One of the biggest risks in retirement planning is also one of the most poorly handled: sequence of returns. Sequence of returns refers to the fact that the order of your returns matters enormously once you’re withdrawing, even if the long-run average is identical. A good decade followed by a bad one is survivable. A bad one at the start, while you’re selling into it, can follow you for the rest of your life.

Compare two scenarios that each start with a $2 million portfolio.

In Scenario 1, the portfolio increases 25% in each of years 1 and 2 then decreases 20% in each of years 3 and 4.

Scenario 2 is the same, except that the portfolio decreases 20% in each of years 1 and 2 and then increases 25% in each of years 3 and 4.

With no withdrawals, both portfolios end up back at $2 million. However, if the investor withdraws $100,000 at the end of each year, in Scenario 1, the investor’s net at the end of year 4 is $1,676,000, and in Scenario 2 the investor’s net at the end of year 4 is $1,493,750.

At the end of year 4, the investor in Scenario 1 has $182,250 more in assets, 12% more, than the investor in Scenario 2 simply because of how the different sequences of returns affected their portfolios.

(These scenarios are purely hypothetical and use round numbers to make the math easy to follow; they do not represent actual financial advice or expected market conditions.)

Planning for a possible poor sequence of returns is a major consideration in financial planning, and all competent advisors will consider it. However, approaches to mitigating the risk vary greatly, and the most common approach has unintended side effects that I believe are almost as damaging as the risk being addressed. Rain Dog uses a strategy that I believe is markedly better. This post explains the details.

The conventional fixes, and why I don’t use them

The investment industry tends to respond to sequence risk in a few familiar ways. The first is to increase the bond allocation.

The problem with increased bond allocation

The core issue with increasing the investor’s allocation to bonds is that this is nearly always done by increasing the allocation to bond funds (mutual funds or ETFs), and bond funds do not behave like individual bonds.

An individual bond has a maturity date and a par value. An investor can hold the bond to maturity and have a high likelihood (with the right kinds of bonds) of getting their money back at the maturity date regardless of how much interest rates fluctuated along the way.

Most bond funds do not have a comparable maturity date. Most bond funds maintain a maturity horizon on a rolling basis–the bonds’ maturity is always 1 year away, or 3 years away, or 7 years away, and so on. There is never a specific date with most bond funds when an investor can count on the fund “pulling to par” when it reaches maturity, like there is with individual bonds.

This is problematic because it means that investors who hold bond funds are not protected from market volatility. An investor who needed to withdraw funds in 2022, for example, would have found that bond funds were down from their peaks anywhere from 6% for short-term Treasuries (as represented by VFIRX), to 15% for intermediate Treasuries (VGIT), to 40% for long-term Treasuries (VGLT). Diversifying the bond portfolio wouldn’t make much difference–the total US bond market was down 17% (BND) and the total worldwide bond market was down 16% (BNDW).

In practice, reaching for safety by moving into bond funds amounts to moving into assets that perform more like low-return stock funds than individual bonds. Consequently, I do not believe an increased allocation to bonds (in bond funds) is an effective response to the risk of a poor sequence of returns.

Hold more cash

A second common response to mitigate a poor sequence of returns is to hold near-term spending in cash, such as money market or CDs. This kind of cash allocation will not drop in value the way that bond funds sometimes have, but allocating 3-5 years of planned spending to cash drags down portfolio performance significantly enough to create problems of its own.

On a typical 30-year planning horizon, allocating 5 years of spending to cash amounts to allocating 15-20% of the portfolio to the lowest-return asset class. While this effectively addresses the risk of poor sequence of returns, it also perpetually degrades the portfolio’s return to a degree that materially affects the investor’s safe spending level.

Overall, I believe this is a better response to addressing sequence of returns risk than increasing the allocation to bonds, but it leaves more money on the table than necessary.

