test whether an average's move was widely shared.

By the late 1960s leadership had already narrowed, so breadth peaked years before the 1972-73 rally. Averages printed new highs while former leaders built large non-accumulation tops and then lost support. A new high in the average was not treated as proof that the list was healthy.

The 1974 low was framed as a tested selling climax. Repeated sessions with more than 70% of issues advancing, plus heavy upside volume, were treated as unusual demand and later recast as advance-to-decline issue and volume ratios.

Weight confirmation over any single reading

Daily conclusions were assembled from a market diary plus rate-of-change and second-order acceleration on major indexes, high-low differentials, most-actives, block activity, multi-length Arms averages, 10-day advance/decline series, and on-balance volume. Weight sat on confirmation and divergence rather than on any single reading.

Rate-of-change acceleration meant a 12-month rate of change on a market average, then the rate of change of that series, used to spot fading momentum while price is still rising.

A 10-day average of the new-high minus new-low differential was marked stretched above +150 and washed out below -150. A 10-day Arms average was marked stretched at 0.80 and washed out above 1.25. On-balance volume added New York volume on up closes and subtracted it on down closes to confirm price direction.

Require the group's relative strength index

Weekly group work used a 15-period relative strength index on a five-year sample, with stochastics and standard-parameter MACD as backups. The relative strength index is a bounded oscillator on a fixed lookback of ordered group or index observations. It was used to compare groups and judge whether leadership was improving.

Candidates then had to sit above the 10-week and 30-week moving averages with a positive comparative-relative-strength reading. Comparative relative strength is a ranking of one price series against another, distinct from the oscillator, and was used on weekly charts to see whether a stock or group was outperforming. The oscillator and the ranking had to agree. The averages had to agree as well.

Keep leadership spreads in view

A capital-goods-to-consumer-goods ratio was kept because long-dormant capital-goods bases were appearing while 1980s consumer-growth leadership had not clearly topped. The ratio asks whether dormant producer and industrial groups are taking leadership from long-running consumer growth names.

Small-capitalization strength was treated as able to persist even if the industrial average declined, as in 1977. An average could sag while another sleeve stayed firm. That was another reason not to let one benchmark stand in for the whole book.

Editorial. Those spreads keep rotation visible. They are not read here as a standing instruction to own capital goods, consumer names, or small-capitalization stocks.

Leave on the same weekly rules

A break of the 30-week average, often joined by fading comparative relative strength, was the practical exit line for trading accounts. Longer-horizon accounts added a company review only if large distribution appeared.

Editorial. The same nest applies on the way out. A broken weekly average is read with fading group rank, with breadth that may already have narrowed, and with a bond-rate-currency regime that may no longer be accommodative. The archive is a sequence of checks, not a license for a lone chart.