Sample breadth is mixed.
A research sample, not the market. The index is loading; read the desk top-down: posture, then the participation series, then the coverage panel that states exactly what this dataset can and cannot support.
What the index is not telling you.
Every participation measure this dataset can support, on one screen, with its full history behind it. Breadth turns before price. With this sample the question is direction, not level.
Where the universe sits.
Each name is classified into one of four trend stages from its price structure and moving-average geometry. Stage 2 pays. The migration between stages — not the level — is the leading information.
Participation, strength, liquidity.
Three things decide whether a sector move is investable: how much of it is working, whether it is beating the benchmark, and whether you can get out. Turnover is the honesty test.
Emerging, and quietly leaving.
A market that cannot produce new leaders is a market being distributed. Leadership is counted both ways: names entering Stage 2 above the 80th RS percentile, and names losing their 200-day.
Description, not forecast.
A regime is a description of present conditions with a memory of what followed similar ones. Five conditions must agree before the label changes; the conditions are shown, so you can disagree with them.
How every figure is built.
Read this before you trade from the terminal. Definitions, exclusions and the places where the model is known to break are the difference between a research tool and a set of charts.
This build runs on one real dataset: 50 Bursa equities across 1,474 daily sessions, retrieved from a public reference endpoint and validated before use. Nothing is fetched live during a session, and nothing is estimated from an index return. Every screen prints the provider, the window and the coverage it is working from.
Universes and exclusions
The configured Bursa universe is 1,066 eligible ordinary equities, drawn from 1,143 listed instruments (77 excluded as REITs, funds, SPACs, Leap Market or incomplete listings). The dataset loaded in this build contains 50 of them — 4.7% of the eligible universe — so every participation figure is a 50-name statistic and is labelled as such on every screen. The rules below are the full-universe rules; they are implemented and tested, but this dataset cannot exercise them:
- Securities with fewer than 200 trading sessions of history (new listings).
- Securities suspended on the calculation date.
- Securities below RM 0.05 or with nil volume in the trailing 20 sessions.
ACE Market names are included in all breadth measures and are not included in the KLCI comparison series. REITs, business trusts, closed-end funds, SPACs and Leap Market issues are excluded from the eligible universe; warrants, loan stocks, preference shares and ETFs are excluded as well — they are not operating businesses and their moving averages are derivative of the underlying.
Moving-average participation
For each name, the simple moving average of the adjusted close is computed over 20, 50, 150 and 200 sessions. The participation figure is the percentage of the universe closing above that average. Adjusted close accounts for splits, bonus issues and rights; cash dividends are not adjusted, which lifts long-horizon participation marginally for high-yield names — roughly +0.3pp on the 200-day measure.
Advance / decline
Advancers and decliners are counted on the closing price change against the previous close. A stock unchanged to the tick counts as unchanged, not as an advance. Untraded names are excluded from the ratio and reported separately. The cumulative A/D line is the running sum of net advances in thousands of names, rebased to zero at the start of the file; its shape is meaningful, its level is not.
New highs and lows
52-week closing extremes, computed on adjusted close, requiring at least 200 sessions of history. A name at its 52-week high counts once, not once per day of the run.
Trend stage classification
Each name is assigned at the close using price relative to a rising or falling 200-day average, the slope and order of the 50- and 150-day averages, and the position of price within its own 26-week range:
- Stage 1 — basing. 200-day flat, price crossing it repeatedly, range contracting.
- Stage 2 — advancing. Price above a rising 200-day, 50 above 150, within 25% of the 52-week high.
- Stage 3 — topping. 200-day flattening, 50 crossing below 150, range expanding to the downside.
- Stage 4 — declining. Price below a falling 200-day, 50 below 150, more than 20% off the high.
Relative strength and the RS percentile
RS is the 126-session total return of a name divided by the same measure for the FBM KLCI. The percentile ranks that ratio against the full universe on the same date. Momentum is the 21-session change in the percentile — a name can be strong and getting weaker, which is the most expensive combination in the market.
Breadth score
A single number, 0–100, equal to the mean percentile of nine measures: the four participation horizons, the advance/decline ratio, the 10-day A/D ratio, net new 52-week highs, the Stage 2 population and the McClellan oscillator. It is a summary, not a signal. Equal weights; no optimisation; no curve fitting.
Regime classification
Six conditions are evaluated at each close; a regime changes only when four agree for five consecutive sessions. The labels are descriptive of the present, never predictive. Transition frequencies quoted on the Regime page are the observed historical frequencies within this file.
Known limitations
Fifty names is not a market. Participation across 50 large caps moves several points on a single stock, so the direction of a series is readable and its level is not. This is a validated pipeline and a research sample, not a breadth product.
A bar dated today, read before the venue close plus a settlement margin, is a live print. Such bars are dropped before any calculation and the count is shown on the desk; the series therefore ends at the last settled session.
The classifier cannot see intraday structure, so a stock that recovers above its 200-day in the afternoon is counted as below it.
Equal weighting over-weights micro-caps relative to their economic significance. A 20-name micro-cap rally can lift participation without moving a single institutional portfolio.
Sector taxonomy follows Bursa's own classification, which is coarse: "Industrial Products & Services" contains both glovemakers and pipe manufacturers.
Stage classification, sector analytics, leadership ranking and regime classification are not published for this dataset. The history is sufficient (1,474 sessions), but 50 large caps carry no sector taxonomy and no ranked universe, so population shares and leadership rankings would be sample statistics dressed as market structure. Those sections state their absence rather than displaying a placeholder.