Follow a safe withdrawal rate

A third common answer is to focus on a so-called safe withdrawal rate, which is usually a close variation on the 4% rule. Like allocating to cash, the 4% rule and its cousins are safe, but too crude to be useful. The 4% rule assumes an investor can withdraw at a constant inflation-adjusted amount over a 30-year retirement horizon, even while the investor’s life is constantly changing—spouses retire at different times (and thus the appropriate withdrawal rate changes), Social Security begins, a mortgage is paid off, RMDs begin, a vacation home is sold and its equity becomes available. Real life presents numerous moving targets, and the 4% rule accounts for none of them.

The institutional solution

There’s a well-defined institutional solution to this problem called the dedicated portfolio strategy. In this strategy, an investor buys individual bonds that mature close to the dates the investor plans to spend the money. If $100,000 is needed in 2027, you buy bonds that mature in 2027. If another $100,000 is needed in 2028, you buy bonds that mature in 2028. For institutional purposes, this is typically treated as a form of liability matching—matching future liabilities with assets that are highly likely to equal the specific liability values in the years the liabilities come due.

The liabilities could be matched with money market funds, but it is more cost-effective to hold bonds and benefit from their higher interest rates.

In this strategy, the right bonds must be held to fund the right years. When individual bonds are held to maturity, interest rates move and the bond prices move correspondingly along the way, but the interest-rate movements do not affect the terminal value of the bonds. When the bonds are held to maturity and the issuer pays as promised, the investor receives the expected par value of the bond.

If the bond is sold before maturity, the investor will be subject to mid-stream volatility in bond prices. The dedicated portfolio strategy depends on committing to hold the bonds to maturity.

While this protects the investor from interest-rate risk, the investor is still subject to credit risk—the risk of the bond’s issuer defaulting. This is typically addressed by investing only in investment grade bonds, which normally have default rates of less than 1%.

Overall, the dedicated portfolio strategy provides a way to fund future liabilities, protect bond investments from interest-rate risk, and safely earn higher rates of return than money markets or CDs.

Institutions have done this for decades. Individual households essentially couldn’t.

Why individual bond ladders do not work well for households

Creating a true dedicated portfolio built from individual bonds is harder than it sounds.

First, the math is complex. The planner is not just buying bonds. The planner needs to match the timing of planned spending to a series of maturity dates for the different bonds, account for coupon payments from each bond at each bond’s specific interest rate, and manage or reinvest the cash generated by the ongoing stream of coupon payments. The math is complex enough to be out of reach for most retail investors and even most financial advisors.

Second, the number of bonds needed for household spending is usually too small to be traded efficiently. Unlike stocks, which are now commonly traded in fractions, bonds are still typically traded in minimum lot sizes of 20 (approximately $20,000), and minimum lot sizes are often even higher. Institutions can buy bonds in lots of hundreds of thousands of dollars. Households’ smaller bond trades suffer from wider bid/ask spreads and worse execution. A bid/ask spread of 5% significantly erodes the nominally higher interest rates of bonds compared to money markets.

The third point is related—effective diversification requires more bonds than most households can justify. If one spending year is funded by only a handful of individual corporate bonds, the result is concentrated credit risk. A default or downgrade in just one bond still affects the household’s bond portfolio overall.

Any one of these factors is significant, and collectively, historically, they have put the dedicated portfolio strategy out of reach at the household level. A sound strategy for addressing sequence of returns has remained solely an institutional practice.

What changed

The emergence of target-maturity bond ETFs over the past decade has made the dedicated portfolio strategy accessible for individual households.

Unlike traditional bond funds, target-maturity bond ETFs are built around a specific maturity year. They hold bonds maturing nominally in the middle of the target year, and they distribute the net asset value in the fund’s target year.

That is not identical to owning individual bonds, but, for purposes of the dedicated portfolio strategy, it is very close. If an investor buys a target-maturity bond ETF and holds it to the maturity year, the investor is insulated from interest rate fluctuations. The ETF value will fluctuate along the way, but the investor can count on receiving the ETF equivalent of the bonds’ par values in the maturity year. In addition, the investor receives all the practical advantages of an ETF: diversification, liquidity, transparency, and far less trading inefficiency than trying to assemble a household-size bond portfolio.

