I’ve noticed an interesting phenomenon, which I call “common knowledge about common knowledge” (doesn’t exactly roll off the tongue, I know). It’s when some piece of conventional wisdom becomes so banal that someone decides to refute it, and then the refutation becomes as widely known as the original factoid, if not more so.
To give an easy example, “everyone knows” that vitamin C strengthens your immune system. But “everyone” also “knows” that’s ridiculous folk wisdom based on a single poorly conducted study in the early 20th century.
The problem is this can cause people to overcorrect, and end up saying “actually vitamin C has no health effects.”
Except vitamin C really is essential to human health!
There’s no Fidelity “dead accounts” study
An ancient piece of financial lore holds that one day, the beancounters at Fidelity decided to study the performance of every account at the firm in order to identify what characteristics were shared by the accounts yielding the highest returns. Small stocks, big stocks, domestic stocks, foreign stocks, frequent trading, infrequent trading, they looked at everything.
After months of crunching the data, the beancounters made a shocking discovery: there was one account type that consistently outperformed all the others. The accounts of dead people! It turned out that people who did no trading in their accounts whatsoever had higher long-term returns than any other strategy.
Now, if you think about this for even a moment, you’ll realize it doesn’t make any sense. How many accounts could Fidelity possibly have in the names of dead people (another version sometimes includes “people who had moved and forgotten about their accounts”)? If Fidelity could easily identify dead accountholders, why hadn’t they notified the appropriate authorities already?
The Fidelity study is fake, not wrong
Whenever I see a reaction like this building, whether it’s “actually vitamin C doesn’t make you healthy” or “actually a dead person’s account would not achieve better returns than an actively managed account,” I like to take a second to ask, what if the conventional wisdom was right all along?
To find out, I used Morningstar’s portfolio tool to check the performance of two portfolios: a simple 60/40 portfolio using the Vanguard 500 (VFINX) and Vanguard Total Bond Market (VBMFX), and a Boglehead 3-fund portfolio substituting a 12% allocation to Vanguard Total International Stock (VGTSX). The 2-fund portfolio performance goes back to 1986, and the 3-fund portfolio to 1996. In both cases, I selected reinvested dividends and capital gains.
For each portfolio, I checked its annual return (and the final value of an initial investment of $100,000) with a monthly, quarterly, semiannual, and annual rebalance, and with no rebalance at all.
For the 2-fund portfolio dating back to 1986, the average annualized returns (and final value) for each rebalance selection were:
- Monthly: 8.73% ($1,526,134)
- Quarterly: 8.79% (1,553,836)
- Semiannual: 8.71% ($1,517,894)
- Annual: 8.73% ($1,526,244)
- No rebalance: 9.02% ($1,665,050)
Here are the same values for the 3-fund portfolio since 1996 (the pattern is similar for the 2-fund portfolio over the same period):
- Monthly: 7.06% ($486,314)
- Quarterly: 7.16% ($496,945)
- Semiannual: 7.28% ($510,056)
- Annual: 7.32% ($514,528)
- No Rebalance: 7.07% ($487,180)
I like this data precisely because it lends itself to multiple interpretations. Looking at the two-fund portfolio since 1986, you’d conclude no rebalancing at all yields the highest returns — the fake Fidelity was right all along, the best investor is a dead investor!
On the other hand, the three-fund data since 1996 suggests annual rebalancing yields the highest returns — having a pulse matters after all.
If you’re reading this, I’ll go ahead and assume you have a pulse, so allow me to split the difference: if you’re invested in a simple 2-fund, 3-fund, or 4-fund portfolio, you’re almost certainly rebalancing too often. If you’re rebalancing monthly, consider switching to quarterly. If you’re rebalancing quarterly, consider switching to semiannually (perhaps on the equinoctes?). And if you’re rebalancing semiannually, consider switching to annually. If you’re rebalancing annually, your blood is already ice cold and you don’t need my advice.
The logic is simple: when you rebalance a portfolio, you’re by definition selling your winners and buying your losers. But the tendency of assets, especially over short periods, is for winners to keep winning and losers to keep losing, meaning frequent rebalances move you out of assets that are performing well and into assets that are performing poorly.
There are good reasons to do that, for example if you are targeting a particular balance in lower-performing, more stable assets, in anticipation of a near-term expense. One of the benefits of diversification is that it tends to increase your risk-adjusted return. But while risk-adjusted returns are an important metric when designing a portfolio, you can’t eat them. For that, you need the ordinary kind.