If the answer is property investment, what was the question?
Even though by now, when one would have thought that we really should know better, there are still too many pension fund asset liability studies being carried out that seem to confine themselves, when looking at the asset side of the equation, to just a study of equities and bonds. It is as if other asset classes really don’t matter.
Talk to trustees and you will see that the advice they receive from many consultants is that the only choice for a mature pension fund is how much to put into bonds and how little to put into equities.
Other assets classes barely get a mention. And yet it doesn’t seem that long ago that real estate, for example, often accounted for 15% or more or a typical pension fund asset allocation.

Certainly any pension fund which had a heavy exposure to property throughout the 1990’s has done remarkably well in the twenty-first century. Even now with yields still high, prospects for commercial property seem quite
reasonable.
I really had thought that the case for diversification in asset strategy was now fully accepted. Certainly, we know that bond and equity investment benefit from a global remit rather than just a domestic bias but overseas investment in either equities or bonds brings very little in the way of diversification advantages. However many of the various alternatives including hedge funds and real estate do provide low correlation to the “main” asset classes. So if we can get good diversification, good yields and apparently good growth prospects, why do so many funds carry so little property in their portfolios?
If we analyse what pension funds might want to see in an alternative asset we would probably look for relatively low volatility, a low correlation with equities or bonds, a reasonable hedge for liabilities, a relatively secure income stream, a liquid investment and low dealing and administration costs.

Now property does not offer all of these things but liquidity need not be too much of problem, as pension fund liabilities often stretch out 20 or 30 years into the future, as long as it is not the only investment a fund has. Similarly, dealing and high stamp duty can be amortised over a number of years and become less of an issue. It is less easy to find a way of reducing administration costs unless one considers indirect investment in the sector.
Against property one might say that it suffers from a much higher tracking error against a benchmark compared to some other investments and its volatility is not as low as the published figures suggest. Certainly pricing and valuation methods seem to me to be more of an art than a science, which must frighten off some potential investors. However they are many good professional fund managers out there offering opportunities to obtain alpha as well as the plain vanilla exposure to the sector (beta).
So it seems that the biggest problems with putting property into an asset liability study and especially into an optimiser are ignorance and agreement of assumptions.
A common approach seems to think of property either as an alternative to bonds or an alternative to equities and to express returns and volatility in terms of one or the other.
At the moment much property investment offers a higher running yield than most comparable bonds and yet still offers the potential for capital appreciation. Many equity advocates argue that the published property volatility figures are misleading as pricing information is so poor it heavily understates the true volatility of the sector. Interestingly the published volatility figures for property are often better than bond volatility statistics. Now true volatility figures may indeed be higher than the published figures but so what. For a long term investor like a pension fund this may actually be a huge advantage.
One of the biggest problems pension funds face today is the dilemma between the need for long term investment and the problem of short term reporting and being made to mark to market. Now if we can find an investment that has low published volatility surely this will help to even out the risks of other investments and the more property (especially with long leases) a fund holds the less volatile the whole portfolio will become.

Most pension funds do not have the luxury of a greenfield situation when preparing their asset liability model, so existing holdings and prejudices tend to be fed into the model. Inevitably therefore the model will only be as good as the parameters and assumptions (and prejudices) fed into it. You may even feel that many consultants fudge ALMs to give the results they want to see anyway. However, it really is incumbent on trustees to use as wide a range of reasonable assumptions and as wide a range of asset classes as possible. If they do they may find it quite difficult to argue that property has no place in a properly diversified pension fund portfolio.