Generalized Protective Momentum
The Generalized Protective Momentum (GPM) strategy was developed by Wouter Keller and Jan Willem Keuning as an evolution of their earlier momentum and trend-following frameworks. It builds on the concept of using momentum signals to select among a diverse set of asset classes while applying a crash-avoidance filter that moves the portfolio to safe assets when market conditions deteriorate. The strategy is documented in Keller and Keuning's publicly available research papers on SSRN. A related strategy from the same authors is Vigilant Asset Allocation G12.
Investment Philosophy
GPM applies momentum ranking across a broad universe of risky assets -- including global equities, bonds, real estate, and commodities -- and combines this with an absolute return filter to avoid holding assets in sustained downtrends. The "generalized" aspect refers to the framework's flexibility: the asset universe, lookback periods, and crash protection parameters can be adjusted within the model's design. The underlying philosophy is that momentum is a persistent, cross-asset return anomaly, and that systematic crash protection can improve risk-adjusted outcomes without sacrificing long-run returns.
Who It's For
This portfolio suits investors who are comfortable with quantitative, rules-based strategies and understand that systematic approaches can underperform in certain market environments. It requires monthly monitoring to implement momentum signals and a willingness to hold cash or bonds when the model dictates a defensive posture.
Pros
- Broad asset class universe provides diversified exposure to the momentum signal
- Crash protection filter helps limit severe drawdowns during prolonged market downturns
- Fully systematic and rules-based, removing emotional bias from allocation decisions
Cons
- Complex to implement compared to static buy-and-hold portfolios
- Can underperform in choppy markets where momentum signals are frequently reversed (whipsaw)
- Multiple adjustable parameters create the risk of overfitting if the strategy is modified based on historical backtests
Technical Notes
Keller and Keuning publish updated research and parameter details through their papers on SSRN. The strategy requires monthly rebalancing and access to performance data for the constituent asset classes.
We cover a curated selection of portfolios in this database.
workspace_premiumSee what's coveredCancel anytime. No lock-in.
Average Allocation
Based on historical average weights across all rebalance periods.
Performance Snapshot
Rolling Returns
| Period | Low | Average | High |
|---|---|---|---|
| 1 Year | -8.4% | +12.2% | +68.5% |
| 3 Year | +0.6% | +12.0% | +36.0% |
| 5 Year | +1.6% | +12.3% | +25.5% |
| 10 Year | +4.3% | +12.3% | +23.5% |
Growth of $10,000
Historical Drawdown
Percentage decline from the portfolio's peak value at each point in time.
Rolling Returns
Annualised return for each rolling period ending on that date.
Annualised return for each 1Y period ending on that date.