Lone wolves may secure the best returns

Some animals instinctively gather as a herd, apparently pension funds are such animals. A new asset allocation study by academics at Maastricht and Yale, presented at the ICPM discussion forum last week, reveals the mob behaviour by funds when it comes to asset allocation, leaving way for security selection to be the differentiator in returns.Of course herding can be an advantageous action for some animals, such as in the face of a predator, or in some animals as an attacking mechanism (apparently whales do it in co-ordinating feeding activities).

But in the context of asset allocation is there any benefit for pension funds to act as a pack, in either defence or offence?

The behavioural finance work of academics, such as Yale’s Robert Shiller shows that individual behaviour that is rational, can produce group behaviour that is irrational.

Now this new research, conducted by Aleksandar Andonov and Rob Bauer from Maastricht, and Martijn Cremers from Yale – under the title of “Can large pension funds beat the market?” – reveals that security selection, and not asset allocation policy, accounts for most of the differences in return between funds.

This is not to say that asset allocation does not still produce the majority of the return of an individual fund (as has been demonstrated by Ibbotson and others) but that between funds it is security selection that differentiates them.

In this analysis, which looked at 774 defined benefits in the US and Canada using the CEM database, the performance of pension funds was decomposed into asset allocation, market timing and security selection components.

Crudely, it found that security selection has a far greater explanatory power: 45-55 per cent in the US, and 48-58 per cent in Canada. Asset allocation decisions explain only 35-41 per cent of the return differences in the US and even less in Canada (24 to 34 per cent), with the balance attributed to market timing.

Pleasingly, the research found that pension funds on average are able to beat the market or their own benchmarks, but that interestingly the larger positive alpha resulted from security selection (45 basis points annual alpha) than the timing of asset allocation decisions (21 basis points).

While a collaborative industry for the most part, pension funds ultimately compete, and increasingly so. While pension funds vie with their contemporaries for access to assets, if the trend of large funds insourcing the management of private assets continues, they will increasingly be competing with traditional asset management firms for clients.

Which highlights the question of differentiation. In this context the ability to act outside the herd may be an advantage.


The academic analysis in this paper also goes on to explore the role of asset size, liquidity and costs for performance, as well as the outcome of decision to use active versus passive, and internal versus external management. A more indepth analysis of the study, and an interview with Aleksandar Andonov will appear in next week’s