Historically, about 1.58% a year—which compounds to quite a lot over a career and retirement.
Even if you don’t consider yourself an “investor,” you likely have a 401(k), IRA, or even a taxable investment account. People, regardless of experience, use these accounts to grow their savings for a variety of goals: retirement, weddings, vacations, or some version of financial independence. Portfolio construction, or asset allocation, plays a significant role in how and when you can achieve those goals.
Most investors have never taken a deep analytical look at their asset allocation decision. At best, they answered a few questions and got sorted like a Hogwarts first-year into an age-based or risk-based portfolio.
You almost certainly have one of these in your 401(k) or brokerage account— a target date fund, a lifecycle fund, or a risk-based portfolio classified somewhere between conservative and “degenerate-gambler.”
Cheap Doesn’t Mean Good
A more practical portfolio should depend on a handful of factors: your goals, how long you have to invest, how much risk you can stomach, and what markets are currently paying you to take that risk. Most off-the-shelf portfolios — target date funds (100 minus age) and model portfolios (60/40) — handle the first three and ignore the fourth completely.
- TDFs ask how old you are or when you want to retire (529s work the same way). You outsource investment decisions based on when they think you’ll need the money, getting less risky over time. This strategy is so popular that it has amassed about $5.2 trillion in retirement money.
- Risk-based allocations ask how you feel about risk. They allocate a percentage to stocks and bonds based on your aversion to risk, and the allocation stays stagnant over the long haul.
Neither of these staple investor allocation strategies asks, “Is this expensive? Is this a good investment?” Just because a target date fund is cheap, or Vanguard says you should buy it, doesn’t mean that the things it’s holding are good investments.
Expected Returns Matter
To produce a better investment experience, we need to consider expected returns in addition to goals, risks, and time horizon. Target date funds and lifecycle funds fail to consider what current market conditions imply about future returns—they buy what they buy at any price, even if that price is absurd.
The piece these products miss is what stocks are expected to return above a safe alternative—usually inflation-protected government bonds (TIPS). When that gap is wide, stocks are paying you well to take equity risk. When it’s narrow, they’re not.
Haghani and White put it well in The Missing Billionaires: ignoring what markets are offering is like deciding to bet the same amount on a biased coin whether it’s 70/30 in your favor or fair at 50/50. The bet size shouldn’t be the same. Neither should your equity allocation.
A proper portfolio construction acknowledges this. Vanguard and Fidelity do not.
An example using historical data
Let’s look at a simple historical scenario that makes my point. Take two 55-year-old investors. One in 1981. The other in 2021. In 1981, real yields on the 10-year Treasury were close to 6%. In 2021, the 10-year real yield was -1.18%. Same age, same time until retirement—should they hold the same portfolio? In 1981, stocks faced a significantly higher hurdle to justify the risk than they did in 2021. Using an age-based or risk-based portfolio, the investors would hold the same thing.
If we were to build a better portfolio construction process, age would not be an explicit factor. We would look at the time horizon, risk aversion (how much loss hurts), and expected returns (what current prices tell us about future valuations). Age does matter, but only in the sense that it affects your earnings and wealth building. Financial markets don’t do anything special because you hit some age target that Fidelity or Vanguard picked for you.
So, what do we do about it?
Well, it’s a bit more work, but the juice is worth the squeeze, so bear with me. In the 1970s, Nobel Prize-winning economist Robert Merton worked out the math for what an optimal portfolio actually looks like. He made something now called the Merton share (of course it is). It depends on three inputs: the expected reward of stocks above a safe asset, the risk of those stocks, and how much risk an individual investor is willing to take. Age is not a factor in the equation. Age matters, but only because it affects your wealth and your remaining earning years—not as an input to the formula itself. That distinction matters a lot.
However, the Merton share is not perfect. The problem with Merton’s framework is that it’s a little abstract, even by finance professor standards. That is somewhat problematic—my clients already read my emails with about as much enthusiasm as the dentist’s, so I’m not going to say “hey, let’s find your coefficient of relative risk aversion,” and expect them to engage.
