Claiming Social Security Early to Invest It: What Rate of Return (Discount Rate) Should We Assume?

A reader writes in, asking:

“In your social security talk at the White Coat Investor conference, you mentioned that when considering whether to delay social security or claim it early and invest it, the appropriate rate of return to assume is the rate of return from TIPS bonds. But I didn’t catch the reason for that. Would you consider discussing that for an article?”

As a bit of background for other readers: when deciding whether to delay Social Security or claim it now and invest the money, you have to make some assumption about the rate of return that you would earn on invested benefits. The higher the rate of return you assume, the more advantageous it is to claim benefits early.

Alternatively, you can think of the analysis as, “what part of my portfolio would I spend down in order to delay Social Security? And what would be the rate of return that I’d be giving up by no longer having those assets in my portfolio?”

For most people, the relevant rate of return is the expected return on the bond portion of their portfolio. The primary exception would be the household that has a 100%-stock allocation and wants to take on even more risk. (Relatedly, by default the Open Social Security calculator uses the current 20-year TIPS yield as the discount rate, but you can select a different discount rate by checking the box at the top of the page for additional options.)

By way of analogy, imagine that your portfolio currently has a 70/30 stock/bond allocation. And imagine that Vanguard releases a new bond fund tomorrow. When evaluating the new fund to determine whether to add it to your portfolio, would you want to know whether the fund’s expected return is:

1. At least as good as the expected return of the stock portion of your portfolio,
2. At least as good as the weighted average expected return of your 70/30 portfolio, or
3. At least as good as the expected return of the bonds fund(s) in your portfolio with similar risk?

Question #3 is the relevant question. If the new fund had, for example, a 0.5% higher yield than a given bond fund in your portfolio with a similar level of risk, it would probably merit inclusion — even if the new fund’s expected return is considerably lower than the expected return for stocks and lower than the weighted average expected return of your 70/30 portfolio.

Same thing with Social Security.

If your portfolio is currently 30% bonds and you are currently considering filing before 70, you can instead choose to swap some of those bonds for more Social Security. In practice, that usually looks like spending more quickly from bonds than you otherwise would have done — “selling” bonds in order to “buy” Social Security.

For somebody who isn’t spending from their portfolio because they have income from other sources, it means shifting the allocation of the portfolio toward stocks. Counterintuitive, but because money is fungible it’s the same concept.

That is, if you have, for example, decided that filing at 62 and using a 70/30 portfolio is an appropriate level of risk, what about instead filing at 70 and using an 80/20 portfolio? (To be clear, I’m just making up the 80/20 figures here. The actual percentage would depend on the size of your Social Security benefit relative to the size of your portfolio. The idea is to keep the same level of risk, but have more Social Security and less bonds.)

Spending Down Bonds to “Buy More” Social Security

As financial planner Allan Roth has been arguing for years (here for example), it doesn’t usually make sense to own bonds earning a certain rate of interest while simultaneously paying a higher rate of interest on your mortgage. It’s generally advantageous to sell the bonds and pay down the mortgage.

A similar concept applies for Social Security.

For an unmarried male, the necessary rate of return that would make claiming Social Security at 62 as good as claiming at 70 is about 1.7% above inflation. For an unmarried female, the necessary return would be about 2.9% above inflation.* If delaying Social Security provides such an expected return, with a low level of risk, it doesn’t usually make sense to forgo additional Social Security in order to continue owning bonds that have a lower expected return (or a similar expected return and a higher level of risk).

*These rates of return use the SSA’s 2014 period life table for life expectancies. This understates the average life expectancy somewhat. As a result, the necessary rates of return would actually be somewhat higher. For a married couple, the “breakeven” rates of return will vary based on their difference in earnings history and difference in age. In general though, the breakeven rate of return for the higher earner will be significantly higher than for an unmarried person (meaning it’s usually super advantageous for this person to delay) and lower for the lower earner (meaning it’s less advantageous for this person to delay).