David Swensen, Yale’s legendary endowment manager, recently delivered a lecture in London to an audience of charity trustees.  Forget stock picking and market timing, he told them:  asset allocation is the only investment decision that adds value.

Swensen justified his conclusion by pointing out that both stock picking and market timing are necessarily zero sum activities:  one investor’s gain is necessarily another’s loss. But this argument proves too much.  All secondary market trade is zero sum:   only investment activity which improves the return on the underlying assets can benefit savers taken as a whole. In aggregate,  alpha is broadly zero.

Swensen’s claim finds stronger support in the research of Ibbotson and Kaplan (Financial Analysts’ Journal, 2000) who analysed the dispersion of investment returns across a sample of asset managers and found that asset allocation accounts for most of the variance.  But this finding is underpinned – as are so many analyses of fund performance – by the prevalence of closet indexation by asset managers who track benchmarks.  Most of the volatility in short term investment performance comes from fickle equity markets, and most institutions hold very similar equity portfolios.   Asset allocation appears to matter because variation in equity proportions accounts for most observed   differentiation of investment  strategy.

         The keys to Swensen’s  success lay in a high equity exposure and  in becoming an early, and large, investor in successful private equity and hedge funds.  But the title of the book he wrote about his approach– Unconventional Success – reveals the limitations of this approach.  The first hedge funds and private equity funds were often run by talented individuals with idiosyncratic styles.  As other investors piled into these ‘alternative’ investments, these came increasingly to resemble mainstream portfolios but with much higher fees.

That evolution reveals a general  problem in emphasising asset allocation. Conventional asset categories are largely arbitrary and change their meaning over time.  Investors have mostly come to recognise that the distinction between domestic and overseas equities lacks relevance. One way of gaining emerging market exposure is to buy Western companies whose brands are prized by China’s rapidly expanding middle class. Shell and Exxon Mobil, like Glaxo and Pfizer, resemble each other more than they resemble other companies whose head offices are located in the same jurisdiction.

But the issue goes wider.  Owning property occupied by a supermarket may not be very different from owning shares in the supermarket, as the recent pressure on the sector has demonstrated.  Yet perhaps there is a difference.  Financial engineering via sale and leaseback deals from large food retailers created investment opportunities  that were classified as ‘property’ but in fact bore more resemblance to indexed bonds.  The proliferation of complex securities made it more necessary to enquire into the exact determinants of the underlying  cash flows on which the security was based; but the global financial crisis showed that many investors had not really understood the need to do this, or did not much care so long as a rating agency gave the answer they sought.

      Diversification is essential to prudent investment but the herding behaviour of modern investors means that in the short term almost every asset is correlated with every other, even if the underlying characteristics of the assets are completely different. Quantitative easing, has boosted all asset prices and  exacerbated this problem.

       The consolation  for charity trustees or others responsible  for endowments is that their time horizons are sufficiently long that they should not have to care.  There are two basic questions.  What is the expected absolute return (over a period long enough for the influence of  underlying cash flows to outweigh the  volatile opinions of market traders)?  And how differentiated are the determinants of these cash flows from those already held in the portfolio? Asset allocation should be the outcome, not the driver, of such investment decisions.