Decumulation and Silver Divorce: “Spike Expense Risk” and 3 other risks


By Peter Neuwirth, FSA, FCA

In 2020, I wrote a series of short essays expanding on each of the 6 principles of Holistic Financial Wellness that I outlined in my book, “Money Mountaineering.” These essays were intended to help readers understand how each principle might operate in practice as well as give me a chance to update what I had written for how the world and my thinking about these principles had changed in the months between the submission of my manuscript to the publisher and the day that the book became available to the general public. Click here to read those essays.

In essay #6, Optimism Bias, I shared my perspective on the financial consequences of my own divorce, a contingency that  I had carefully considered before I got married, but because of “optimism bias,” had inadequately prepared myself for. I want to go beyond what I shared last year and discuss how divorce is an example of what I consider “Spike Expense Risk” – a risk that has been largely ignored by financial analysts and researchers but can have grave consequences for all retirees.

While marriage can fall apart at any time in one’s adult life, experiencing a divorce when one is older and has retired from full-time employment can be particularly disruptive. Unfortunately, “Silver Divorce” has become increasingly common and a subject of much study and discussion. For those interested, click here to read about the perspective my collaborators and I are providing to that discussion.   

Today, however, I want to talk more about the actuarial aspects of  “spike expense risk”, and how important it is to consider that risk along with the other unique decumulation risks  when  planning for retirement or to advise clients on their retirement income strategy

In the months since “Money Mountaineering” was published in October 2021, I have spent much of the last year speaking virtually and in the “real world”  to thousands of people about my ideas and how others might use my insights to help individuals navigate through the financial wilderness. It has been a wonderful experience, far more enjoyable than I could have expected, but I am now back home. I have been spending much of my time researching and writing about the unique financial risks that are present after an individual stops working full time, where they accumulate assets and begins drawing down those assets and enters the “decumulation” phase of life.

Planning for Retirement: Accumulation vs. Decumulation

With tens of millions of Americans now retired and tens of millions more planning/hoping to retire in the next decade, it has become critical for investment advisors and financial planners to communicate clearly about the significant differences between planning strategies that are most effective while an individual is living on their employment income and accumulating assets to be used in retirement and the decumulation phase of life when an individual needs to begin drawing down those assets to sustain themselves for the rest of their life.

While both aspects of retirement planning focus on ensuring sufficient income during retirement, the decumulation problem is distinctly different from the challenge of accumulating assets during one’s working years

The most obvious difference is that during the accumulation phase, the individual’s cash flow is positive (e.g., 401(k) deferrals, company contributions, and personal savings). During decumulation, cash flow is negative (e.g., withdrawals from 401(k), savings accounts, and other personal assets like Home Equity Cash Value Life Insurance policies, etc.).

Less obvious is that the time horizons differ in that the end point of accumulation (i.e., retirement date) is  (usually) within the individual’s control. The endpoint of decumulation is determined by life contingencies (e.g., death) and is governed by random variables that are not easily calculable on an individual basis.

Financial Risks Unique to Decumulation

There are 6 distinct financial risks that a retiree faces as they try to sustain themselves for the remainder of their lives. Because of the cashflow and time, horizon differences just noted,  only the first two of those 6  are present during the accumulation phase.

Specifically, there is a large body of investment theory addressing inflation/investment risk during the accumulation phase beginning with  Harry Markowitz’s Modern Portfolio Theory and the many refinements added in the last few decades.

There is also a significant body of literature on investment/asset allocation strategies that protect against investment liquidity risk while accumulating assets for retirement. Investment liquidity risk is a second distinct risk that must be considered along with inflation/investment risk during both accumulation and decumulation and reflects the risk that changing the asset allocation of an investment portfolio could require selling an investment in a down market.

Once an individual begins decumulating assets (i.e. retires and begins to draw down assets to meet living expenses), 4 more risks that were either non-existent or immaterial while the individual was accumulating assets begin to rear their ugly heads.

The four separate risks that are highly significant just during decumulation are:

  1. Sequence of Returns Risk
  2. Longevity Risk
  3. Tax Risk
  4. Spike Expense Risk

What follows is a brief description of each and some commentary on the approaches that might be most effective for mitigating each. It is important to note, however, that decumulation remains a very hard problem – one with no “closed form” solution and one where research so far  (focusing on just decumulation)  has been very limited and, at least for the foreseeable future, it seems unlikely that a comprehensive general strategy can be developed to address all 4 risks for every (or even most) people who are developing financial plans for their retirement years.

Sequence of Returns Risk

Sequence of Returns Risk is a risk well studied by actuaries and others, but until recently, it was not recognized as being vitally important in decumulation while being largely immaterial (from a planning perspective) while accumulating assets

Sequence of returns risk applies, by definition, to retirees whose primary source of income is a securities portfolio — typically, but not necessarily, a 401(k) account or a rollover IRA.

It is the risk that arises from the confluence of 2 factors: 1) The volatility of the securities portfolio from which the retiree is distributing income, and 2) The need to sell some of the securities to meet living expenses when the portfolio is at a low point in its volatility cycle.

Selling low is, of course, the route to premature portfolio exhaustion.

The good news is that research has shown that this risk can be addressed for many retirees with a drawdown strategy that utilizes a  “coordinated strategy,” which incorporates an asset whose return is uncorrelated with the invested portfolio. An example of such assets is a Home Equity Conversion Mortgage (“HECM”) that can be drawn on rather than the portfolio when the portfolio is down

For those interested in the mechanics of how this works, the methodology, and its effect on mitigating this risk, click here.

