Hulic Co., Ltd.
The long-horizon signal remains in force, but the shortest horizon now points the other way. That is an early loss of persistence, not yet a regime flip.
When the close is above the boundary the primary state is positive; below it, negative. Regime shading records what was known at each completed week.
The zero line is the regime boundary. Distance from zero shows how much of the trailing return would have to unwind before direction changes.
Four means all tracked horizons agree with the current 52-week direction; lower readings reveal internal tension.
Each observation uses only the direction and horizon agreement available at that completed week. Samples are spaced by the measurement horizon to reduce repeated overlap. Fewer than five observations are shown but marked as too thin for a dependable inference.
| Forward window | Matched samples | Median move | Directional follow-through | Regime survived | Versus all periods | Read |
|---|---|---|---|---|---|---|
| 4 weeks | 9 | +3.6%range -9.5% to +6.9% | +77.8% | +100.0% | +2.5%median difference | Usable descriptive sample |
| 13 weeks | 4 | -1.3%range -8.0% to +10.6% | +25.0% | +50.0% | -4.5%median difference | Too little matched history |
| 26 weeks | 3 | -4.5%range -6.0% to +19.1% | +33.3% | +33.3% | -9.2%median difference | Too little matched history |
| 52 weeks | 3 | +2.9%range -8.7% to +60.6% | +66.7% | +66.7% | -10.1%median difference | Too little matched history |
| Direction | Start | End | Duration | Move inside regime |
|---|---|---|---|---|
| Positive | 2025-04-18 | 2026-07-31 | 68w | +23.1% |
| Negative | 2024-11-29 | 2025-04-11 | 20w | +7.8% |
| Positive | 2023-03-31 | 2024-11-22 | 87w | +38.7% |
| Negative | 2023-03-24 | 2023-03-24 | 1w | 0.0% |
| Positive | 2023-01-13 | 2023-03-17 | 10w | +4.5% |
| Negative | 2022-12-23 | 2023-01-06 | 3w | -1.2% |
| Positive | 2022-11-11 | 2022-12-16 | 6w | +2.4% |
| Negative | 2022-01-14 | 2022-11-04 | 43w | +6.5% |
| Positive | 2022-01-07 | 2022-01-07 | 1w | 0.0% |
| Negative | 2021-12-31 | 2021-12-31 | 1w | 0.0% |
| Positive | 2021-02-26 | 2021-12-24 | 44w | -5.2% |
| Negative | 2021-01-01 | 2021-02-19 | 8w | +4.4% |
The realised variability of weekly returns over the recent 26-week window.
Recent 13-week volatility relative to the broader 52-week estimate. Above 1 means risk has accelerated.
A capped 10% annual-risk normalization. It is a comparison tool, not a position recommendation.
The current boundary margin is not unusually small relative to recent weekly volatility.
Close-to-close movement since the present 52-week direction first became active.
The largest counter-trend move observed without the primary state changing direction.
Rising volatility can make an otherwise persistent direction harder to hold and increases the chance of crossing a boundary abruptly.
These are separate Sharemaestro readings from the same latest completed week. They can explain participation, value or relative pressure, but none is allowed to overwrite the transparent 52-week signal.
Positive when the latest adjusted weekly close is above the adjusted close exactly 52 completed observations earlier; negative when below; neutral only when equal. Discontinuous histories are excluded. This is a stock-and-ETF research adaptation, not a replication of a diversified futures strategy.
The same calculation is repeated at 13, 26 and 39 weeks. Their agreement measures coherence. They cannot veto or replace the 52-week state.
Forward returns start after each historical decision week. Matching uses the direction and number of aligned horizons known then, with spaced samples and visible counts.
Trend persistence can reverse, suffer repeated whipsaws, or cross its boundary by a gap. Transaction costs, tax, borrow availability and implementation slippage are not modelled.
Independent market research for education. Not investment advice, a recommendation, a forecast or an execution instruction. Past states and outcomes do not establish what will happen next.
Search by ticker symbol or company name and select a result to open its chart.