I confess to being late in getting around to the announcement by MSCI Barra, which was reported on January 12 in Global Pensions, of their intention to add 20,000 indices in the coming months, which would bring the total to over 100,000:
Global index provider MSCI Barra will be launching some 20,000 new indices over the next few months in order to remain in step with growing trends in pension fund investments….
MSCI Barra COO David Brierwood explained the reasons behind the restructure. “Financial instruments are constantly being created to lower the total cost structure of investments,” he said, “and this therefore has allowed pension funds to invest in a wider spread of assets.”…
Currently MSCI Barra has around 80,000 indices and Brierwood said the number would be sure to reach 100,000 within the next few months and possibly go beyond that figure in the future.
There are a couple of ways to look at this supposed progress. One it the tack taken by Warren Buffet in his 2005 letter to Berkshire Hathaway shareholders, where he points out: “With unimportant exceptions,… the most that owners in aggregate can earn between now and Judgment Day is what their businesses in aggregate earn.” He then goes through the fate of a hapless family, the Gotrocks, who have various Helpers, first, brokers, next investment managers, then financial planners and institutional consultants, and finally private equity and hedge fund managers. Buffet concludes:
A record portion of the earnings that would go in their entirety to
owners – if they all just stayed in their rocking chairs – is now going to a swelling army of Helpers. Particularly expensive is the recent pandemic of profit arrangements under which Helpers receive large portions of the winnings when they are smart or lucky, and leave family members with all of the losses – and large fixed fees to boot – when the Helpers are dumb or unlucky (or occasionally crooked)….Today, in fact, the family’s frictional costs of all sorts may well amount to 20% of the earnings of American business. In other words, the burden of paying Helpers may cause American equity investors, overall, to earn only 80% or so of what they would earn if they just sat still and listened to no one.
For those of us who aren’t Warren Buffet, and might want more specific reasons to take the proliferation of indices with a grain of salt, here are a few:
Indices aren’t asset classes. There has been a good deal of academic work on the fact that the more asset classes one is invested in, the better off one is on a risk/return basis (this notion is sometime described as the efficient investment frontier). As a crude generalization, being in more asset classes does not increase your returns, but lowers risk. However, the peer reviewed papers have looked at stocks, bonds, cash, sometime real estate, occasionally commodities. That leaves you with a number considerable lower than 100,000.
The problem is that investors (encouraged or one might say misled by the vendors of narrowly-focused funds and indices) have come to regard investment styles, say health care stocks or a value fund, as being the same as an asset class. They will buy several types of equity or bond funds, thinking they are getting further diversification, when all they are doing is incurring more fees than if they bought a simple S&P 500 or a Russell 2000 index fund. While private equity and hedge funds are treated as asset classes, there isn’t much work to support this view (although, per an earlier post, some claim that hedge funds are a good way to get “synthetic beta.”)
Investors don’t get index returns. Even with big indexes that trade highly liquid securities, like the S&P 500, index funds will show tracking error due to transaction costs and the need to have some cash on hand to manage redemptions. The tracking error can be large with more exotic funds. The blog Seeking Alpha pointed out that “the MSCI Hedge Invest Index, the investable index fund that tracks the MSCI Hedge Fund Composite Index is up 2.9% (year-to-date), while the index itself is up 5.1% during the same period.”
Covariances are not consistent over time. The DJIA and the S&P 500 have been around a long time, and various companies have entered and exited. But what can we make of a new index? How will MSCI Barra go about creating its history? Investors are leery of black box trading strategies that rely on backtesting to prove their results, but seem content to accept it with indices.
Personally, I would be leery of any analysis of covariances that used an index with less than 20 years of history. Why? These covariances aren’t as stable as the consultants would like you to believe. Indeed, they may not be meaningful at all. From The New Palgrave Dictionary of Money and Finance (no online source):
Most asset-pricing theories link expected returns on assets with their expected variances and covariances. A large body of literature in empirical finance, as well as episodes such as the stock-market crashes on 1929 and 1987, make it clear that these conditional variances and covariances change randomly over time. Understanding the dynamic behavior of expected returns therefore requires an understanding of conditional heteroskedasticity in returns.
I don’t pretend to understand what the third sentence says. But I do understand what the first two sentences say, that the expected returns on assets and covariance of returns among assets vary. That isn’t something you see acknowledged in the slightest way in most investment manager and consultant presentations. They treat the covariance of the investment style under question versus the broader index (say an energy fund versus the S&P 500) as being a single value, not something that is variable (remember, Palgrave said “changes randomly over time“). Their whole methodology goes out the window if it doesn’t have any predictive value.
And I bet you there isn’t anyone at the vast majority of these investment consulting firms who knows what that third sentence says either. That in turn begs the question of whether they are selling expertise or shamanism. Warren Buffet has already given us his answer.