The financial instruments that support practical implementation of a dedicated portfolio strategy finally exist, which makes this institutional strategy available at household scale.

Rain Dog’s implementation: the dedicated spending sleeve

The dedicated portfolio strategy is the general institutional practice described above. I call Rain Dog’s implementation of the approach a dedicated spending sleeve, which we use to pre-fund a client’s planned withdrawals for the next 2-4 years (nominally 4).

Here’s how our dedicated spending sleeve works.

How the sleeve is funded

The sleeve holds 4 years of planned spending in target-maturity funds (nominally).

The number of years it holds when we first implement a client’s sleeve depends on whether the core portion of the client’s portfolio has been up. We want to fund the sleeve only during up years, which is how we combat a poor sequence of returns. In a down market, we simply don’t fund the sleeve; we wait for a better year to start.

Assuming conditions are right to fund the sleeve, the specific funding amount is based on the client’s year-by-year spending needs, all things considered. We account for the client’s planned spending level, anticipated one-time expenses, onset of social security or pension, and all other factors that will affect spending.

If the client’s plan requires $100,000 to be withdrawn from the portfolio to support 2027’s planned spending, then roughly $100,000 of 2027-maturity funds go into the sleeve. We repeat the analysis for 2028, 2029, and so on.

How spending is taken from the sleeve

Once the sleeve is in use, spending each year comes out of the sleeve rather than out of the client’s core portfolio. (I’m using core portfolio to refer to everything in the client’s investment portfolio that isn’t part of the spending sleeve.)

How the sleeve is extended

The sleeve is extended after years in which the client’s core portfolio returns at the 60th percentile or better, measured against the historical return distribution for that client’s asset allocation. That threshold matters:

• Refilling at lower percentiles would mean refilling after mediocre or down years. The result would be funding the sleeve by selling core assets after down years, which is exactly what the sleeve is trying to prevent.
• Refilling only after exceptional years increases the likelihood of depleting the sleeve and being forced to spend out of the core portfolio after down years, which is another example of doing what the sleeve is intended to prevent.

What happens if the sleeve runs out

In a series of bad years, the sleeve can draw down to 3 years, 2 years, 1 year, or possibly zero years. In the event that the sleeve runs down to zero, that is not necessarily a bad outcome. That only happens after 4 years of spending have already come out of the sleeve, so the client was protected through the first 4 years of bad returns.

Second, the goal is not perfection; the goal is to avoid spending large amounts out of the core portfolio while the market is significantly down. Preventing 4 years of spending in a long down cycle has significantly accomplished that–it has significantly mitigated the risk of poor sequence of returns.

Why not fund longer sleeves

We find that clients get comfort from knowing that their planned spending is covered for the next 2-4 years. If that’s the case, why not extend the sleeve out to 5, 6, or 7 years?

The practical answer is, because it isn’t needed and a good idea taken too far starts to create problems of its own.

Mathematically, a 4-year sleeve that is refilled after 60th percentile years can be expected to run out in approximately 2.5% of years. Over a 30-year planning horizon, that implies about a 50/50 chance that a client will ever deplete a 4-year spending sleeve. We believe that is comfortable for most clients.

Sometimes it makes sense to extend a sleeve to 5 years. That length sleeve will run out in approximately 1% of years, with about a 1 in 4 chance of depleting the sleeve over a 30-year retirement horizon.

Longer sleeves decrease the chances of ever running out, but the tradeoff is that each year added to the sleeve removes a year from the core portion of the client’s portfolio. That lowers expected returns, and that can create its own failure path. Being too cautious can become a self-fulfilling prophecy.

Rain Dog’s portfolios are designed to reduce the need for longer sleeves

Our portfolio optimization process concentrates on recovery time as a key risk metric in our portfolio construction. We fund down cycles out of stable assets via the dedicated spending sleeve, and we optimize the portfolio as a whole to shorten the down cycles, which increases the chance that longer sleeves won’t be needed.

With this approach, our analysis has found that 2-4 years is the sweet spot for mitigating risk of sequence of returns without dragging down overall portfolio returns.

