On Multiple Asset Allocation Methods

From a reader who is a dear friend of mine:

There are obvious many disparate approaches to asset allocation.  Similar to the disparate approaches of any style of investing, each asset allocation approach has its own particular pitfalls.  Some of these you can plan for and perhaps hedge against or at least mitigate the potential negative impact from those pitfalls, while some booby traps spring up out of nowhere.  Risk Parity issues revolve around leverage, negative skew, and potential negative returns from certain levered asset classes.  Long-term strategic asset allocation may suffer from the quality of initial assumptions and typically relies on stable volatility profiles and correlations between asset classes.  And so on.  Every professional investor – let’s take an endowment for instance – diversified its portfolio among several asset classes and styles of management.  But what is interesting to me is that I’m not sure I’ve ever seen an institutional (or even HNW) investor diversify its portfolio among multiple asset allocation approaches.  Theoretically, splitting up a portfolio between 3-5 different AA approaches (strategic, risk-based, tactical with an opportunistic value lens, tactical with a momentum/trend-riding lens, etc.) mitigates the pitfalls of each one.  What are your thoughts here?  I have a few of my own, but I don’t want to muddy your own intellectual waters ahead of time.  🙂

My personal approach to asset allocation is similar to Warren Buffett, or Value Line.  I invest mostly in stocks, and keep a bunch of safe assets for liquidity.  As the market rises, I add to my safe assets.  As the market falls, I buy stocks.  In October of 2002, things were so bad that I depleted my safe assets, an everything was in stocks.

In general, I think most complex asset allocation strategies are overly complex.  In general, there are safe and risky assets.  Asset allocation should first focus on the division between the two.  Typically the safe assets are high quality bonds and cash equivalents.  Sometimes there are more opportunities, sometimes fewer.  Safe asset levels should reflect that.

The second focus of asset allocation should be liquidity needs.  Even if there are a lot of promising opportunities to deploy cash, if the liability that funds the assets needs cash, have cash ready for it.  If you invest in limited partnerships or private companies where the assets are locked up for a period of time, have a sense of what your maximum level of illiquidity is (what will you with certainty never need to tap?), and ladder the investments so that like a laddered bond portfolio, you always have some illiquid investments maturing each year, providing fresh cash for deployment where current opportunities are most promising.  These top two ideas are very basic, but even experts neglect them at times.

The third focus of asset allocation is choice of risk assets, which is how I view your question.  There my view of asset allocation is like that of GMO.  Forecast future returns off of free cash flow yields; invest accordingly.

Don’t pay much attention to volatility, but aim for what is most likely, and bend a little in the direction of what can go wrong.  Most of the time, over longer periods of time, what is most likely happens on average; that’s why it is most likely.

Maybe “Too many cooks spoil the broth.”  I have enough trouble trying to work with momentum versus mean reversion.  I would lean toward having one AA strategy that fits with my broader asset management practices.  But on the other hand…

Suppose we did have five asset allocation models, and what their results were encouraging various investors to do.  If we thought that one of the models had been too hot of late, and was attracting too much money, and distorting ordinary market relationships, maybe that could give us a signal to make sure our asset allocation de-emphasized the results of that method.  Timing of course would be difficult, it always is, but seeing the results of the five methods could provide a fuller view of choices faced by our competitors.

I’m not sure that using the average of a number of asset allocation models will provide the best result, but I think that understanding what other players in the market are doing could lead to better decisions.

I’m open to your thoughts, and the thoughts of other readers here.  Anyone have a better idea?