Skip to main content
Track Industry rotation
2 / 23
Library

1988issue C111-9

Constructing industry-group breadth and rotation measures

An industry-group index is a composite of related stocks treated as its own small market. Once the group exists, advance-decline, high-low, diffusion, absolute-breadth, and overbought-oversold constructions can be calculated at the group level instead of only for a single name or the broad average.

  • Build each industry group as a composite of related stocks so advances, declines, and new highs and lows can be counted at the group level.
  • Overlay the group advance-decline line and the high-low index on the industry index to test whether a new high or low is still widely shared.
  • Classify the advance-to-decline ratio with fixed overbought and oversold position lines, then use diffusion and absolute-breadth readings to see whether participation is turning.
  • Apply the same five constructions to major exchange composites and to equal-weighted group indices so large-capitalization names do not dominate the path.
Entries in this reading3 entries

Rebuild the group as a miniature market

An industry group index is a composite of many related stocks. Once that composite is in place, advances, declines, and new highs and lows can be measured at the group level instead of only for the broad market or a single name.

Stocks in the same industry tend to move together, so a group can remain in its own advance while the overall market is declining. That shared motion is the construction reason for analyzing rotation at the industry-index level.

The group advance-decline line

The group advance-decline line is a cumulative total of the daily or weekly percent of advancing issues minus the percent of declining issues, then overlaid on the industry index.

When a weighted industry index makes new highs while the advance-decline line is already falling, the divergence is read as a top warning. Those comparisons are judged inside a one-year window because the line has a long-term downward bias independent of short-term index weakness.

The high-low index

The high-low index is the net count of 52-week highs and lows over five days for studies of one year or less, or over five weeks for longer studies, expressed as a percent of issues in the group so membership changes do not distort the series.

A new high or low in the group index that is not matched by the high-low index is treated as a late-stage move that may be near a reversal, because fewer members are still making new extremes.

Overbought and oversold position lines

The overbought-oversold construction is an advance-to-decline ratio plotted against fixed position lines. Readings above the upper line are labeled overbought. Readings below the lower line are labeled oversold.

Diffusion and absolute breadth

The diffusion index averages the percentage of advancing stocks over the prior 10 days or 10 weeks. A downturn from a high reading while the group index is still rising is treated as a negative participation reading, and the reverse reading is used at bottoms.

Absolute breadth is the absolute difference between advancing and declining counts, then smoothed with a 10-day moving average for studies of one year or less or a 10-week average for longer studies. High readings are treated as constructive. Very low readings are treated as weakening.

The same constructions on composites and unweighted paths

The same five breadth constructions can be applied to major exchange composites and to equal-weighted group indices that add the average percent change of every member. Each stock then has equal influence, and the unweighted path can turn before a capitalization-weighted index.

Educational research material, not investment advice. Historical source context does not establish present-day performance.
2 of 23 in the Industry rotation track
19921-16 pp.Next on Industry rotationTrendline holds, trailing stops, and industry rotationAfter a stock breaks out and begins an uptrend, the trendline is the holding rule: profits are allowed to run until a close through the line fails the hypothesis.
All readings on this track · 23 readings
  1. 1985Industry leadership carryover as a bull-regime test
  2. 1988Constructing industry-group breadth and rotation measures
  3. 1992Trendline holds, trailing stops, and industry rotation
  4. 1994Inflation-deflation regimes inside the stock cycle
  5. 1996Sector rotation across economic cycle phases
  6. 2001Rebased relative performance charts for sector rotation
  7. 2001Place a small-cap growth idea inside a regime map
  8. 2004Rebuild every industry as a share of one rank scoreboard
  9. 2004Rate-hike regimes and sector rotation as a case study
  10. 2005A two-name style-index sleeve makes rank rotation one procedure
  11. 2006Consumer staples after a smokestack cycle
  12. 2007An intra-sector regime split between builders and equity REITs
  13. 2008Country and sector weights in an Africa regional-sleeve
  14. 2011Trend permission, priced entries, and sector rotation
  15. 2012Construct a regime-aware context from sector rotation
  16. 2012Regime overlays versus rank rotation
  17. 2014Rank-based sector rotation as a portfolio test
  18. 2017Real estate as a ranked industry sleeve
  19. 2017Theme sleeves: liquidity and commission filters before industry rotation
  20. 2018Retail sleeve construction through channel rotation and daily leverage
  21. 2020Water sleeve construction: satellite size, industry mix, and liquidity
  22. 2020A ranked research terminal as a three-layer watchlist procedure
  23. 2020Regression channels for sector rotation context
All 29 readings tagged Industry rotation
Also on Industry rotation5 readings