Some investors labor under the delusion that the realm of asset allocation and its influences don't apply to their portfolios. At a recent conference, for instance, I found myself in a casual conversation with a fairly wealthy individual who told me that asset allocation was to be avoided at all costs. He talked a good game, but he didn't realize that he could no more escape decisions on asset allocation than he could walk the earth and avoid gravity.
Consider an investor who picks stocks, one at a time, looking for so-called beaten-down shares that appear to be trading at a discount to book value or some other accounting metric. Let's also assume that the investor invests exclusively in this strategy and holds no other assets. To some eyes, it appears that asset allocation is null and void for this portfolio. But first impressions can be deceiving, and that's true here.
In fact, holding a portfolio of hand-picked value stocks, each selected with great care, is still an asset allocation choice. An extreme one, but an asset allocation choice nonetheless. Although this investor may think that this strategy has nothing to do with asset allocation, the truth is that he's making a decision to emphasize--in the extreme--a particular equity beta with a specific alpha overlay. Embedded in that choice is a decision to avoid everything else. Assuming that the first decision is the only choice of relevance, and that the second decision has no influence on end results, is to ignore reality.
Every portfolio strategy is driven by what it holds, and what it avoids. Imagine a world with just two assets, A and B. Let's say that asset A earned 5% per year over a 10-year period while asset B's price remained unchanged over that span. The benchmark for this world, a 50/50 mix of the two assets, earned 2.5%. But if you held one or the other asset in isolation, your return would be 5% or 0%, depending on which asset you picked. Here's the key point, and one that's too often ignored: your return in holding one or the other asset would have changed considerably relative to the 50/50 benchmark because of two factors, not one. That is, earning 5% in this example is a function of choosing to hold the higher-performing asset and sidestepping asset B and its zero return. There's no other way to earn 5% except as a matter of making two decisions, not one.
The negative-selection factor (NSF) isn't always so stark in terms of outcomes. For a broadly diversified portfolio, NSF's influence may be virtually nil. If your strategy holds all but one of the major asset classes, for instance, NSF isn't likely to be a major influence on risk and return. By contrast, if you hold just one asset class, the NSF influence is potentially huge. Many investors don' think in these terms, focusing instead on what they own and ignoring what they don't. But depending on how a portfolio is structured, the assets avoided may be a substantial influence on the strategy's results. In some cases, what you don't own may be more important than what you own in terms of the end results.
Asset allocation, in short, applies to every portfolio, no matter how extreme. The implication: you should have compelling reasons for holding a given set of asset classes, but the rule is no less crucial when it comes to avoiding asset classes. Short sellers understand this subtle but crucial point, but it's usually overlooked otherwise.
Recognized or not, negative influences matter in money management, sometimes a lot. Combine that with the fact that few investors have strong views (if any) about what they don't own and it's clear that this combination can be quite powerful. But in a world where NSF can be a critical driver of results, ignoring this aspect of risk and return can have large and unintended consequences, and not necessarily in your favor.
No wonder, then, that holding all the major asset classes as a strategy tends to perform competitively against a broad set of actively managed asset allocation strategies. Yes, it's difficult to develop and maintain views on each of the major asset classes, much less their subsets. Even if you managed to stay up-to-date on each asset class, it's still hard to make reliable, high-quality forecasts on a regular basis across such a diverse realm.
Given all this, it's fair to say that a broadly defined set of asset classes is (still) likely to generate average-to-above-average results through time compared with all the world's attempts at doing better. That doesn't sound like much... until you consider the likely outcome from most investors who try to second guess Mr. Market's asset allocation.