January 2026
This note documents the empirical results of a trend persistence study based on a triple structural filter applied to daily equity price data.
The purpose of this analysis is not return forecasting.
Instead, the goal is to quantify:
- how long statistically “valid” trend regimes persist,
- how much internal drawdown they typically experience,
- how often they recover after stress events,
- and how fast downside and recovery processes unfold.
This information is intended to support risk-aware screening, regime management, and option-selling decisions.
Dataset Overview
The study analyzes 4,104 individual trend episodes extracted across a large equity universe using a triple structural filter.
Each episode corresponds to a contiguous price segment where:
- both fast and slow rolling slopes are positive,
- distance from the 200-period moving average is constrained (≤ +20%),
- and the regime ends when at least one slope turns non-positive.
All statistics are computed at the regime level, not on raw daily data.
Regime Duration Characteristics
The distribution of trend duration shows meaningful persistence:
| Percentile | Bars (Trading Days) |
|---|---|
| 10th | 24 |
| 25th | 53 |
| Median | 90 |
| 75th | 159 |
| 90th | 257 |
Interpretation
Most filtered regimes are not short-lived noise:
- The median regime lasts roughly 90 trading days.
- One quarter persist longer than 159 days.
- The longest deciles extend beyond 257 days.
This confirms that the triple structural filter isolates structurally persistent trends, not transient fluctuations.
For systematic strategies, this persistence creates time windows long enough to justify multi-week position structures.
Performance Potential vs Internal Volatility
Maximum Favorable Movement
Inside each regime:
- Median max return: +10.35%
- 75th percentile: +26.18%
- 90th percentile: +53.41%
These values indicate that the filtered regimes capture meaningful upside potential.
However, this upside is accompanied by significant internal volatility.
Maximum Drawdown Inside Regimes
Even within positive trend regimes:
- Median max drawdown: −14.61%
- 75th percentile: −10.78%
- 10th percentile (worst cases): −26.92%
Interpretation
Trend regimes are not smooth.
Double-digit drawdowns are typical even when the overall trend remains positive.
This confirms that any strategy relying on trend alignment must explicitly manage internal stress and volatility.
Recovery Probability After Drawdowns
A central question is how often regimes recover after internal stress events:
| Drawdown Level | Recovery Probability |
|---|---|
| −10% | 63.1% |
| −15% | 51.4% |
| −20% | 42.9% |
Interpretation
- After a −10% drawdown, roughly two out of three regimes eventually recover to break-even.
- At −15%, recovery becomes about as likely as not.
- At −20%, the regime is more likely to fail than to recover.
This identifies a practical stress boundary between temporary stress and structural regime breakdown centered between −15% and −20%.
Recovery Time Dynamics
While recovery probabilities remain moderate for mild drawdowns, the time required for recovery is long.
After −10% Drawdown
- Median recovery time: 78 bars
- 25th percentile: 48 bars
- 75th percentile: 124 bars
Probability of recovery within short windows:
| Window | P(Recovery) |
|---|---|
| 11 bars | 0.5% |
| 18 bars | 2.9% |
| 25 bars | 6.3% |
After −15% Drawdown
- Median recovery: 122.5 bars
- Recovery within 25 bars: ~1.6%
After −20% Drawdown
- Median recovery: 151 bars
- Recovery within 25 bars: ~1.0%
Interpretation
Recovery, when it happens, unfolds over many months.
Short-term mean reversion is statistically rare after significant stress events.
This creates a strong temporal asymmetry:
- downside develops relatively quickly,
- recovery unfolds slowly.
This asymmetry is one of the most important structural findings of the study.
Drawdown Timing: Speed of Downside
While recovery is slow, drawdowns occur relatively fast.
Probability of Hitting Drawdowns Within 25 Trading Days
| Drawdown | P(DD ≤ level within 25 bars) |
|---|---|
| −10% | 24.6% |
| −15% | 14.3% |
| −20% | 11.3% |
Interpretation
Within one trading month:
- roughly one in four regimes experiences at least a −10% drawdown,
- more than one in ten experiences at least a −20% drawdown.
This confirms that stress events are front-loaded in time.
Operational Implications
Trend Filters Do Not Eliminate Drawdowns
Even well-filtered regimes regularly experience double-digit drawdowns.
Trend confirmation alone is insufficient as a risk control mechanism.
Short-Horizon Recovery Should Not Be Assumed
The probability of recovering from stress within typical option expiries (2–4 weeks) is extremely low.
This invalidates strategies that rely on fast mean reversion to “save” short premium exposure.
Structural Risk Thresholds Exist
The sharp degradation in recovery probability beyond −15% suggests a natural regime stress boundary.
This provides an empirical basis for:
- risk flags,
- dynamic position management rules,
- regime invalidation thresholds.
Methodological Interpretation
This study is intentionally simple and operational:
- it does not forecast returns,
- it does not optimize parameters,
- it does not attempt statistical inference beyond regime behavior.
Instead, it measures observable path-dependent behavior inside filtered trend regimes.
The output should be interpreted as:
an empirical risk and time profile of trend-conditioned price dynamics.
Closing Note
The main question addressed by this work is simple:
How fragile are trend regimes once they experience internal stress?
The answer is clear:
- trend persistence exists,
- upside potential exists,
- but downside shocks are frequent, and recovery is slow.
This asymmetry is not a flaw of the model — it is a structural property of financial markets that systematic strategies must explicitly account for.
AIQ Notes
Independent Trader · AI-assisted Coding & Systematic Analysis
G. D. P.