Where money moved across the 30 PSX sectors, and how each sector's stocks are currently positioned, jump to the rotation table. End-of-day data; educational analytics, not investment advice.
Each row is the median of that sector's liquid constituents, a stock counts only if it averaged 50,000+ shares a day over the last 20 sessions, so one thin ticker cannot move a sector. Medians, not averages, so a single outlier cannot either. This is a record of what already happened; it is educational analytics, not investment advice and not a buy list.
| # | Sector | Score | Velocity (WoW pp) | 5d | 20d | 60d | 250d | Volume 5d/20d | MTF aligned | Quality phase | Foreign net (MTD) | Leaders (5d) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Loading sector rotation… | ||||||||||||
A high score means money has already moved into a sector. It is not evidence that money will keep moving there. We tested exactly that question against 10.6 years of our own price database, and the honest answer is on the record below.
Method. 75 of the score's 100 points come from price and volume alone (20-session performance 30, week-over-week acceleration 25, volume momentum 20) and can be replayed at any past date. The other 25 points (trend alignment 15, operator phase 10) come from snapshots that keep no history, so they cannot be replayed and are excluded. At each of 124 non-overlapping rebalance dates from 2017 to 2026 we scored every sector using only data available on that date, then measured each sector's actual median constituent return over the following 20 sessions, 3,176 sector-date observations in total.
| Score bucket (of 75) | n | Mean next 20d | Median next 20d | Positive |
|---|---|---|---|---|
| 0–19 pts | 1,289 | −0.25% | −0.33% | 48.3% |
| 20–34 pts | 921 | +0.86% | +0.03% | 50.2% |
| 35–49 pts | 478 | +1.14% | +0.19% | 50.6% |
| 50–64 pts | 305 | +0.74% | −0.23% | 47.9% |
| 65–75 pts (highest) | 183 | −0.97% | −1.03% | 44.3% |
| All sectors | 3,176 | +0.34% | −0.20% | 48.9% |
Result. The highest-scoring bucket produced the worst forward 20-session return of any bucket and the lowest hit rate. Ranking sectors into quintiles each date and measuring excess return over that same date's median sector, the top quintile trailed the bottom quintile by 1.05 percentage points over the next 20 sessions (n = 591 per quintile, Welch t = −2.36). Across other horizons the spread is statistically indistinguishable from zero: −0.25pp at 5 sessions (t = −1.07), −0.23pp at 10 (t = −0.72), +0.01pp at 40 (t = +0.01), +0.31pp at 60 (t = +0.31). No horizon showed a positive edge, and the single significant result points the wrong way. That is why nothing on this page tells you to buy anything.
Limits of that test. Sector membership is taken from today's classification, so companies that delisted are absent. A survivorship bias that flatters the past. Trading costs are not deducted; a PSX round trip costs roughly 3.04 percentage points, which is larger than every spread above. Rebalances are non-overlapping but sectors move together, so the effective sample is smaller than 3,176.
What the numbers below are still good for: describing breadth and dispersion, which parts of the market participated in a move and which sat it out, and flagging sectors where volume behaviour changed. Treat every row as history, not as a forecast.
| Sector | Stocks | FL + Launching | Distributing | Avg Confidence | Top Signal | Status |
|---|---|---|---|---|---|---|
| Loading... | ||||||