Stacking Trend Following on a Glidepath

A typical investment glidepath is built around the idea that an investor’s balance sheet changes through time. Early in life, most of an investor’s wealth is human capital: decades of future earnings that behave like a long-duration bond. They pay out steadily and are largely insulated from the stock market.

With that bond-like asset already in place, the financial portfolio can hold mostly stocks, because the investor needs growth and future earnings can offset the impact of any equity market shocks. As retirement approaches, the human capital is spent down and the earnings stream comes to an end. The portfolio then has to supply the stability that a paycheck once provided, so it swaps stocks for bonds.

Diversifying return streams - such as managed futures trend following, gold, merger arbitrage, and futures yield (carry) - are usually excluded from traditional glide path models. This may be a missed opportunity for both savers and spenders! Because they have historically moved independently of stocks and bonds, diversifiers have the potential to lift the portfolio’s compound growth rate during accumulation periods, and cushion drawdowns during withdrawals (when losses are hardest to recover from).

The catch is that an allocation to them has to come from somewhere. A traditional portfolio can hold only $1.00 of assets for every $1.00 invested, so funding diversifiers means selling stocks and potentially giving up growth, or selling bonds and giving up stability.

Funding diversifiers means selling stocks and potentially giving up growth, or selling bonds and giving up stability.

Return stacking (also known as Portable Alpha) offers a solution. By applying a modest amount of leverage, an investor can layer diversifying return streams on top of the portfolio they already own. The result is more than $1.00 of exposure for each $1.00 invested.

The glidepath can stay exactly as it is while the returns of trend following are “stacked” on top.¹

Keep the Path, Add the Stack

We started with a representative industry glidepath modeled on Vanguard’s target-date fund glide path: 90/10 stocks and bonds through age 40, de-risking to 50/50 at 65, and landing at 30/70 at 72, interpolated between those published knots. A simulated investor saves along this path over a 40-year career, retires at 65, and funds a 4% real withdrawal plan through age 95².

To find out whether stacking improves this path, we ran the base glidepath and a stacked version of it side by side through 2,000 simulated lifetimes, each calibrated to conservative forward-looking return assumptions.³ We limited the stack to a single diversifier, managed futures trend following.

The stacked version follows three rules:

For each simulated lifetime, an optimizer determines the stack size at every age. It chooses the total portfolio combination with the highest return that still keeps volatility within the limit. Any one simulated history reflects a particular sequence of market outcomes, so we do not rely on it alone. Instead, the recommended allocation at each age is the average of the optimal allocations across all 2,000 runs. That average reflects the full range of market paths the investor might experience rather than a single lucky or unlucky one.⁴

Figure 2 shows the optimization results. Through the working years the optimizer deploys nearly the full 30% budget in trend following, then releases it gradually until it lands near 15% at age 72.

As the core de-risks toward 30/70, its volatility falls, and the stack must de-lever with it to stay inside the risk budget.

Accumulation: Arriving at Retirement Wealthier

By running the stacked and base glidepaths through identical market simulations, we can compare their outcomes world by world and see how each would have fared under the same conditions. In each world, we record the ratio of the stacked investor’s wealth to their base twin’s at every age. Figure 3 plots different percentiles of that ratio across all simulations, through the working years.

A value of 1.10 means the stacked investor holds 10% more wealth than the base investor at that age. At each age, the median line marks the middle outcome; the 5th and 95th percentile lines bracket the unusually bad and unusually good ones.

The whole fan drifts upward with age. The median world reaches 22% more wealth by 65, and half of all worlds finish between 13% and 31% ahead.

The median world reaches 22% more wealth by 65, and half of all worlds finish between 13% and 31% ahead.

Given financial planning is more concerned with bad luck than good, it’s worth explaining the 5th percentile outcomes in more depth. It drops below 1.0 in the first year and stays there until roughly age 60. In these worlds trend following earns less than its financing cost, so the stack subtracts from wealth instead of adding to it. Fortunately, in these cases, the adverse effect is fairly small.

At its worst, near age 40, the stacked investor still has 95% of the base investor’s wealth. The gap then narrows through the 50s, and even the 5th percentile finishes 1% ahead by age 65.

Table 1 summarizes the arrival at retirement.

The advantage realized by the stacked investor did not depend on friendly markets. In the worst decile of late-career equity markets (the lowest 10% of real stock returns between ages 55 and 65), the median improvement grew to 27% and the stack finished ahead in more than 99% of worlds.