Implementation details matter

The strategy is simple to state and less simple to run well. Here are some of the fine points.

Source of sleeve funding. The sleeve is funded from the bond allocation, not added on top of it. Otherwise, the client ends up with too little in growth assets overall.

Sleeve location. Target-maturity bond ETFs generate interest, which is taxed as ordinary income. We locate the sleeve in an IRA or other tax-deferred account when possible.

Withdrawal of funds from the sleeve. If the sleeve is held in an IRA, we don’t necessarily withdraw money directly from the IRA to support client spending. When circumstances permit, we can take spending from taxable assets while making offsetting adjustments inside the IRA for a more tax-efficient withdrawal.

Effect of sleeve’s interest on planning in later years. If the client expects to need $100,000 from the portfolio in 3 years, we fund the level that is expected to grow into that spending amount in 3 years. We expect inflation to offset growth somewhat, but the sleeve earns a real return, so we can fund somewhat less than the nominal figure even after accounting for inflation.

Types of target-maturity bond ETFs that are in the sleeve. The sleeve contains mostly ETFs that invest in investment grade bonds for reliability, with some ETFs that invest in high-yield bonds for increased return. The specific mixes are mathematically determined and account for interest rate and credit risk tradeoffs over each time horizon.

Client variations. A more risk-sensitive client can carry a longer sleeve, a heavier allocation to ETFs that hold investment grade bonds, or both. A more risk-tolerant client can carry a shorter sleeve, a heavier allocation to ETFs that hold high-yield bonds, or both.

The sleeve’s payoff

In the example at the beginning of this article, the investor in the bad-years-first scenario ended up with less money because the sequence of returns was poor. That is the primary scenario the sleeve is designed to avoid.

At current interest rates, in the bad-years-first scenario, the sleeve strategy would leave that investor with about 7% more at the end of four years. In the good-years-first scenario, the sleeve would leave the investor with about 3% more. In both cases, the benefit derives from not withdrawing money from the core portion of the portfolio during years the portfolio is down.

That does not mean the sleeve wins in every possible scenario. If the investor’s portfolio rises consistently, year after year, the sleeve will degrade performance to a degree because part of the portfolio was invested in lower-return sleeve assets instead of the core portfolio. For many investors that tradeoff is desirable, but that varies based on individual circumstances.

The biggest benefit: ability to support clients as human beings

The explanation so far has concentrated on math. I like math, and the math works, but here’s why I really care about this strategy.

I think my job is to make as much money for my clients as possible, while still allowing them to get a good night’s sleep every night.

A client who knows the next several years of spending are already funded is less likely to panic during a market decline, and possibly less likely to feel even a little bit anxious.

When the market goes down and the client asks, “Should I sell?” the spending sleeve makes it easy to say, “Our plan accounts for this. This is the reason we put the spending sleeve in place. This is when the spending sleeve really earns its keep. You won’t need to sell anything for at least 2-4 years.” They can look at their sleeve and verify that they have literally years to wait for the market to recover.

The spending sleeve also buys us time. With a funded sleeve, a client and I have literally years to talk through a bad market before we have to make any hard decisions. Nothing has to be done today, or this quarter, or this year. Urgency stops being the enemy and time to slow down becomes our friend—we get to sit on the sidelines for a long time and let the market do whatever it’s going to do.

In some cases, the spending sleeve lets us go a step further. Because the spending sleeve reduces fear, some clients are able to hold more growth assets than they otherwise would. The somewhat-conservative sleeve provides comfort that allows for more risk-taking and more growth in other areas, and their returns can increase over time.

The spending sleeve doesn’t require a client to be emotionally bulletproof. It gives the client a structure that makes it easier to commit to and follow their best investment plan.

That is what a retirement income strategy should do.

---

Note: The bond-fund drawdown figures in this post are drawn from Portfolio Visualizer. The sleeve depletion statistics were calculated by Rain Dog based on returns data from YCharts. The sleeve interest rate was assumed to be 4.96%, which is the blended average SEC yields of ETFs for a 4-year sleeve as of 6/29/2026.