Edward Chancellor: Asset Allocation Insights

by: IndexUniverse Europe

Edward Chancellor is a member of the asset allocation team at GMO LLC. Prior to joining GMO he worked as a financial journalist, receiving the 2007 George Polk Award. Chancellor is the author of acclaimed books on the financial markets, including Crunch Time for Credit (2005) and Devil Take the Hindmost: A History of Financial Speculation (1999). In an interview last week with Paul Amery, editor of IndexUniverse.eu, Chancellor talked about his firm’s approach to asset allocation.

IU.eu: Edward, how does GMO approach asset allocation from a theoretical perspective?

Chancellor: Our asset allocation approach is very much valuation-driven. Many allocation models are based on historical asset class returns and correlations, perhaps followed by an optimization process. While this approach is fine in theory, in practice it tends to lead to disaster, since it ignores the fact that asset class returns are primarily a function of starting valuations.

For example, ten years ago there was great euphoria about equities, everyone was reading Jeremy Siegel’s “Stocks for the Long Run”, and all this was based on hundred-year historical returns from share investment. But if you decomposed the historical equity return you could see that the starting point was a low double-digit price/earnings ratio, and two thirds of the overall return came from dividends. Given that the dividend yield a decade ago was less than half the long-term average, and that valuations were much higher than those in the past, forward-looking returns were much less attractive than those you could expect from extrapolating history.

The study of market bubbles and busts, from both a historical and behavioral perspective, reinforces confidence in our valuation models and enables us to add a certain contrarianism to our investment approach. So, a decade ago, the observation of various kinds of euphoria in the markets would have added to the conviction that we were indeed in a bubble.

It’s always worth examining asset allocation claims that are based on historical returns with some scepticism. Take commodities, for example. There’s been a lot of hype about raw materials as an alternative asset class over the last seven or eight years. Promoters of this idea point to the returns on the GSCI index from 1970 onwards and the low correlation of commodity returns with those on other assets. But it turns out that a large part of the overall commodity index return came in the 1970s, and a major component was due to roll yield – rolling from one commodity futures contract to the next while the market was in backwardation. Now we have the opposite situation, where an investor faces contango and a cost in rolling forward. The component of a commodity index return that came from investing in treasury bills has also now largely disappeared. Finally, the low correlation argument took a hit last year when commodities fell with everything else.

If the last year has demonstrated anything, it’s the importance of a dynamic asset allocation process, shifting funds to asset classes where the risk/reward trade-off is best, based upon forward-looking valuations rather than historical returns.

This approach helped GMO’s asset allocation group to highlight negative expected returns from US equities in 1999 and, conversely, very high expected returns from emerging market equities in the same period, at a time when people were still highly skeptical of this asset class following the Asia crisis and Russian default.

More recently, during the credit bubble period, we observed that the price of all risk assets – equities, high yield, emerging market debt, even commodities – was being bid up, and Ben Inker, who heads the asset allocation team, produced a chart in 2007 showing that the more risk you took on, the less return you were getting. Conventional asset allocation models at that time were showing that you should increase your exposure to private equity and hedge funds, for example. Because Ben used forward-looking returns, he was able to show that the efficient frontier – the line matching risk and return, which should theoretically be upward-sloping – had become flat, and in fact cash had the highest risk-adjusted return.

IU.eu: We’re marking the second anniversary of the start of the credit crunch – it was in summer 2007 that two Bear Stearns hedge funds that were invested in sub-prime debt failed. How far do you think we’ve gone in unwinding the excesses of the bubble?


Chancellor: Looking at things from the perspective of valuation, we decided in February and March this year to add to equity positions – in high-quality US stocks, and some international and emerging market shares – as we calculated expected returns of above 10% from all three areas at that time. Clearly equities are less attractive now that the S&P 500 index has risen from under 700 to over 1000.

Our equity asset allocation model, which is based upon the concept of mean reversion of profit margins and the mean reversion of valuations over time, suggests a fair value for the S&P 500 of around 850. So the market is now looking slightly expensive, albeit nowhere near the degree of overvaluation we’ve seen over the past fifteen years. However, given the pretty miserable economic outlook, one could argue that we should be trading below fair value rather than above it.

Since one should expect markets to trade below the mean long-term value for as long as they’re trading above it, the fact that the stock market only fell below its historical mean valuation for a blink of an eye is slightly disturbing. If you’re value-driven, you’d have liked to see the market stay cheap for a while. We’ve certainly had a panic, but not so far a “proper” grinding bear market.

This is a matter of concern for us, not because we enjoy bear markets as the result of a sort of schadenfreude, but because they represent a good time to invest.

Having said all that, the prospective returns on equities are not too bad compared to where they were a decade ago. On bonds, by contrast, the prospective returns look pretty miserable.

With government bond yields below 4%, you may still make some money if we enter into a deflationary period and yields drop further, but if inflation stays near the recent 2-2.5% average you’ll make very little, and there’s a huge tail risk to your investment in bonds if higher inflation re-emerges.

IU.eu: What about commodities? Do your earlier comments mean that you wouldn’t use them as a long-term investment?

Chancellor: No, I was referring more to the way in which commodities have historically been sold to investors. There are two things that make this asset class potentially attractive. The first is that it’s a very efficient inflation hedge. The second – and this is a subject that is close to our Chairman, Jeremy Grantham’s heart – is the possibility that we are getting to the stage where the supply of raw materials doesn’t keep up with demand. If one were to buy into a “peak oil” or “peak commodities” thesis, then it might be expected that real commodity prices would rise, rather than the historically flat price trend we’ve seen for oil and gold, and the negative price trend we’ve seen for agricultural commodities.

If you accept this scarcity thesis – and it’s important to remember that there have been prior scares about the world running out of oil or copper, plus you have to assume that technological development will not lead to the more efficient extraction or usage of existing commodities – then a commodities strategy would play a part in a long-term asset allocation strategy.

Of course you would still need to pay attention to the question of roll yield and contango, and to the possible changes in the regulations surrounding commodity investment which are now being discussed.

IU.eu: What about inflation-linked bonds as another asset class that would hedge you against inflation?

Chancellor: They’re no longer as attractively priced as they were during the deleveraging scare last autumn, but on our models US inflation-linked bonds offered a prospective return of 1.6% at the end of June, so they represent a bedrock, low-volatility asset for a portfolio. Although some people say that real yields on inflation-linked bonds may rise during an inflationary period, we see this more as an inflation risk premium in a fixed income bond, and if inflation were to re-emerge in the next ten years or so I doubt we’d see a large rise in real yields. In summary, inflation-linked bonds are slightly expensive, but priced in line with the long-term real returns on bonds.

Original post