Archives for March 2025

A reader writes in, asking:

“This might be a question other people have too, but it’s certainly on my mind: you’ve spoken a lot about social security bridges using TIPs. If I were to use a TIPS fund instead of individual TIPS bonds, and the start of the bridge is about 1-2 years away, would it be better to use a shorter duration fund like VTIP rather than an intermediate duration fund?”

With TIPS mutual funds rather than individual TIPS, in theory the idea is that the fund’s average duration should match the average timing of the cash outflows.

For example, if the first payment were 1 year from now and the last were 8 years from now, we’d want an average duration of 4.5 years: (1+2+3+4+5+6+7+8)/8 = 4.5.

But a key point here is that the duration of a bond fund will stay more or less constant, whereas we actually want the duration to decline over time, as the desired cash outflows come closer and closer to the present.

In our example above, after one year has elapsed, we’d want the average duration to be 4 years: (1+2+3+4+5+6+7)/7 = 4. Another year later, we’d want the average duration to be 3.5 years: (1+2+3+4+5+6)/6 = 3.5. And so on.

So, for the best match, we need to use two different funds and adjust the balance between them over time (i.e., shifting more toward the fund with the shorter duration) in order to keep the weighted average duration where we want it.

But at that point, it’s actually more work than just buying 8 individual TIPS all in one sitting to begin with. (With a ladder of individual TIPS, the work is all up-front. Once they have been purchased, you just let them sit there until they mature and you spend the money. And if we’re talking about a fairly short ladder such as an 8-year Social Security bridge, you’re only buying a handful of different bonds.)

So in short:

• In most cases, I’d suggest just buying the actual individual bonds. It’s usually going to be less work, and you’ll have a better match between cash inflows (from bonds maturing) and cash outflows (i.e., spending the money).
• If we’re talking about a situation where you don’t have access to individual bonds (e.g., we’re talking about an employer-sponsored retirement account that provides access to TIPS funds but not the ability to buy individual bonds), then I’d follow the approach described above with a shifting allocation between two different funds.
• And if we’re talking about a situation where we’re placing a very high premium on simplicity rather than on risk mitigation, I’d just buy a fund that has a shorter duration than the current average duration of the cash flows, to account for the fact that the average duration of the cash flows will be declining over time. But it would be important to recognize that that isn’t a perfect fit, because the duration will currently be too short, and toward the end it will be too long.

A reader writes in, asking:

“Could you please clarify Morningstar’s new findings from Jason Kephart that delaying social security until age 70 while using one’s retirement portfolio may not be the optimal approach. Confusing after years of reading research advising of benefits to use portfolio to cover bridge from start of retirement to age 70 for highest income person.”

For reference, for anybody who hasn’t encountered it yet:

• You can download the actual paper here.
• Here’s an article summarizing the Social Security-related findings in the paper.
• And here’s an interview discussing those findings as well.

I would say that nothing in the Morningstar paper, as it relates to Social Security, should be surprising. However, one of the key tables (exhibit 1 in the paper, which shows up again later as exhibit 19, and which also appears in the article linked above) is causing some confusion, in particular due to using the term “bridge” in a way that is different than what is typical.

What’s a Social Security Bridge?

Typically, a Social Security “bridge” refers to allocating a part of the portfolio to some fixed-income product (most often, a TIPS ladder), to satisfy the extra level of spending that is necessary until Social Security kicks in. For example, you retire at age 62, but delay Social Security until age 70. And you buy an 8-year TIPS ladder (your “bridge”) that will provide an annual amount of spending during those 8 years of delay that is equal to what your Social Security payment would have been.

However, in the table in question, when the authors use the term “Social Security at 70 w/ Bridge” what they mean is a case in which something external to the portfolio performs the role of a bridge. In other words, this is essentially a “delay Social Security to 70 by working part-time until 70” strategy.

And what they find about that strategy is that it results in a higher amount of lifetime spending (and a higher combined “lifetime spending plus ending portfolio balance”) than a strategy of stopping work at 67 and filing for Social Security at 67. That’s pretty intuitive and uncontroversial. If you work longer, you’ll be able to spend more per year in retirement.

Delaying Filing, While Spending from Savings

The title of the interview above is, “Maybe You Shouldn’t Delay Taking Your Social Security Benefits After All.” The sub-headline states, “New Morningstar research suggests that in some cases, the advantage of delayed filing may be overstated.”

Both of those statements are factually true. Some people shouldn’t delay. And sometimes the benefit of delaying is overstated. (See for example, anybody who suggests that you get an 8% return when you delay.)

But I think it’s still very fair to say that most people should delay, and this paper doesn’t really pose a major challenge that assertion.

The big point that those two headlines are getting at is the following finding:

• In their “base case” scenario, they’re assuming a person retiring at 67, filing for Social Security at 67, and spending $73,000 in the first year of retirement, using a 60% stock, 40% bond portfolio.
• When they changed the assumption to retiring at 67 and filing for Social Security at 70 (thereby needing to spend more from the portfolio during those first 3 years), the spending level went up by ~5.5% (from $73,000 to $77,000), but the median ending portfolio value (at the end of the 30-year simulated retirement) went down by ~13.5%.

So the case they’re making is that if you care more about the ending portfolio value than about spending level, you would want to file earlier rather than later.

The result they find shouldn’t surprise you, when you look carefully at the assumptions. Specifically, they’re assuming spending from the 60/40 portfolio in order to fund the Social Security delay. In other words, stocks make up a major part (60%) of what they are assuming is given up, in exchange for more Social Security. And when you give up the risky asset with the high expected returns, in order to get more Social Security (i.e., a very safe asset), you get a higher level of safe spending, and a lower portfolio balance after 30 years. That’s exactly what we should expect.

For some people, that’s a good tradeoff. For others, it isn’t.

But before deciding that you shouldn’t delay Social Security, we have to take the analysis one step further. We have to look at (what is usually meant by) a Social Security bridge strategy, in which you specifically use fixed-income assets in order to fund the additional level of spending from the portfolio that’s necessary for the years of delay. The paper doesn’t look at that question specifically. But when that question is addressed in other research, what we typically see is that trading bonds for Social Security results in, for most people, an increase in the total available dollars over their lifetime (i.e., an increase in spending, an increase in ending portfolio balance, or some combination of the two).

Other Relevant Assumptions

Before closing, there are two additional very important assumptions in the paper. Neither is a flaw, as any model must make simplifying assumptions. But they’re both relevant in terms of the outcome here.

First, the paper is not considering the interplay of benefits for a couple (i.e., spousal and survivor benefits, as applicable). When we do consider those, what we see for married couples is that the case for the higher earner in the couple to delay is considerably stronger than for an unmarried person, and the case for the lower earner to delay is considerably weaker than for an unmarried person.

Second, the paper assumes a fixed rate of inflation (2.3%) per year. One of the major benefits of Social Security (and thus one of the benefits of maximizing your Social Security by delaying) is that it adjusts each year based on inflation. In other words, delaying Social Security gives you some additional protection against inflation risk. And the model in the paper isn’t accounting for that, because it’s assuming there is no inflation risk (i.e., fixed inflation).

Again, to be clear, I don’t think either of the above is a flaw of the paper. It’s just important to be aware of what assumptions are being used in the model before you take the resulting conclusions and apply them to your own personal financial planning.