2014issue C0242-54
Equal-weight sector ranks as a rotation procedure
The 2014 case built market direction from 15 equally weighted sectors, ranked those sleeves on short, intermediate, and long clocks, and treated a single name as the last layer after sector regime and market breadth were in view.
- Build each equal-weight-sleeve-price from reference-normalized constituent prices so the 15 sectors, their industry groups, and the market can be ranked without one name dominating.
- Use rank-rotation on the 10-, 20-, and 50-day clocks and study quartile-rank-history, not a one-day snapshot, so a weak sleeve can be seen improving before it reaches the top bin.
- Place a single name last in the industry-rotation map: market direction first, then sectors moving with that direction and correlated to it, then how large each sector move was versus the others.
- Editorial: keep market-breadth-context beside the market price, and keep a single-name idea off the book until rank improvement and those internals pass the same test.
The name sat last
The 2014 case treated a single name as the last layer of a nested map. Market direction came first. Next came the sectors that were moving with that direction and were correlated to it. Only then did the case ask how large each sector move was versus the others.
That order is industry-rotation in the archive sense: a portfolio map that places a single name inside equally weighted industry groups and sectors so the idea is judged against group and sector regime, not in isolation.
How the sleeves were built
Market direction was built from 15 equally weighted sectors. Each sector price was averaged from equally weighted industry-group prices. Each group price was averaged from constituent prices divided by a reference price. That equal-weight-sleeve-price kept any one name from dominating the sleeve.
Names that did not move with their assigned industry group were parked in a noncorrelated set and later rechecked for membership in a group they did resemble.
Three clocks and a rank path
Sector momentum was the ratio of the current sleeve price to the price 10, 20, or 50 days earlier, labeled short, intermediate, and long. Rank-rotation ranks sleeves on those stated horizons and turns that rank path into one testable rule for entering, exiting, or standing aside.
The object of study was rank history, not a one-day snapshot. Each of the 15 sectors could be ordered by rank or by raw momentum, on daily or weekly bars, for one horizon or all three.
When the three horizons had to become one number, a momentum-rank-score compressed them: three times the short rank plus twice the intermediate rank plus the long-term rank.
Quartile color showed improvement
Rank cells were colored by quartile, with the top quartile green and the bottom quartile red. Quartile-rank-history is that time-ordered color map. A sleeve that had been weak could be seen improving before it reached the top bin.
One documented international-sector row sat through six bottom-quartile readings, then a second-quartile week, then a top-quartile reading in the week of 16 August 2013.
Group action and market breadth
Industry-group action was characterized as quartile changes in momentum rank, the largest rank changes, and power shifts or pivots, with a correlation pane comparing the sleeve with a broad equity index.
Market-breadth-context sat beside the market price as counts of groups above 10-, 50-, and 200-day averages, group advance-decline and volume, an Arms-style TRIN, and new-high versus new-low series. Those internals were the check on whether the whole market could support a rotation.
The same test at every layer
Editorial: the written test does not change when the map moves from market to sector to name. A name stays off the book until the sleeve shows rank improvement on the chosen clock and market-breadth-context does not contradict that improvement.