Behavioural finance guru James Montier has been considering how we allocate our assets.
One of the biggest questions facing investors, and anyone planning a pension, is how to allocate their funds between the various asset classes such as shares, bonds, property and cash.
It's an important question because, as Tudor Davies shows in a recent series of articles, that decision determines the future volatility and return on your portfolio. This point was driven home over the past two years for anyone too heavily invested in the 'wrong' asset mix.
But according to a new paper from James Montier, many of us make some fundamental mistakes when considering asset allocation, and the principal mistake is ignoring valuation.
Returns, or ease of measurement?
One of Montier's pet hates is the way in which investors appear happy to sacrifice return for ease of measurement; he sees benchmarking against an index as an accident of history, and says that if we were starting afresh today, we would probably do it differently.
Focusing on relative performance and tracking errors distracts us from the goal of absolute performance.
Problems with policy portfolios
Many investment plans follow a defined split between equities and bonds, with the intention of getting a reasonable return without the high volatility of a purely share-based portfolio. A typical split is 60% shares and 40% bonds, and often this mix will shift more towards bonds as the investor approaches retirement -- the so-called 'glide path' approach.
Montier sees several problems with this:
- Risk is not volatility. Using volatility as a proxy for risk makes it feasible to quantify and optimise, and satisfies our desire to measure performance, but these are not the same thing: "Risk clearly isn't a number. It is a multifaceted concept, and it would be foolhardy to try to reduce it to a single figure."
- Indifference to valuation: The strategic allocation model requires us to maintain a defined exposure to particular assets regardless of their valuations. Periodic re-balancing may result in the culling of some over-valued assets, but essentially we would still have a weighting of (say) 60% equities whether the market is on a multiple of 45x or 10x.
- Benchmarking alters behaviour. "Everything becomes relative (risk, return, and valuation) in a benchmarked world", but as he points out, few of us get to retire on relative performance.
- Not enough return: While American pension funds assume 8% total return for a 60/40 fund, Montier predicts 4.5%.
What are the options?
Many funds have broadened the range of assets into which they invest, following the lead set by Yale. These assets include private equity and hedge funds, although many would not regard these as asset classes, as such. Also, the returns on these categories turned out to be more correlated than was expected, so we did not have true diversification.
And once again, the diversification was 'returns chasing', as people try to do today what they should have done yesterday.
One characteristic of the 60/40 split is that more than 90% of the portfolio's volatility comes from equity portion, so another approach is to shift the asset mix dramatically towards bonds until the volatilities are equally balanced. This shift comes at the expense of returns, so these 'risk-parity' portfolios compensate for this by adding leverage, i.e. by borrowing money to make a small return appear larger.
You can probably predict the sort of difficulty Montier has with this strategy; it again conflates risk and volatility, and it increases the financing risk. "The fact that pension funds are looking to use leverage so soon after the most recent crisis should surely send shivers down one's spine. Leverage is a dangerous beast. It can't ever turn a bad investment good, but it can turn a good investment bad. Simply piling leverage onto an investment with a small return doesn't transform it into a good idea."
Another solution
Montier urges a break from benchmarking, and suggests that fund managers should be chosen on the basis of process rather than past returns. This obviously requires both trust and understanding on the part of the investor.
He also advocates what is often termed 'tactical asset allocation' (although that's not how he refers to it), so that asset classes are weighted based on value criteria, rather than on preordained weightings derived from historical data. "Changing the asset mix of your asset allocation in response to the fluctuating opportunity set offered by Mr. Market seems like common sense to me. Sadly, of course, common sense tends to count for little in the world of high finance!"
More from Padraig O'Hannelly:
