Truisms are, by definition, things that we either know or believe to be true. In life there are many things we take at face value as being true without seeking out evidence to affirm (or contradict) those beliefs. So let’s consider one of greatest truisms in life (and investing): “Do not put all your eggs in one basket”. Could it be the case that this old adage is not quite as water-tight as we might think? Surely that cast-iron notion that diversification is a good thing dare not be challenged?! Well, actually, I think it should at the very least be revisited!
There are many conventions and truisms that are adhered to in business, which, often with the benefit of hindsight, are not all what they appear. One could cast an eye back to many conventions and assumptions that we understood to hold true pre-financial crisis. For example, “Too big to fail” post Lehmans brothers collapse and perhaps the notion of a “guarantee”, which I think most folks will now agree does not really exist.
Let’s consider how vast swathes of money is managed in our industry, in whatever country you reside. Multi-asset or balanced funds abound in most jurisdictions and are a work horse, read cash-cow, for many an asset manager and life company. Generally speaking, such funds have exposure to equities, bonds, property, cash and alternative assets. Equities will probably account for 50-70% of the fund with the balance split out across the other asset classes. These funds are usually managed within ranges so as to have some set criteria as to maxima and minima in each asset class.
Let’s take it, for the sake of argument, that these funds by investing across these asset classes cover all of the investable universe. Here’s my issue: It simply cannot be the case that all asset classes in the known world are attractive at the same time. Think about it – these funds own all of the investable universe, as we’ve defined it, at all times. Of course it will be argued by those running such funds that they make active asset allocation calls on each asset class depending on their view of the world. They will therefore take a view and adjust their allocation to an underweight position if they have a negative view on an asset class. In my experience, this active call is usually mere tweaking. The asset class might be reduced by at the very maximum 5% at any one time.
Again, in my experience, this large a move would be rare and usually the reduction in a holding is very gradual and happens over time. Also, the view expressed on the asset class will usually not be negative or positive per se but some wishy-washy middle ground where the manager or institution can sit on the fence and not have too much explaining to do and will keep their job if the call is wrong. I understand that investing is a nuanced and a shades-of-grey type world but fundamentally if you have a negative view on an asset class then don’t own it!
One of the basic tenets of these funds is that by investing in assets classes which are uncorrelated one gets an optimized portfolio which will deliver better risk adjusted returns than the alternative. Many smarter than me have argued that the benefit of using historic correlation data in attempt to build an optimized portfolio is questionable. Firstly, there is not a particularly high degree of certainty with regard to those correlations being consistent into the future. Indeed, there have been periods in the past when those correlations have broken down. Secondly, when one most needs diversification and non-correlated assets classes – in a crisis – those very correlations tend to increase thus negating their benefit. (An old and bad joke: The only thing that goes up in a crisis is correlation!). And thirdly, fans of chaos theory will know that altering, even slightly, any single input into a stochastic system can lead to vastly different outcomes at the other end. That is to say that if your assumptions/inputs have a margin for error, the predictability of outcomes will be exponentially worse.
Let’s take an even more extreme view and suggest that a portfolio should consist only of one asset class. The obvious choice for most investors with a long-term horizon would be equities. If you had thrown out notions of diversification, across assets classes at least, and simply held equities over the long run, you’d have done pretty well thank you very much. Of course returns is only part of the argument I accept and many investors don’t have quite the long term horizons that textbooks might assume.
So back in the real world what am I actually suggesting? I am suggesting that investors should only hold asset classes which they believe to be reasonably valued and that the macro-economic backdrop is at least somewhat favourable to that asset class in general. If neither of these conditions exist then don’t own the asset class. If you have a negative view on an asset class and your reasoning therefor is sound, then your chances of being right are probably better than 50% and your opportunity cost if you’re wrong is likely quite low anyway. Imagine a not-so-hypothetical scenario as follows: Bonds have had a multi-year (even decades-long) bull run; bonds are at historically low nominal and real yields; interest rates are zero and there is an expectation, read certainty, that interest rates will go up. Absent any other information, one would reasonably avoid bonds in this scenario and those of longer duration like the plague. If you’re wrong, yields can’t compress a whole lot more, the actual yield/income isn’t so great and so you’ll not lose out too much.
Finally, in a nutshell, always question old adages. To coin a new one: Just because you say it, doesn’t make it so.
Author: Peter Murphy, Managing Director, P. M. Murphy Ltd. Regulated by the Central Bank of Ireland.http://www.pmmurphy.com/
Disclaimer: The content of this article should not be construed as financial advice. You should consult a financial adviser to obtain advice appropriate to your individual circumstances.