Last week, I wrote about the Ivy endowments, and why they have been shifting their alternative focus from Hedge Funds to Private Equity. I criticized this as a futile attempt to ignore the reality of a higher priced stock and bond markets, and that they should temper future return expectations.
There was plenty of pushback: Hedgies who disagreed with my entire assessment, PE managers who said it was the other guys with suspect accounting (not them).
A few people pointed out comparing a broadly diversified endowment with a domestic portfolio was wrong. My response was I always thought of 60/40 as a globally diversified (see various Vanguard mutual funds, etc).
As it turns out, I was wrong.
The 60/40 referenced within the MPI column I linked to was domestic only. Several 0f you pointed this out to me, and one of you steered me to what was in their 60/40 portfolio, in footnote #5:
[These results significantly lag a 60-40 portfolio]: “For this measure of a traditional balanced portfolio that is typically used by pensions to benchmark performance Wilshire used the Wilshire 5000 Total Market Index to represent U.S. equities and the Wilshire Bond Index for fixed income. That said, the result (9.1%) is similar to the 9.9% return of a domestic 60-40 portfolio using the S&P 500 Index and Bloomberg Barclays U.S. Aggregate Bond Index, rebalanced quarterly, which we calculated. (Emphasis mine)
Ugh. My stomach dropped when that was pointed out to me the week after the column ran. Of course you cannot compare a globally diversified endowment with a domestic-only portfolio, especially following a period of U.S. outperformance.
That is simply not a fair comparison.
Having a 57% exposure to alternatives is taking on a lot (too much?) risk, but the under-performance comparison I made to what was a domestic 60/40 now is a #fail. The better benchmark — an international 60/40 portfolio — has the endowments now out-performing over the prior decade.
My assumption was the 60/40 portfolio referenced as a benchmark was international. That assumption was in error — one I could have avoided had I read all of the footnotes…
