I have been developing concerns about the possibility that hedge fund investment flows have become a destabilizing force in world financial markets. Following the maxim “think globally, act locally,” I decided to take a look at how the behavior of our local pension funds may be one small part of that phenomenom. And I have some recommendations to make on behalf of San Diego County residents and the world at large.
There are any number of financial instruments that would allow you to construct an investment portfolio that could show high returns with limited variability for many years in a row, even though the investment strategy could actually involve enormous risk or even negative expected returns. An example is MIT Professor Andrew Lo’s Capital Decimation Partners. Lo’s hypothetical hedge fund simply sells far out-of-the-money puts, earning spectacular returns until the day that it is inevitably wiped out completely.
Such financial instruments could be particularly problematic given a potential conflict between the incentives facing an individual pension fund manager and the best interests of the future retirees whose assets he or she is managing. If the manager can turn in well-above-market returns five years in a row, the manager is likely to be spectacularly rewarded over that period. Whether the beneficiaries are actually paid many years later down the road will be somebody else’s problem.
When I heard about the disastrously irresponsible investments made by the Amaranth hedge fund, my first reaction was, who would be so stupid to have put up the margin requirements for such a scheme? The answer turned out to be found in my own backyard– the San Diego County Employees Retirement Association apparently donated over a hundred million dollars to this worthy cause.
I took a look at the 2006 SDCERA Annual Report to try to find out what was really going on. If you only look at the “summary of retirement portfolio by manager/asset type” on page 54, all appears healthy. Two billion dollars, or 27% of their $7.3 billion portfolio, were reported to have been held in the form of domestic equity, split among a dozen different large, diversified equity funds. There’s another 25% in international equity, 26% in fixed income, and 10% for real estate and “alternative equity”. There are some other categories such as “commodity swaps” and “overlay” that worry me a little, but these don’t amount to that much of the total. So where do you squeeze a couple hundred million for Amaranth in there?
The really interesting stuff is to be found on page 55, which is labeled “summary of derivative financial instruments.” Here Amaranth is included as one of a dozen entries in a group of “alpha engine managers.” The total market value of these alpha engine managers comes to $1.5 billion, which would amount to 20% of SDCERA’s net assets. That’s on top of the $1 billion (14%) in “exposure for currency overlay”, $350 million for “exposure to policy overlay”, and close to another billion in a few other exotic categories.
Obviously these two pages must be counting the same assets twice, or rather, page 54 is not regarding these “derivative financial instruments” as part of SDCERA’s “retirement portfolio.” Instead the annual report seems to take the position that the fund is still holding the $2 billion in domestic equity, and these other details are some clever little tricks the fund employs to allow it to earn a higher rate of return on those stock holdings than you would if you followed the old-fashioned practice of, you know, hanging on to your assets.
What the fund is actually doing is selling risk. SDCERA is receiving a payment for taking exposure to the risk of what is hopefully a low-probability event, just as you would if you were selling far out-of-the-money puts, and pledging the fund’s hard assets as collateral with which to play the game. The result is a far riskier portfolio than would be suggested by a naive reading of the fund’s purported asset allocation from page 54.
But as far as I’m aware, there’s no place you can go to find an audited statement of the assets and liabilities of these “alpha fund managers”, so there’s no way to verify exactly what the nature and magnitude of the risk is that’s being palmed off on the county. We have only the word of the county fund managers that they’re doing something smart. They in turn are likely basing their own confidence primarily on the word of the alpha fund managers. At least in the case of Amaranth, we know how much that word is worth, and it ain’t $233,830,268.
And that’s just the San Diego County pension fund. The pension fund of the city of San Diego is an even more gruesome story, which perhaps I should take up another time.
Buyer beware, says Dave Altig. Actually, being a man of great erudition, or at least someone who reads Forbes, Dave says it in Latin. But when the buyer is acting on behalf of the government, to me it seems very appropriate for the government to set statutory limits on the extent to which managers’ interests can be permitted to deviate from those of the beneficiaries. Specifically, I recommend that California’s County Employees Retirement Law be amended to specify that if one calculates the sum of all investments, pledges of collateral, financial liabilities and exposures, and margin deposits and calls made by a county retirement fund in institutions for which there are not publicly available annually audited balance sheets of those institutions’ assets and liabilities, the sum of all such commitments across all such institutions can not exceed 10% of the retirement fund’s total gross assets.
Maybe Dave can help me to say that in Latin.