SIS Momentum Intelligence
Savaria Corporation · Industrials
See the direction. Understand the journey.
Uptrend taking shape.
4 of 4 horizons are positive; 0 are negative. The last three months contain substantial movement but little net progress.
1 month
4 weeks3 months
13 weeks6 months
26 weeks12 months
52 weeksThe small charts trace each horizon's momentum reading over the last 26 weeks.
One price. Four perspectives.
Follow the price alongside the direction of each horizon. Select a week to inspect the evidence at that close.
Strip colours show direction, not a trading signal. The flat band is −1% to +1%; darker colours indicate a larger move relative to recent volatility. *Unavailable cells are marked with a dash.
Readings use completed weeks only.
Positive returns extend across most measured horizons. This describes the stock's own price history, independently of sector rankings.
What followed similar conditions?
Same instrument, directional balance and path quality. Only outcomes completed by the inspected date.
13 weeks later
6 observations6 of 8 required observations. No aggregate estimate is shown.
26 weeks later
6 observations6 of 8 required observations. No aggregate estimate is shown.
52 weeks later
3 observations3 of 8 required observations. No aggregate estimate is shown.
At least 8 non-overlapping observations are needed for aggregate statistics. Small samples remain inconclusive; even larger samples are historical descriptions, not probabilities or expected returns. Downside is the lowest weekly close relative to the starting close.
Momentum is direction.
Quality is the journey.
A rising price can have weak momentum at longer horizons. A falling stock can rebound sharply while its broader history remains negative. Reading both together makes those differences visible.
Three or four horizons on one side suggest a broader directional pattern.
Shorter horizons often move first. Disagreement identifies a change to investigate.
Orderly movement describes consistency. It does not tell us that a trend will continue.
Compare outcomes with ordinary periods too. A positive outcome alone does not establish an advantage.
Week by week
| Week ending | 1m | 3m | 6m | 12m |
|---|---|---|---|---|
| 2026-09-18 | +2.1% | +1.2% | +16.4% | +42.9% |
| 2026-09-11 | -3.1% | -3.0% | +14.0% | +39.5% |
| 2026-09-04 | -6.8% | -2.7% | +11.4% | +39.3% |
| 2026-08-28 | -3.0% | -0.6% | +12.2% | +37.7% |
| 2026-08-21 | -7.5% | -1.2% | +12.1% | +38.0% |
| 2026-08-14 | -6.7% | +4.0% | +16.0% | +39.4% |
| 2026-08-07 | -2.2% | +7.5% | +19.8% | +48.3% |
| 2026-07-31 | -3.5% | -1.3% | +23.5% | +49.4% |
Month labels represent 4, 13, 26 and 52 calendar weeks. Values are local-currency price changes.
Research foundationWhat this page measures—and how to read it
Own-history momentum
Each horizon measures the percentage change in this instrument's saved closing price. Readings above +1% are positive, below −1% are negative, and between those values are flat. Three positive or negative horizons establish the directional balance; “taking shape” covers its first six consecutive weeks.
One-, three-, six- and twelve-month labels use 4, 13, 26 and 52 weeks. These are price returns, excluding dividends, cash interest, fees and borrowing costs. This is an analytical adaptation, not a replication of an excess-return futures strategy.
Quality, risk and comparisons
Path efficiency is the absolute net log-price change divided by the sum of absolute weekly log changes over 13 weeks. At least 35% is orderly, 15–35% uneven, and below 15% choppy. These are descriptive thresholds, not fitted probabilities. Volatility uses 26 weekly log returns, annualised by √52.
Historical matches share the inspected balance and quality category. Within each horizon, samples are chosen chronologically without overlap. “All conditions” uses separate non-overlapping periods with sufficient history. Comparisons are unadjusted price outcomes, not strategy returns; the samples can cover different market environments.
Dates and data integrity
Only completed, published weeks are used. Missing or invalid prices and unresolved stock splits reset the calculation. Known full exchange closures carry the preceding close. Saved price repairs automatically refresh the report; historical readings reflect today's saved dataset, rather than an archived record of what was known then.
The chart defaults to a logarithmic scale so equal percentage changes have equal visual height. Replay recalculates the narrative and withholds outcomes that had not finished by the selected date.
Read the underlying research
Moskowitz, Ooi & Pedersen · Time series momentum (2012) ↗
Hurst, Ooi & Pedersen · A Century of Evidence on Trend-Following Investing (2017) ↗
The studies examine diversified markets and volatility-scaled strategies. They do not validate these individual-stock classifications or establish a trading edge for this page.