There is an easier way to do the same thing.
Fortunately, we have other ways to finagle this number. James Choi, a professor at Yale, and his coauthors built an analytical approximation—that a normie can run in a spreadsheet.
The Numbers
I promised this lengthy post would be worth it, so let’s look at the value added. Haghani and White ran a test from 1900 to 2022, comparing a static 65/35 stock/bond (risk/safety) portfolio to a dynamic allocation based on the highlighted points above. Both were rebalanced monthly. Both used the same assets. The only difference was that the dynamic version paid attention to the “excess earnings yield,” which is a fancy way of saying what stocks pay above inflation-adjusted government bonds.
Over the course of their study, the static 65/35 delivered an 8.47% annualized return. The dynamic version delivered 10.05%. That’s a 1.58% annual difference! Compounded over the observation period, $1 invested in the static portfolio grew to $22,158. In the dynamic version, it grew to $130,879. The same starting amount yielded six times as much.
Intuitively, this makes sense. If stocks and bonds are offering relatively the same yields, why bother holding stocks? On the other hand, if bonds are offering negative real yields, why bother holding a guaranteed loser? We don’t need a finance professor to answer these questions.
Different approach, same results
We can come to the same conclusion if we look at this from another angle: the consumption side. Choi and his coauthors wanted to look at changes in annual consumption of a dynamic process vs the status quo. The “100 minus your age” rule (target date funds) comes out to about a 2% permanent loss on annual spending. A static 60/40 portfolio comes out to roughly a 3.75% annual loss in spending.
So, we have two different processes showing that age- and risk-based portfolio constructions leave meaningful value on the table. Obviously, past performance doesn’t predict future results. However, the research shows that a meaningful gap occurs when we leave expected returns out of the equation. The framework you’re handed at your 401(k) provider is the cheapest solution, not the best for you.
But wait, there’s more
That gap becomes even wider when we include momentum in the asset allocation model. Haghani and Dewey ran a similar analysis from 1975 to 2013 across twelve asset classes and found that adjusting allocations based on valuation added about 0.86% per year to returns compared to a static portfolio. Adding in a momentum factor added over 2.5% a year over a static allocation.
This sounds like market timing, which we all know is bad.
It’s not. Not really. Market timing depends on markets being wrong(ish)—the idea that you can find something over- or undervalued before others do and profit from it. Most retail trading is some version of this, whether it’s chasing recent returns or some WallStreetBets Reddit narrative.
A dynamic allocation works differently. It does not require some informational or research edge. Even in perfectly efficient markets (you have no informational edge), the expected reward for taking equity risk changes over time as the supply and demand for capital change. I made this point earlier with our 55-year-old investors in 1981 and 2021.
Target date funds aren’t all bad
None of this is to say target date funds and risk-based portfolios are bad. They aren’t. They’re cheap, automatic, and a meaningful upgrade over what most people would do on their own—which is usually nothing, or something bad. For an investor with no financial expertise, a target date fund is a small miracle. I think a target date fund is great for a worker just entering the workforce, as it is cheap and holds almost all stocks—it’s better than nothing.
But “better than nothing” is not the same as “right for you.” If you can remember Carson Daly’s TRL, you’re probably old enough to give your life’s savings more thought than ‘100 minus your age’—you need to start taking your asset allocation more seriously. After all, asset allocation explains most of your returns.
Most of my clients came to me because their financial life had something weird going on—an inherited IRA nobody explained to them, equity comp that never quite got integrated into a plan, a side business generating income that made everything more complicated, or a life that is getting increasingly busy. In situations like that, the value isn’t in any single decision. It’s in coordinating tax planning, investment management (what we talked about here today), and everything else, so they actually work with each other instead of past each other. If that sounds like your situation, take a look at how I work or reach out, and we can talk through the specifics.