Why are time-weighted returns called time-weighted?
Before I answer this question, it might help to explain where the terms “time-weighting” and “time-weighted” come from.
It’s not what you may think. No, Peter Dietz, who gave us the “Original Dietz” as well as the “asset-weighted Dietz” (aka, “Modified Dietz”) formulas was not the one who coined this term. It actually comes from the Bank Administration Institute’s 1968 publication, Measuring the Investment Performance of Pension Funds, the first “standard” for performance measurement.
At one time it was common for firms to indicate that they were “BAI compliant,” or “complied with the BAI standards.” Not the case any longer. However, these standards were quite important in establishing time-weighting as the preferred approach to evaluate the performance of asset managers. AND, two of their return formulas as well as the term “time-weighted” remain with us.
What did the BAI standards give us?
In addition to the term “time-weighted,” which I’ll address shortly, they provided us with three approaches to deriving rates of return:
- the Linked IRR, which is an approximation to the true, exact, time-weighted rate of return; today we know it better as “Modified BAI”
- the Exact method, which we know of today as, well, the “exact method,” or simply “time-weighted,” as it is the true, exact, TWRR. It involves revaluing the portfolio whenever a cash flow occurs
- the Regression method, which, to my knowledge, no one uses; it’s quite complex, and not worth the effort.
One other thing they gave us: a way to link returns across time. They didn’t use geometric linking; rather, their method involved taking an average of the subperiod returns, thus …
Finally: the meaning of the term “time-weighted”
Within the BAI Standards we find the following:
“The recommended rate is called ‘time-weighted’ because it is simply the weighted average of internal rates of return for the subperiods between cash flows with each weight being only the length of its corresponding subperiod.” <emphasis added>
This method involves annualizing the subperiod returns, and taking a weighted average, based on the length of the subperiod.
But, NO ONE does this.
NO ONE! NOBODY!
We all use geometric linking. This idea probably sounded like a reasonable approach, but it has since been deemed inappropriate. At least if we judge by the consensus of the investment community.
BUT, the term “time-weighting” has stuck, even though WE DO NOT WEIGHT TIME!
Imagine that! Kind of funny, right?
No wonder many find what we do confusing.