Vincent, you’ve recently written a great paper on portfolio construction and real estate debts.
What are the key takeaways?
The premise (of the paper) was that modern portfolio theory (MPT) by Markowitz1 – which is the popular framework for constructing investment portfolios – is based on the idea that the risk of an investment is the volatility of the return that it provides. If you look at a stock portfolio, that seems exactly right.
The price goes up, the price goes down, and at any given point you can buy in, or you can sell out, and the value that you see on the screen is the value at which you can buy. The conclusion of that theory is that diversification across different stocks that have low correlation with each other is good, and the more diversification, the better.
When we look at real assets that are not liquid – that don’t trade daily, that at best have a valuation that’s a proxy for what you might sell it at, although the liquidity means it would take time to actually act on that valuation and sell it – it’s worth asking whether or not the conclusions hold if the premise doesn’t hold, or the assumptions don’t hold.
We concluded that of course some diversification is a good thing, but the purpose of diversification in the Markowitz world is to achieve the market return – to diversify away the specific risks of individual stocks and get to a market return.
That is hard to do in real estate where there isn’t really a market return that is as observable as it is in the stock market. You have an index you can target, but the index holds, by definition, assets that you cannot own. They are owned by other parties, not by you. So, you inhabit a world where the benchmark is, at best, ten years, or in real estate debt where it’s completely unobservable.
So, the purpose is not to achieve the market return, but rather to build a portfolio where some of the specific risks are diversified away, but not necessarily to achieve some utopian market return.
On top of that, in real estate, the size of an investment is often dictated by the size of the building, and so you cannot say: ‘I want 10 million of this, and 20 million of that’. You might have to take 15 or 30 (million) depending on the size of the asset. The same is true in real estate debt.
What you end up with is a portfolio that is skewed from what you might think of as an ‘ideal allocation’. The purpose is to ask ‘where should it skew?’, and the way we have determined that is to say it should skew in such a way that the largest assets are those that are the best credits. There, the benefit of diversification is typically lower because the chance that the future return that you will actually get deviates from your underwritten level is much smaller.
So, if you have investment grade, private loans, sub-investment grade private loans in one combined portfolio, the sub-investment grade assets should be smaller. Those investments should be smaller than the investment grade ones, and that is a continuum.
So, the view was, if modern portfolio theory permeates all thinking of investors, it’s worth taking a step back and recognising that there are certain elements in real assets that are fundamentally different and require a slightly different approach.
We still conclude that some diversification is a good thing. We wouldn’t go as far as to say it’s not. But I think if you look at senior loan portfolios, you probably get more benefits from careful asset selection than from heavy diversification.
