‘Success,’ More or Less?

 What does it mean to have a 75% probability of success?

I’ve never been particularly fond of the probability-of-success measures commonly used in retirement planning. It’s not that the calculations aren’t useful — or that I have a better crystal ball. I’m just not convinced that most people understand what the resulting percentage means, much less how to apply it to their retirement decisions.

After all, a 75% probability of success sounds like a grade — and not a particularly good one. It also sounds as though there is a 25% chance that your retirement will be a complete and unmitigated failure. Neither interpretation is necessarily accurate.

ad space

A new paper from David Blanchett at PGIM, aptly titled “Successfully Failing,” takes on that conventional measure and suggests that it may not simply be confusing. It may actually lead retirees — and those advising them — to make less-than-optimal decisions.

Success ‘Measures’

Probability-of-success calculations generally run a retirement strategy through hundreds or thousands of different scenarios. If the retiree’s assets last through the prescribed retirement period, the scenario is labeled a success. If the money runs out before the end, it is deemed a failure. 

It’s all — or nothing.

ad space

A scenario that comes up $1 short is a failure. So is one that comes up $100,000 short. A portfolio exhausted in the final month of a 30-year retirement receives the same failing grade as one depleted after 15 years.

For that matter, the success side can be just as uninformative. A plan that finishes the period with $1 remaining is successful. So is one that leaves the retiree with $1 million —though those outcomes may say very different things about how much the retiree could have enjoyed spending along the way.

In effect, probability of success answers one narrowly defined question: In how many of our modeled scenarios did the portfolio avoid hitting zero before a date we selected? That may be useful information for academics or retirement planners, but it doesn’t strike me as the question actual people are trying to answer.

Goal ‘Oriented?’

Blanchett suggests an alternative: goal completion percentage. Rather than sorting every outcome into one of two buckets — success or failure — it measures how much of the desired spending was actually funded.

ad space

Consider the paper’s simple example: a retiree wants to generate $100 annually for 10 years. Across 10 modeled scenarios, only half provide the entire $1,000. That produces a probability of success of just 50% — a number likely to send most retirees scrambling for the nearest spending cut. And yet, averaged across those same scenarios, 96% of the desired spending is funded.

Same assumptions. Same outcomes. Very different — and arguably much more useful —description of the result.

Most people can probably get their arms around being able to fund 96% of what they hope to spend. They can consider what comprises the other 4%, whether they are willing to do without it and what adjustments might close the gap. That seems more tangible than being told that their retirement plan has a 50% chance of “failure.”

ad space

It also acknowledges something these models frequently overlook: retirees (not to mention non-retirees) don’t generally set a spending plan on the day they retire and then blindly follow it for the next 30 years. They adjust. They postpone a trip, replace a car later than anticipated, reduce gifts or make other changes as their circumstances evolve. A disappointing market doesn’t automatically cause them to spend their portfolio down to zero without noticing.

Indeed, Blanchett estimates that viewing the same risks through goal completion rather than a traditional probability-of-success threshold could allow some retirees to spend 20% more without taking on additional risk. That is potentially a significant improvement in retirement — not because the investments performed better, but because the measurement did. Not to mention the understanding of what the measurement means.

Successfully ‘Failing?’

That said, goal completion percentage isn’t a perfect measure. Knowing that a plan funds 90% of projected spending still doesn’t tell us when the shortfall occurs — or what kind of spending will have to be sacrificed. Funding 90% of a budget containing substantial discretionary travel is different from funding 90% of one already pared down to food, shelter and healthcare.

ad space

Still, it gives retirees something that probability of success generally doesn’t: a sense of the size of the potential problem. And that creates an opportunity to make informed choices rather than merely reacting to the absolutism of a passing or failing grade.

After all, a retirement that delivers 96% of what you hoped for may technically have “failed” by some measures.

But a measurement that can’t help folks distinguish that kind of outcome from financial catastrophe surely has.

• Nevin E. Adams, JD

 

Could ESG Options Undermine Participant Outcomes?

Despite surveys to the contrary, a new study finds that overall interest in ESG strategies by participants is “relatively weak” and “driven by naïve diversification.”

The difference may, of course, be attributed to the difference between what individuals say—and what they actually do. Unlike surveys that purport to capture participant (and plan sponsor) sentiments, the research by David Blanchett of PGIM and Zhikun Liu of the Employee Benefit Research Institute (EBRI) looks at the actual allocation decisions of 9,324[i] newly enrolled DC participants who are self-directing their accounts in a DC plan that offers at least one ESG fund. 