Longevity Risk

Longevity risk is theoretically present during the accumulation phase, but, as with sequence of returns risk, the risk of outliving one’s assets is much more significant during the decumulation phase. Unlike most other retirement risks, longevity risk grows dramatically more acute over time as individuals age and stay healthy. Throughout the decumulation phase, longevity risk becomes exponentially more important – especially as the retiree attains an age where expected mortality increases significantly.

This dynamic aspect of longevity risk is rarely addressed, even in the actuarial literature, and most mitigation approaches (e.g., annuitization) tend to be financially inefficient. As a result, a more targeted approach to addressing this risk is needed. Fortunately, some approaches show promise for cost-effectively mitigating longevity risk.

For example, the purchase of a QLAC (pure longevity insurance) is one possibility, but the question of when, during decumulation, to purchase a QLAC is tricky. Too early in retirement and the “premium” for such insurance may be too high (from a purely actuarial perspective), but waiting too long may make  the purchase  prohibitively expensive relative to the resources that a retiree may have available

Maintaining a buffer asset appears to be a potentially more practical  alternative, but as with most life contingencies, direct insurance of the risk appears to be a more efficient means of mitigating the risk

For buffer assets, the challenge is to preserve the asset’s value while maintaining a desired standard of living, while the challenge of using annuity products is cost and timing

Tax Risk

While tax is almost always a very important consideration for an individual as they accumulate assets in anticipation of retirement, it is far beyond the scope of this essay to discuss the results of that research or the extraordinary complexity and variety of effective tax strategies that have been and continue to be developed by tax experts spanning an individual’s complete lifespan and beyond.

When considering that range of strategies and how they might be rendered more or less effective as the tax laws change, it is important to reflect again on one of the fundamental differences between an individual’s situation during the decumulation phase versus their situation while they are accumulating assets and have yet to begin distributions from each of the assets that they will own when they stop working.

Because, by definition, those in retirement will be receiving cash from either debt they take on (e.g., a Reverse Mortgage), assets that they have already accumulated (e.g., 401(k), IRA, etc.), or from a 3rd party by contract (e.g., annuity, social security, employer-provided pension, etc.) there will be substantially more variability and less control that an individual has concerning their annual tax expense after retirement than before.

Therefore while tax risk is very much present during the accumulation phase of life, once an individual stops working and begins decumulating assets, this risk becomes both more significant and less manageable, and as a result, I believe that it is useful for retirees to consider not just tax planning but also “tax expense risk” mitigation.”  Annual Tax Expense has not, to my knowledge, ever been considered a unique risk during decumulation to be addressed from an actuarial perspective in the same way that longevity and sequence of returns risk have been. Yet, it seems that the actuarial perspective may be useful here as well – considering a retiree’s annual tax expense as a random variable subject to many uncertain and hard-to-predict factors whose likelihood is hard, if not impossible, to estimate (e.g., the possibility that future marginal tax rates will change via future law changes). During the accumulation phase, long-term tax expense will only have a major impact on the level and mix of the accumulated assets an individual will retire with, while in the decumulation phase, annual tax expense will directly impact both the retiree’s lifestyle and the risk of a retiree outliving their resources.

There don’t appear to be any “pure insurance” solutions to this problem. Still, one area of some potential is converting a Traditional/Rollover IRA into a Roth IRA and considering the upfront cost of conversion as an insurance premium against the future volatility of the tax expense associated with distributions from a Traditional/Rollover IRA. Initial research results are promising, but much more work needs to be done before anything definitive can be said.

Spike Expense Risk

Finally, we come to perhaps the most important understudied decumulation risk – the risk that a well-designed and well-executed retirement income strategy can be completely derailed due to unexpected “spikes” in necessary living expenses while drawing down one’s assets. Mathematically, this risk is equivalent to amplifying sequence of returns risk with two other unfortunate aspects – first, that the incidence, frequency, and magnitude of the spikes are subject to variability that is very hard to model, and second that the aggregate magnitude of the spikes an individual experiences throughout retirement may be significant enough to warrant long term adjustments to one’s ongoing lifestyle expectations and/or basic draw down strategy (e.g., how much of an annual draw to take each year).

Note that each spike risk might need to be addressed separately. Divorce is a spike risk that can be addressed via techniques referred to earlier when “Silver Divorce” was discussed. Specific insurance products or riders on existing policies might be purchased for others (e.g., the potential for Long Term Care). Still, many possible spikes (a family crisis or the impact of a natural disaster) are not insurable or manageable.

The real challenge for Financial Planners is to enumerate for each individual the range of possible spikes that might affect them and ensure adequate attention has been paid to each. Since it is impossible to enumerate all possibilities, our current research into possible risk mitigation techniques includes reducing annual draws and/or setting aside a “buffer asset” to absorb the impact of possible spikes partially, but as with Tax expense risk, much work remains to be done

Concluding Thoughts

The decumulation problem is one of the most important retirement planning challenges facing individuals and those organizations involved in assisting people with their financial plans.  In some sense, it is still an open research problem for which only partial/incomplete solutions have been developed.

In my view, the fundamental nature of the decumulation problem is different than that of the accumulating assets for retirement. While the extensive body of investment/asset allocation theory developed over the last 50 years is often applicable, decumulation is more of a  risk management problem where actuarial science and the techniques developed by actuaries should be brought to bear.

At its core, decumulation is an asset/liability and cashflow matching problem governed by random variables from unruly and difficult-to-determine probability distributions. It will take the efforts of both researchers in the investment and actuarial professions to make progress.

This challenge and the scope of potential research on decumulation in this area are significant. The benefits of making progress are vitally important to millions of current and future retirees.