Byron McKeeby has long been a friend to this blog. Here he questions the over reliance on asset allocation to pay collective pensions, when the chief consideration is the short -term volatility of the asset classes. What follows is Byron’s argument displayed on comments, you can see other comments of Byron on this subject on this link
I did note this in PLSA’s published response to the LGPS consolidation consultation:
“It is widely recognised that setting the right asset allocation is the most important factor in driving long-term investment returns for a scheme.”
PLSA’s scribes therefore still seem to be peddling the flawed Brinson research from the 1970s and 1980s in the mid 2020s.
Unfortunately, both the research conclusions and the interpretation of those conclusions are wrong.
The problem with Brinson’s analysis was its focus on explaining short-term portfolio volatility rather than total portfolio returns.
LGPS surely should be more concerned with the range of likely outcomes over their longer term investment planning horizon than the volatility of short-term returns.
Their pools will be contemplating LDI and buying gilts next?
Asset allocation policy explains only a small fraction of ten-year returns, but a large fraction of the variation of short-term returns.
Persistent small increments to periodic returns compound over time, while the volatility in returns grows more slowly (and often reduces) as the investment period is lengthened.
When analysing returns for short periods of time, it is easy to miss the significance of small persistent increments to returns in all the noise of marking to market.
To give more detail, several issues have been identified with the Brinson research, but those who trumpet that asset allocation trumps everything have probably never looked into these:
1. Empirical flaws
The original studies by Brinson et al. were based on a limited dataset of US mutual funds and pension plans, and did not have actual data on their strategic asset allocations.
They assumed that the 10-year mean average holding of each asset class was sufficient to approximate the appropriate normal holding, which is a significant limitation.
2. Methodological Flaws:
The studies did not account for the detailed mandates given to portfolio managers, which often include constraints that limit the potential impact of security selection.
This means the studies may have underestimated the importance of security selection in particular
3. General Applicability:
The conclusion that strategic asset allocation dominates other factors has been widely accepted by investment consultants (and the PLSA’s scribes) without sufficient empirical validation across different contexts and time periods.
Drawing general conclusions from such limited research is methodologically unsound.
4. Practical Implications:
The emphasis on strategic asset allocation has led to a more rigid approach in portfolio management, where the potential benefits of active management and security selection are often overlooked or constrained by the terms of market-relative mandates.
(Much so-called “active” management, however, may be closer to “passive” because of the use of market benchmarks and a fixation upon short-term volatility.)
Brinson’s research oversimplifies the complexities of portfolio management.
A more balanced approach, recognising the importance of both asset allocation and security selection, as well as market timing and transaction costs, is preferable.
Sources: such as Journal of Financial Planning: William W. Jahnke’s February 1997 piece, “The Asset Allocation Hoax.”