‘Weak Preferences’

They do so in a paper titled “ESG Fund Allocations Among New, Do-It-Yourself Defined Contribution Plan Participants,” they claim to find that overall interest in ESG strategies among these participants is “relatively weak,” with only 8.9% of participants having any allocation to an ESG fund and average allocations to ESG strategies of just 18.7% among those holding any ESG funds.[ii] Indeed, while they note “some clear demographic preferences for ESG funds (e.g., among younger participants with higher incomes),” they find that ESG allocations appear to be “primarily a function of weak preferences, driven by naïve diversification.”

Now, that hardly sounds like the heightened interest and engagement with those options that some participant surveys have captured (well, aside from that by younger participants with higher deferral rates and higher incomes). However, the research claims that the two factors which appeared to drive the largest allocations to ESG funds were not related to participant demographics, but rather the number of funds in the participant portfolio and the percentage of participants in the respective DC plan allocating to an ESG fund. 

If that seems a confusing descriptor, they found a “notable increase” in the probability of owning an ESG fund as the number of portfolio holdings increases—basically, the more funds the individual holds, the more likely he or she is to have an ESG offering among them. This tendency they characterized as attributable to “naïve diversification”—again, basically, if you’re simply picking a larger number of funds overall, then they concluded that the decision to allocate to the ESG fund is “likely based on a weak preference, not necessarily conviction in ESG.” Said another way, if you’re picking a lot of different funds, the more you pick, the better the odds that an ESG fund will (randomly) be among them.

On the other hand, those looking for a more optimistic future for ESG might take heart from their conclusion that “the fact ESG allocations increase as more participants in a plan allocate to ESG funds suggests plan interest effects could be an especially strong driver of future growth in ESG funds (despite relatively low usage today).” In fact, they noted a “notable plan interest effect, whereby ESG allocations are significantly higher in plans where general ESG usage is higher.”

Plan Sponsor Cautions

That said, the current decision-making by those participants appears to be “sub-optimal” (worse than you might expect) from a return standpoint—with the researchers here basically finding that participants who self-direct their portfolios have significantly lower expected returns than those using professionally managed investment options, such as target-date funds—something that proponents of professionally managed asset allocation solutions shouldn’t find surprising. To put it another way, those more likely to pick ESG funds are more likely to be the “do it yourself” (DIY) types—and those don’t do as well as those professionally managed solutions. This, as the researchers point out, can be an “important consideration for plan sponsors when adding ESG funds to the core menu to the extent they entice participants to self-direct their accounts.” So, adding an ESG fund might encourage more DIY investing by those interested in ESG—and that interest pulls them away from the professionally managed, higher-returning alternatives.   

In fact, an additional analysis suggests that those DIY participants have expected returns that are approximately 100 basis points lower than investors using professionally managed portfolios, such as target-date funds and managed accounts. And this, the researchers comment, suggests that adding ESG funds to core menus may create additional implicit return “costs” for participants—by adding those options that encourage participants to make choices other than professionally managed multi-asset options (e.g., target-date funds).[iii]

Overall, the researchers comment that their analysis paints a “mixed picture about the actual participant interest, and drivers of demand, for ESG funds in DC plans and suggests that plan sponsors should take a thoughtful approach when considering adding ESG funds to an existing core menu.”

Or—it seems fair to say—when adding (or subtracting) any funds at all.

- Nevin E. Adams, JD

---

[i] Of the 9,324 participants included in the dataset, only 833 had some allocation to an ESG fund, which is 8.9% of the total.

[ii] Among participants with an allocation to an ESG fund, the average allocation was 18.7%, with a standard deviation of 19.0%. The total average balance allocation to ESG funds is 1.7% (including all participants). There are only 56 participants (0.6% of the total) with ESG allocations greater than 50% of their balance and only 19 participants (0.2% of the total) with 100% of their balance in ESG funds. “In other words, even among participants who select the ESG funds, they almost always play a relatively supporting role as part of the overall portfolio.”

[iii] Some of the issues here are no doubt a consequence of current menu constructions. In the sampling studied, no plan offered more than five ESG funds, and the vast majority (approximately 76%) offered only one ESG fund. “This suggests it would be relatively difficult to build a diversified portfolio using only the ESG funds in DC plans currently,” the authors note. Moreover—and adding to the reality that it is “relatively difficult to build a truly diversified portfolio using only ESG funds”—they explain that roughly half of all ESG funds available are large blend funds. Only 13 of the funds (8.7% of the identifiable category total) are fixed income funds, and only 12 (8.1% of the identifiable total) are balanced funds. “The difficulty associated with building a diversified portfolio with only ESG funds has important implications on overall portfolio efficiency. If allocating to ESG funds requires participants to opt out of using a professionally managed portfolio option (e.g., target-date funds or retirement managed accounts), it may negatively impact future expected returns”—a cost the authors say they plan to quantify in a future work.