Protective Asset Allocation (PAA) Overview

Protective Asset Allocation (PAA) is the 2016 strategy that started everything else Wouter Keller and Jan Willem Keuning went on to publish. It was the first to use breadth momentum: instead of asking whether a particular holding is failing, count how many assets across a whole universe are failing, and let that count decide how much of the portfolio comes out of the market.

Each month PAA scores twelve global assets against their own 12-month average price. The number with negative scores sets a floor on the bond allocation. What is left over is spread across the six strongest assets, and any of those six that is not itself rising gives its share back to the bond side too.

Dual Momentum Systems runs the high-protection configuration, which reaches a fully defensive portfolio when half the universe has turned down.

The Backstory

A strategy designed to a promise

Most tactical strategies are built to maximize something: return, return per unit of drawdown, some blend of the two. PAA was built to keep a promise, and the promise came first.

The paper appeared in April 2016, after the 2008 crisis and a sharp pullback at the end of 2015 had left a lot of people wanting something other than equity risk, at a time when savings accounts and term deposits paid almost nothing. The authors set themselves an absolute-return requirement: across rolling one-year periods, at least 95% should come in above zero, and 99% above minus five percent.

That framing explains everything about the design. The subtitle of the paper calls PAA an alternative to a one-year term deposit, and the audience it had in mind was someone choosing between a certificate of deposit paying nearly nothing and a portfolio that might actually lose money. For that reader, a strategy that occasionally returns 25% is worth far less than one that almost never returns less than zero.

Breadth momentum

The mechanism they invented to keep that promise is the one every later model in the family inherited.

A traditional dual momentum strategy tests each holding: rank the universe, hold the best few, and replace any pick that has turned negative with cash. The problem is timing. If you hold the top six of twelve, your own six picks only start failing well after the broader market has begun breaking down.

PAA instead counts failures across the entire twelve-asset universe, whether or not it holds them. Four assets down out of twelve is information about market health that a strategy watching only its own six positions would not see. That count sets the minimum bond allocation, which rises smoothly as conditions deteriorate rather than flipping at a threshold.

The name follows from the behavior: protection that scales with the danger instead of arriving all at once.

The protection factor

PAA ships with a dial. The protection factor sets how quickly the bond allocation climbs as assets fail, and the authors described three settings: low, medium and high.

At the high setting, which is the one implemented here and the one their absolute-return target pointed to, the portfolio is fully defensive once half the universe has non-positive momentum. It does not wait for all twelve to break down; six is enough. That is a cautious dial setting, and it is the reason PAA's returns sit below its successors while its declines are shallower.

Two safety assets, not one

The paper's original safety allocation was a single bond fund. The authors' later recommendation, which this implementation follows, is to hold whichever of two Treasury funds has the stronger momentum: short-term or intermediate-term.

The reason is what happens to bonds in different kinds of trouble. In a deflationary scare, intermediate Treasuries rally hard and provide genuine crisis alpha. In a rising-rate year, they fall, and the short end is the better place to wait. Choosing between the two each month instead of committing to one is a small refinement that matters more in some decades than others. The paper also explicitly discourages long Treasuries as the safety asset, on the grounds that the volatility works against the absolute-return goal.

What came after

PAA is the root of the family tree. Generalized Protective Momentum kept its breadth machinery and changed the scoring to account for how much assets overlap. Vigilant Asset Allocation sharpened breadth into something far more decisive. Defensive Asset Allocation moved the counting onto a separate set of assets, which became the canary concept. Bold and Hybrid Asset Allocation carried that forward.

Every one of those models is, at bottom, an argument about how to do better what PAA did first.

Core Strategy Logic

PAA is evaluated once a month. Signals come from month-end data and set the holdings for the following month.

Every asset is scored the same way: its current price against the average of the last thirteen month-end prices. Above that average is a positive score, below it is negative. This is a slow measure by design, since the whole point is to respond to sustained deterioration rather than to one bad month.

Step 1 - Count the failures, set the bond floor

Count how many of the twelve risky assets have a score at or below zero. That count sets the minimum share of the portfolio going to the safety side. At the high-protection setting used here, six failures out of twelve puts the entire portfolio on the safety side; zero failures puts none of it there; and the fraction moves smoothly in between.

Step 2 - Rank and fill the risk slots

Rank all twelve by score and take the top six. Each of those six is entitled to an equal share of whatever is left after the bond floor.

Step 3 - Make each pick earn its slot

A pick only receives its share if its own score is positive. Being in the top six is not enough; a slot held by a falling asset gets nothing, and its share moves to the safety side instead.

This is why the bond fraction from Step 1 is a floor rather than a final answer. When the strongest assets are themselves falling, the actual defensive allocation ends up higher than breadth alone would suggest.

Step 4 - Choose the safety asset

The entire defensive allocation goes into whichever of the two Treasury funds has the higher score, regardless of whether that score is positive.

The portfolio is rebalanced to fresh targets every month.