Drawdowns: Depth and Duration

The natural concern with adding leverage to a portfolio is that it falls further when markets do. To explore this concern, in Figure 4 we plot the percent of simulations that each approach exceeded a given drawdown level. Note that we measure drawdowns on investment return alone, before any contributions or withdrawals, to compare the approaches like-for-like.

At every depth, the stacked glidepath was less likely to have ever fallen that far. The median worst decline was -48%, versus -52% for the base glidepath, and the odds of a real drawdown worse than 60% fell from roughly 28% to 14%.

Depth is only part of a drawdown’s cost. A loss that recovers in three years does far less damage to a retirement plan than one that takes fifteen, especially during withdrawals (and to say nothing of the fact that long stretches underwater are when investors tend to abandon a strategy).

A loss that recovers in three years does far less damage to a retirement plan than one that takes fifteen.

The larger difference between the two glidepaths shows up in recovery time. Figure 5 shows, on the same basis, how quickly each strategy recovered from its deepest drawdown, and how often each sat in a deep drawdown at every age.

Regardless of approach, recoveries measured in real terms are slow because inflation raises the high-water mark every year. The gap between the two strategies is still wide.

The median deepest drawdown took 17.5 years to fully recover for the base glidepath and 11.7 years with the stack. The stack recovered faster in 83% of paired worlds, and the share of worlds whose deepest drawdown never recovered by age 95 fell from 16% to 6%. Past age 70, the base glidepath was in a deep drawdown in roughly 45% of worlds at any given time; the stacked glidepath, in 30%. The drawdowns were about as deep, but they ended sooner.

Decumulation: Making the Withdrawals Last

Wealth at retirement only matters if it lasts. A lifecycle portfolio has to fund withdrawals through retirement, and here, the stack’s advantages continued. Across all simulations, the stacked glidepath sustained the 4% plan to age 95 in 79% of worlds, versus 64% for the base glidepath, and median wealth remaining at age 95 nearly tripled.

The stacked glidepath sustained the 4% plan to age 95 in 79% of worlds, versus 64% for the base glidepath, and median wealth remaining at age 95 nearly tripled.

Figure 6 asks whether the money lasts. The left panel tracks the share of worlds still funded at each age under the 4% plan. The gap opens through the late 80s and reaches 15 percentage points by 95. The right panel repeats the test inside the worst 10% of early-retirement markets, ranked by real stock returns over the first five years after retiring. There, the base glidepath sustained the plan just 12.5% of the time; the stacked glidepath, 34.5%.

Neither number in the worst decile is comfortable, and stacking is no cure for a hostile decade. Even so, the stacked glidepath carried trend following, a return stream that does not depend on equities. It kept working through stretches where stocks and bonds struggled together, which slowed the erosion of capital in the years when withdrawals did the most damage.

A retiree willing to accept the base glidepath’s odds under the 4% rule could instead have spent roughly 0.3 percentage points more each year with the stack, about 8% more real income in every year of retirement.

Figure 7 translates the same question into spending terms. The left panel shows, for every constant real withdrawal rate, the percentage of simulations where wealth lasted to 95. A retiree willing to accept the base glidepath’s odds under the 4% rule could instead have spent roughly 0.3 percentage points more each year with the stack, about 8% more real income in every year of retirement. The right panel shows the median sustainable rate, which the stack lifted from 4.3% to 4.6% across all worlds and from 3.3% to 3.7% in the worst early-retirement markets.

No Silver Bullets, But Maybe Stacked Odds

A bad market spares no portfolio. The investor who retires into a lost decade will feel it, stacked or not. What stacking changed in our simulations was the sensitivity a plan’s outcome had to equity market growth. Across 2,000 lifetimes, stacking trend following increased the wealth an investor had at retirement, reduced both the depth and duration of drawdowns, and helped that wealth last longer once withdrawals began.

We cannot solve a bad market, but we can try to stack the odds.

Key Takeaways

• 1 Trend following improved both sides of retirement. In our simulations, the median world retired with 22% more wealth, and the odds of funding a 4% withdrawal plan to age 95 rose from 64% to 79%.
• 2 The glidepath and its risk never changed. The stock/bond core was untouched, and a volatility budget kept total risk within 5% of the base glidepath’s; the optimizer even tapered the stack toward retirement on its own. Drawdowns were no deeper and recovered roughly six years sooner in the median world.
• 3 The edge was largest in the worst markets. In the worst decile of late-career equity markets, the median wealth improvement grew to 27%, and in the worst early-retirement markets, the odds of sustaining the 4% plan rose from 12.5% to 34.5%.