Theory meets practice

Ostensibly, prices in these two dance to different beats, but lately their movements have been more like synchronised swimmers and this sort of pattern is being replicated all over the poor man’s portfolio. What’s happening and does this spell the end of conventional portfolio theory – where there is no greater virtue than diversification – as we know it?

I suspect these questions are being asked in all corners of the globe because – to use a headline from The Economist earlier this month – we are dealing with The Bull Market in Everything. If asset prices of all sorts are heading north and at roughly the same pace, then their price changes will – to use statistics jargon – be ‘positively correlated’. That’s okay while prices continue to rise. But the implication is that, when they fall, all will head earthwards at the same pace, rendering diversification useless and making a mockery of portfolio theory.

Reality is messier than theory. Diversification is a fine aim, but more difficult to achieve than theory suggests. Probably it was ever thus even if, as we speak, the price movements in assets of all sorts seem more connected than ever.

First, it’s helpful to grasp why diversification is so important and not just because of the ‘eggs in baskets’ thing. For a given return, the value of a properly diversified portfolio will bounce around less than a badly diversified one and – for those who like it rough – for given volatility, a diversified portfolio will generate better returns than its non-diversified opposite.

Yet these points hold more in theory than in practice, not least because it’s so hard to find a bundle of assets whose value will change independently of each other and in opposite directions. That shouldn’t be too surprising since, in theory and practice, there is a common driver propelling the value of all investible things – the cost of money (or interest rates).

The theoretical connection is that, implicitly or explicitly, interest rates lie behind every valuation model you care to mention. In practice, interest rates always influence the inclination to take risks or, conversely, not to bother. If rates are low – like now – that’s a come-on; if they’re high – and, implicitly, risk-free rates are halfway decent – why bother?

In so far as that holds, you can see that the values of TR Property and Pacific Assets should move in something approximating to harmony. To quantify that, we use the correlation co-efficient, a key tool in regression analysis. Using monthly returns over the past five years, the correlation for these two is 0.48 on a scale going from 1.0 (perfect correlation) to minus 1.0 (perfect negative correlation). That’s pretty typical for London-quoted equities; in comparison, the co-efficient for the Bearbull Income Portfolio versus the FTSE All-Share index is 0.56.

Yet this still leaves us pondering whether or not asset-price movements have become more synchronised than in the past. Rather unscientifically, the table tests this, using a pair of assets – the FTSE 100 and the dollar oil price – whose values could often be expected to move in opposite directions. In five-year batches, the table groups returns for the two over the 25 years from 1992 to 2017.

It shows that, generally, oil was better to hold than the FTSE 100, but only for those who could live with lots of volatility in returns. In every five-year period the volatility of the oil price, as measured by its standard deviation, exceeded that for equities. Meanwhile, the correlation co-efficient (right-hand column) shows that putting the two into the same portfolio was a great thing to do in the 15 years from 1992 to 2007. That has been less since the financial crisis. 

From 2007-12 the Footsie and oil behaved much like any pair to equities. But in the latest five years the correlation has widened, reverting closer to what common sense would predict.

Rightly, this leaves us sceptical of the notion that profound changes have taken place in the relationships of asset prices. Even so, this exercise is helpful to the extent that it makes us focus on the make-up of our portfolios and to be aware of the truth that we are always less diversified than we think, even when we bring all of our assets – listed and unlisted, liquid and illiquid – into the reckoning. As for the application of regression analysis to equities alone, don’t let a positive correlation with other holdings in your portfolio put you off a really good investment proposition.