Key insights
- The author discusses concentrated portfolios under efficient and inefficient market assumptions. In inefficient markets, concentration can build wealth through exploiting informational edges. In efficient markets, concentration may be viable with firms exposed to minimal firm-specific risk. Overall, the post leans slightly bearish as it highlights the risks associated with concentrated portfolios in efficient markets.
These are my thoughts on concentrated portfolios under different assumptions about market efficiency. I'm curious to know, what do you guys think about concentration? Cheers!
Inefficient Markets: It's a common mantra in value investing circles that "concentration builds wealth" and "diversification keeps it", but under an efficient market, this isn't theoretically true. The assumption, then, is that the market is NOT efficient. It is this inefficiency that value investors hope to exploit, concentrating on positions where they believe themselves to hold an 'edge' (aka, asymmetric information about the position/security that the rest of the market is ignoring).
As a corollary, the investor has a privileged opportunity to know more about this business, because it lies 'in their circle of competence' (which means it is a type of business that they know very well from personal experience), and/or because they have done the sufficient 'scuttlebutt' to know more (effectively investigative research, boots on the ground knowledge). While I assume far less likely, it could also mean the investor has gleamed something from the financial reports that no one else has yet.
If the market is inefficient, value investors stand to gain tremendously by sticking to this strategy. While it requires a tremendous amount of skill and due diligence to outperform the market in this way, historical and anecdotal evidence points to this being certainly possible.
Efficient Markets: Under the aforementioned assumptions, concentration makes perfect sense for someone who knows what they're doing. Does concentration ever make sense, if we instead accept that the market is efficient (or at least mostly efficient)? I propose that yes, it can make sense - under the specific conditions where an investor holds a small set of firms that are exposed to as little firm-specific risk as possible.
In the capital asset pricing model, it is assumed that the marginal investor (those with the significant capital necessary to set prices) is fully diversified. If they are, they are exposed to nearly no firm-specific risk from any of their holdings, which results in them bidding up prices on individual stocks until firm-specific risk is more or less uncompensated among them.
The problem is, it is extremely difficult to find single firms with little to no firm-specific risk. It is easier, though, to find a few companies with comparatively little firm risk (aka, those with no more than 60% firm risk, or an R squared of at least 40% with the market). These are usually large, mature firms with predictable cash flows (likely holding competitive advantages).
If an investor holds as few as 5 of such firms (and they are uncorrelated), their R squared with the market approaches 90%, meaning the concentrated investor should be willing to pay close to the same price for these firms as the completely diversified investor. If this is done successfully, mathematically, an investor will have a total beta relatively close to the market beta of all their holdings. In other words, the concentrated investor is exposed to as little uncompensated risk as possible, compared to the diversified investor.
It would, however, be an exercise in futility, because you would not be expected to benefit whatsoever by concentrating in this way. Alpha, in a completely efficient market, is assumed to be down to luck and only luck. There is also the logical thought that, just because a firm has had little firm risk in the past, that doesn't mean it will continue having little to none of it in the future.