Updated: May 19, 2022
As is the case with many things in the financial world, there is no one clear definition of the term ‘strategic asset allocation.’ While this may not have been the case when the term was first used in or around the 1970s, in the decades since its meaning has become increasingly ambiguous.
"While bond bull and bear markets average 30 years, the average for equities is 6 years"
The term ‘Strategic Asset Allocation’ (SAA) was essentially a product of Modern Portfolio Theory and the idea that a client’s asset allocation with respect to equities, bonds and cash should be determined by real long-term expected returns from each, as well as volatilities and correlations. Since long-term expected returns would by definition not change, the SAA for a particular client would not change (volatilities and correlations measured over the longer terms are also very stable).
‘Tactical asset allocation’ (TAA) quickly became associated with SAA and referred to the process by which actual asset class weights would deviate from the SAA to take account of expectations that returns from one or more asset classes would in the shorter term be different to the aforementioned long-term expected returns. In practice, TAA deviations from SAA were never more than a few percentage points.
Thus the value added from TAA has been generally either low (if the manager ’s views were correct) or negative (if they weren’t).
I like the use of the terms SAA and TAA but I have a very different view of what they mean and how they should be applied with respect to management of investors’ portfolios. For starters, real long-term expected returns from asset classes should not be considered stable and indeed can vary substantially from decade to decade - or even from generation to generation.
Take bonds. Using the US as an example – it has the best data history – the table below shows annualised real returns from long bonds over certain periods. The point that I would make is that it would be foolish to have a high strategic allocation to bonds during multi-decade periods in which they produce substantially and consistently negative real returns.
Interestingly, equities in the US haven’t exhibited the same long-term cycles as bonds. While bond bull and bear markets average 30 years, the average for equities is 6 years. The reason for this is that bonds follow inflation cycles which are long while equities are more synchronised with business cycles which are much shorter. In other words, equity market bull and bear markets can occur within long-term bond bear and bull markets. For example, the last two equity bear markets (2000-2002 and 2007-2009) both occurred within the bond bull market that began in the early 80 s. Conversely, the two equity bull markets of the 1970s happened during a period of poor real bond returns . Go back further and you’ll find other examples.
To demonstrate how these findings can be used to build a more effective SAA (i.e. one that is not static but at the same time one that does not change frequently) I apply some very simple rules as follows:
• SAA starting point is 50% equities/50% bonds
• Move the equities allocation to 75% if the real equities index falls to 40% below its all-time high
• Move the equities allocation to 100% if the real equities index falls to 75% below its all-time high (very rare!)
• Move the equities allocation back to 50% five years after previous increase
• Move the bonds allocation to 0% if the real bonds index falls to 60% above its 30 year moving average
• Move the bonds allocation back to 50% if the real bonds index rises to 40% below its 30 year moving average
• The cash weighting is the residual of the above changes and can not be negative.
The impact these rules would have had on the SAA can be seen in the chart below.
The main point to note is that very few changes are made. There are 14 triggers to change the equity allocation, which equates to one every ten or so years. As for bonds, there are even fewer triggers to reduce or increase its weighting: four to be precise, or once e very 34 or so years (it should be noted that the bond allocation may get reduced because equities are increased, but such reductions are not ‘active’ ones). In total, therefore, there are 18 active decisions to change the SAA over the 135 or so years under consideration, equating to one every 7.5 years.
The big question of course is what impact these changes would have had on the performance of the rules-based SAA in relation to the static 50/50 SAA.
The answer is that they would have added 1 percentage point per annum (remember, this relates only to SAA and thus takes no account of value that can be added from TAA or security selection). Furthermore, there were only two months (out of 1260) when the rolling 30 year annualised return was lower for the rules-based SAA than for the static version. These findings can be seen in the chart below .
While the value added varied between 0% and 3% per annum, it is important to note that it was never negative (other than in the case of the aforementioned two months which indeed were only very slightly negative.)
To be clear, the rules-based SAA outlined above is for demonstration purposes only . The message I am seeking to communicate is that fixed SAA weights are a bad idea and that with a small number of simple rules, a substantial amount of value can be added. Furthermore, those SAA frameworks that are more flexible with respect to weighting changes tend to reduce equities at the wrong time. This is because optimisation models incorporate volatility as an input. When equity markets fall, volatility generally rises and as a result recommended SAA weights get reduced.
This is in stark contrast to the rules-based SAA outlined above in which when equity markets fall, the SAA recommended weight is increased not decreased. In fact, the rules-based SAA never seeks to anticipate equity bear markets, only to respond sensibly when they happen by raising targets. Indeed this appro ach is endorsed by investment adviser William Bengen in his 1994 paper , Determining Withdrawal Rates Using Historical Data . In it, he wrote:
“Admittedly, increasing stock allocation to 100 percent after a long period of miserable returns requires unusual foresight and fortitude on the part of the advisor, as well as the client. If you can convince your client just to maintain the 75-percent allocation under such conditions, you have won a major battle. However, the client is still faced with a shorter than-average portfolio longevity, and with much less wealth to pass on to heirs than originally hoped for.”
Bengen’s point was that although severe bear markets will always damage portfolios, the damage will be even worse if you do not take advantage of them, or, heaven forbid, reduce positions during them.
Published in Investment Letter, September 2015
The views expressed in this communication are those of Peter Elston at the time of writing and are subject to change without notice. They do not constitute investment advice and whilst all reasonable efforts have been used to ensure the accuracy of the information contained in this communication, the reliability, completeness or accuracy of the content cannot be guaranteed. This communication provides information for professional use only and should not be relied upon by retail investors as the sole basis for investment.