Illinois Teachers’ Pensions Increase Allocation to Alternative Investments in Order to Do the Impossible

Filed Under (Retirement Policy, U.S. Fiscal Policy) by Jeffrey Brown on Apr 21, 2011

Just a short note today (which I learned from reading Institutional Investor) to point out that the Illinois Teachers’ Retirement System on April 8 approved a change in portfolio allocation.  Specifically, they are going to increase by 5 percentage points the fraction of the pension portfolio going to hedge funds and private equity.  The increase is coming primarily from a reduction in domestic equities from 26% to 20% of the portfolio.

What I find ironic (indeed, it would be amusing if the stakes were not so high) is the statement by the Chief Investment Officer that they are doing this to “minimize risk and maximize returns.”

There is a major problem with this statement.  Namely, it is impossible to do both.

It is possible to minimize risk, while holding the return constant.  Or one can maximize returns, while holding the risk constant.  Indeed, these are two different ways of – in financial economics lingo – to get to an “efficient” portfolio.

But it is impossible to do both simultaneously.  In financial markets increased market risk and increased expected returns go hand-in-hand.  As Frank Sinatra said about love and marriage, “you can’t have one without the other.”

 

You Can’t Diversify Your Investment in Your House!

Filed Under (Finance, U.S. Fiscal Policy) by Don Fullerton on Mar 5, 2010

Jim Berkovec was a colleague and co-author of mine when I was at the University of Virginia.  He then became an economist at the Federal Reserve Board, and later moved to Freddie Mac.  Just a couple of months ago, I got some very bad news that Jim had died in a bicycle accident.  At the young age of 50, he was out for some exercise near his home in the Washington DC area.  He was going too fast down a hill and around a curve, and hit a slippery patch.

So in honor of my friend Jim, I want to tell you about our joint research project, a 1992 paper in the Journal of Political Economy, called “A General Equilibrium Model of Housing, Taxes, and Portfolio Choice.”

This research starts with the observation that a wise diversification strategy would include not only U.S. stocks and bonds, but also real estate (and even assets in other countries).  You don’t want all your eggs in one basket.  And for real estate, diversification would mean spreading your investments across properties in many different locations.  That kind of diversification might be possible for rental real estate, but for two reasons it is not possible for your owner-occupied house. 

First of all, you can’t live in all those diversified locations.  The fact that you own the house where you live means that you must have all your eggs in one basket.  And the federal government encourages people to own their own homes, through tax advantages for homeownership, so the government is discouraging diversification.  When a bad event occurs, the homeowner loses the whole basket.

A second reason the homeowner cannot diversify properly is that the amount of the housing investment must be tied to the amount of housing one wants to consume.  You can’t pick the house size that is best for your investment portfolio, if you have to choose the house size that fits your family or other life style choices.  Some people like to spend their money on vacations or electronics, while others like to live in a great house.  Choosing based on those preferences means not balancing your investment portfolio. 

Having an unbalanced portfolio means taking extra risks, and it makes us worse off.  If you invest in one stock or one asset, you get the normal expected rate of return and a large variance (large risk).  If you invest in a diversified portfolio, you still get the same normal expected rate of return, and low variance (low risk).

That might all seem like an unavoidable fact of life, if folks are to own their own homes.  But government policy can make it better or worse.  Just compare the following alternatives.   If government were to take away some housing incentives by disallowing property tax deductions or mortgage interest deductions, then you would still own your whole house.  Those measures do not help share this risk.  But suppose that government were to treat housing just like any other asset.  That is, suppose you had to treat yourself as a renter in your own home: (1) calculate the amount your house would rent for, if it were rented to someone else, (2) include that imputed rent as income on your income taxes, and (3) continue to deduct mortgage interest paid and property taxes paid, just as a landlord would do.  You could also (4) take additional deductions for maintenance and insurance, as a landlord would do.

If homeowners were all in a 33% tax rate bracket, then the government would be taking one-third of all returns to all housing investments, and it could then return that money to everybody through some other tax reduction.  We would NOT be paying higher taxes, overall, but the net effect is that we each would “own” only 2/3 of our own home, and we would each – through our government – become part owner in one-third of everybody else’s home.

In other words, we all would be better diversified, effectively sharing in the returns to all real estate all around the country, with less of an investment in the one house we choose to live in.  According to the economic model in my research with Jim Berkovec, that policy would make us all unambiguously better off.

Jim was a pretty smart guy, and a great researcher and friend.  We’ll miss him.  Here is the official “abstract” of the more technical version of this research paper:

We describe a model in which rental and owner housing are risky assets, tenure choice is endogenous, and each household is constrained to consume the same amount of owner housing that it has in its investment portfolio. At each iteration in the search for an equilibrium, we determine the new taxable income for each of 3,578 households (from the Survey of Consumer Finances), and we use statutory schedules to find the marginal rate and tax paid. Equilibrium net rates of return are major determinants of the amount of owner housing, but a logit model indicates that demographic factors are the main determinants of ownership rates. In our simulation, taxes on owner housing would raise welfare not only by reallocating capital but also by the government’s taking part of the risk from individual properties and diversifying it away. Measures to disallow property tax or mortgage interest deductions do not help share this risk. Simulations of the 1986 tax reform indicate a small shift from rental to owner housing and welfare gains from reallocating risk.