The universes

Risky, twelve candidates, top six eligible:

Category	Assets
U.S. equities	large cap (VOO), Nasdaq 100 (QQQ), small cap (IWM)
International equities	Europe (VGK), Japan (EWJ), emerging markets (EEM)
Real assets	real estate (VNQ), commodities (PDBC), gold (SGOL)
Credit and duration	high yield (HYG), corporate bonds (LQD), long Treasuries (TLT)

Safety, two candidates, best one held:

Asset	Role
SHY	Short Treasuries
IEF	Intermediate Treasuries

Performance Highlights

Over the full published history, alongside its direct descendant and the benchmark the paper set out to replace:

February 1980 through August 2026, net of trading friction:

	CAGR	Max Drawdown	UPI
Protective Asset Allocation	+9.8%	-13.9%	1.56
Generalized Protective Momentum	+9.9%	-11.7%	1.67
60/40	+9.8%	-32.3%	0.90
Triad	+14.6%	-13.8%	3.85
S&P 500	+11.9%	-51.0%	0.61

The third column is the Ulcer Performance Index: return above cash divided by how deep the strategy went and how long it stayed there. Max Drawdown reports the worst single moment; UPI reports the experience of holding it. Higher is better.

The 60/40 row is the one the paper would want you to look at, since PAA was pitched as a tactical version of exactly that portfolio. Over more than four decades PAA matches 60/40's return while taking less than half its worst decline, and its UPI is nearly double. Judged against what it set out to replace, it did the job.

Judged as a growth strategy it looks unremarkable, trailing the index by a couple of points a year. That is the correct reading and not a criticism: a strategy built to keep rolling one-year returns above zero is not trying to win that comparison.

Generalized Protective Momentum, which took PAA's breadth machinery and added correlation-aware scoring, lands slightly ahead on both drawdown and UPI at a nearly identical return. A modest improvement from a modest change, which is honest.

Triad is included as the closest comparison among strategies built here.

The same five, measured from January 2000:

January 2000 through August 2026, net of trading friction:

	CAGR	Max Drawdown	UPI
Protective Asset Allocation	+7.9%	-11.0%	1.90
Generalized Protective Momentum	+7.2%	-11.7%	1.45
60/40	+6.6%	-32.3%	0.62
Triad	+14.2%	-8.6%	4.93
S&P 500	+8.0%	-51.0%	0.39

PAA is the rare strategy on this site whose recent record reads better than its full one. Its worst decline is smaller here than over the whole period, meaning its deepest drawdown came before 2000, and its UPI improves rather than slips. It also moves ahead of Generalized Protective Momentum in this window, reversing the full-history ordering.

Two cautions apply as always. A window starting in January 2000 begins shortly before a major bear market, which flatters strategies that step aside during one. And a single figure covering decades says nothing about the order the returns arrived in, which is most of what an investor lives through.

Portfolio Characteristics

• The original breadth design. The first strategy to set its defensive allocation by counting failures across a universe.
• Gradual, not binary. The bond fraction climbs smoothly as conditions deteriorate. There is no single switch.
• Cautious dial setting. The high-protection configuration reaches fully defensive at half the universe, not all of it.
• A floor, not a ceiling. Breadth sets the minimum defensive allocation; failing picks push the actual figure higher.
• Every pick must be rising. A top-six ranking alone never buys anything.
• Slow momentum measure. Price against a 12-month average, which reacts to sustained trends rather than to single months.
• Two-speed safety. Short or intermediate Treasuries, whichever is stronger, with long Treasuries deliberately excluded.
• Built for consistency, not growth. The design target was rolling one-year returns above zero, and the returns reflect that priority.
• Monthly rebalancing. Fresh targets every month.
• No leverage. PAA never holds a leveraged fund.
• Best suited to tax-deferred accounts. Monthly rebalancing generates short-term gains. In a taxable account, plan accordingly.
• Fully mechanical. Every allocation follows from published rules with no discretionary override.

Who It's For

PAA is designed for investors who want:

• Consistency above growth, with the shallowest practical drawdowns.
• A tactical alternative to a conservative balanced portfolio rather than a way to beat the index.
• Risk that comes off gradually as conditions worsen, in small increments.
• A defensive position that adjusts its duration to the environment.
• A rules-based process with no market-timing judgment calls.

It is a poor fit for investors seeking maximum long-run growth, who would be frustrated trailing equities through a long bull market, or who trade in a taxable account. Investors who like this structure should compare it with Generalized Protective Momentum, which refines the scoring, and with Hybrid Asset Allocation, the authors' most recent and best-performing model. Those wanting shallow drawdowns from strategies built here should look at Triad and Quiet Compounding.

---

Protective Asset Allocation was created by Wouter Keller and Jan Willem Keuning and is presented here as an independent implementation, in the high-protection configuration with standard fund substitutions. Dual Momentum Systems is not affiliated with or endorsed by either author. The original paper is available at SSRN.

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