GPMv Strategy Overview

GPMv is a breadth-driven, correlation-aware tactical allocation strategy operated by Randy Harris of Dual Momentum Systems. It is a Dual Momentum Systems variant of Generalized Protective Momentum (GPM), the strategy published by JW Keuning and Wouter Keller, modernized in its universe and tuned to stay invested slightly longer than the original while keeping the protective core intact.

Where most momentum strategies ask only "what is going up?", GPMv asks two questions at once: what is going up, and what is going up for reasons unrelated to everything else it owns. It then lets the breadth of the market decide how much of the portfolio should be at risk at all.

Core Strategy Logic

Every month, GPMv scores eleven risk assets on two dimensions, combines those scores into one number, and uses the result to set both the holdings and the size of the defensive position.

The Two Scores

Momentum. Each asset's momentum score is the equal-weight average of its trailing 1, 3, 6, and 12-month total returns. Blending four lookbacks rather than relying on one keeps the signal from being hostage to a single sharp month at either end of the window.

Correlation penalty. Each asset is then measured against the equal-weight return of the whole risk universe over the trailing twelve months. An asset moving in lockstep with everything else is penalized. An asset moving independently, or against the crowd, keeps its full score or better.

The two combine into a single composite: momentum multiplied by one minus correlation. A strong performer that is simply riding the same wave as the rest of the portfolio scores below a moderate performer that is genuinely doing its own thing. An asset with negative momentum scores negative regardless of how uncorrelated it is, because diversification is not a reason to own something falling.

Breadth Sets the Risk Dial

The strategy counts how many of the eleven risk assets carry a positive composite score. That count is the risk dial.

• When 10 or 11 score positive, the portfolio is fully invested.
• Each asset below that threshold moves one sixth of the portfolio, roughly 17%, into the defensive position.
• When 5 or fewer score positive, the portfolio is 100% defensive.

Risk comes off in steps rather than in a single binary switch, and the trigger is market breadth rather than the trend of any one index. A market where most things are still working keeps most of the portfolio at work.

The Risk Universe

Eleven candidates span equities, credit, real assets and duration:

• U.S. large cap (IWB)
• U.S. large cap growth (QQQ)
• U.S. mid cap (IWR)
• Asia-Pacific equities (VPL)
• European equities (VGK)
• Gold (SGOL)
• Broad commodities (PDBC)
• U.S. real estate (VNQ)
• High yield bonds (HYG)
• Investment grade corporates (LQD)
• Long-term Treasuries (TLT)

Whatever risk budget the breadth count allows is divided equally among the top three assets by composite score. At full investment that is three positions of roughly 33% each. The universe is deliberately broad so that the top three can land almost anywhere: gold and commodities and real estate compete on the same footing as equities, and in some months they win outright.

The Defensive Destination

The defensive position is intermediate-duration Treasuries (IEF), with one override. If IEF's own momentum score is negative, the defensive allocation goes to BOXX instead, a short-duration Treasury-bill alternative.

That override exists because the default assumption of every protective strategy, that Treasuries rally when equities fall, is not always true. In 2022 both fell together. When intermediate duration is itself in a downtrend, GPMv steps down to the short end rather than retreating into a second losing position.

What GPMv Changes From the Published Model

The original GPM is implemented separately in Dual Momentum Systems as a benchmark, reproduced faithfully and never improved, so that GPMv's results can be measured against something honest. GPMv makes five changes:

1. A modernized universe. IWB replaces SPY, IWR replaces IWM, VPL replaces EWJ with broader Asia-Pacific coverage, and emerging markets are dropped entirely. Eleven risk assets rather than twelve.
2. BOXX replaces SHY as the short-duration safe harbor.
3. A one-step breadth credit. One is added to the positive-score count before the crash-protection formula is applied, which keeps GPMv slightly more invested than the base model at any given level of market breadth.
4. A crash-protection scale matched to the eleven-asset universe, so the step size stays consistent with the smaller candidate set.
5. The IEF safety override described above, which the published model does not have.

Individually these are small. Together they open the strategy up modestly on the upside while adding a defense the original lacks on the downside.

How GPMv Trades

Signals are evaluated every month. Weights are not.

GPMv rebalances only when its target allocation has actually moved. If the model asks for the same three assets at the same risk level it asked for last month, the existing positions are left to drift rather than being trimmed and topped back to exact weights. Trading real money to undo a drift the model never objected to is a cost with nothing on the other side of it.

The effect here is deliberately modest. GPMv's top three churn frequently enough that its target genuinely changes in most months, so the drift allowance skips roughly one and a half rebalances a year out of a schedule that still trades in about three months out of every four. It was adopted to remove pointless trades, not to change returns.

This is worth being plain about: GPMv is an active strategy. It is not a low-turnover holding. The trade-off is deliberate, and it is what the risk profile below is bought with.

Performance Highlights

Backtested, 1980 through July 2026, frictionless:

	Annualized	Max Drawdown	Ulcer Index	UPI
GPMv	13.4%	-13.1%	3.4%	3.94

Characteristics worth noting across the 46-year record:

• Maximum drawdown just over 13%, against roughly 50% for U.S. equities over the same period.
• Equity-like compounding from a portfolio that is frequently not all equities. Gold, commodities, real estate and credit all take turns in the top three.
• Gradual de-risking. The defensive allocation scales in sixths with market breadth rather than switching on and off.
• Applying trading friction at 10 basis points one-way brings the annualized figure to roughly 12.5%. Given the turnover, that gap is real and is stated here rather than buried.

Portfolio Characteristics

Concentrated but genuinely diversified. Typically four positions: three risk assets chosen for low mutual correlation, plus the defensive sleeve. In fully defensive months, one.

Drawdown control is the design objective. GPMv is built to keep the Ulcer Index low and the maximum drawdown shallow. It is not built to win a bull market outright, and it will lag a rising equity index in strong years.

Active turnover. The top three change often. GPMv trades in roughly three quarters of all months.

Best suited to tax-deferred accounts. This follows directly from the turnover. The great majority of GPMv's gains are short-term, and the drift allowance reduces the number of trades without converting short-term gains into long-term ones. In a taxable account, plan accordingly.

Fully mechanical. Every allocation follows from published rules with no discretionary override.

No leverage. GPMv has no leveraged variant and applies no Smart Leverage overlay. All positions are held at 1X.

Scalable. All positions are liquid, exchange-traded funds.

Who It's For

GPMv is designed for investors who want:

• A conservative alternative to 60/40 that adapts rather than holding a fixed split.
• Shallow drawdowns as the primary objective, with competitive compounding as the result rather than the goal.
• Diversification enforced by the rules themselves, through a correlation penalty rather than a fixed allocation.
• A defensive position that recognizes when Treasuries are not the safe answer.
• A rules-based process with no market-timing judgment calls.

It is likely a poor fit for investors who want maximum growth and can tolerate deep drawdowns to get it, who prefer a strategy that trades a few times a year, who cannot hold commodities or gold, or who are investing in a taxable account and are highly sensitive to short-term capital gains.

---

Past performance, including backtested results, is not indicative of future results. Backtested performance is hypothetical, does not reflect actual trading, and benefits from hindsight in the selection of strategy rules. Portions of the long-term history rely on reconstructed proxy data for funds that did not exist for the full period. Investors should carefully consider their risk tolerance and consult with a financial advisor.

For the latest details, visit www.DualMomentumSystems.com