When I started out as a value investor my original plan was to hold 10 companies. I thought that this would be enough diversification to reduce volatility by a sufficient amount such that I would be able to sleep well enough to stick with the investment program.
At the time my approach involved fairly strict checks on each company’s balance sheet to make sure that a total failure due to leverage was highly unlikely, so the odds of a company going bust, and my losing 10% of the portfolio, were acceptably small.
On that basis holding 10 companies seemed like a good idea as it would allow me to concentrate on those stocks that were at the bleeding-edge of value, the absolute top percentile. There was a flaw in that plan though. In fact, there was more than one.
The value premium is hard to spot, so a big net is better than a small one
Most academic studies into the various value effects, like a low price relative to book value, sales or earnings, generally split the investment universe into quintiles or deciles. The most finely grained results I have seen were split into a 10×10 grid, splitting the market into 100 separate portfolios. Even at that level of granularity the US market (on which the study was based) would have over 50 stocks in each portfolio.
The value premium, when it appears, is generally found across a wide pool of stocks of a particular type rather than by looking for it in individual cases. More importantly, much of the value premium comes from a small number of stocks that are very difficult to spot in advance. Holding low price to book stocks (for example) just increases the chances that you will own those stocks that will outperform, and the outperformance of these few stocks is enough to outweigh the weaker performance of the majority.
Large holdings increase the fear factor
Not only is a highly concentrated portfolio more likely to miss the big winners that help value investors outperform as a group, but the volatility of each holding, since it is a relatively large piece of the portfolio pie, may keep you awake at night or worse, make you sell out at the bottom to avoid the terror of further losses.
If a portfolio has only 10 or 20 positions and 2 or 3 of them drop by 50% (which is actually quite likely), that’s going to be a big test of nerves for all but the most stoic.
More diversification means that each position has less impact on the portfolio as well as the investor’s mental health. This is important because as an investor you have to be able to sit through 50% drops in the value of a holding, assuming you can see no underlying change in the company’s fundamentals.
If volatility shakes you out of a position from which you would otherwise have stayed in, then you are one step closer to becoming a speculator.
Changes to my own investment policy
Investing is a process of constant learning and so based on my continued thinking about diversification I’ll be changing the number of holdings in both my personal portfolio as well as the model portfolio that I follow from the Defensive Value Report. Currently the target is 20 FTSE 350 companies that have on average faster growth, more earnings power yield and a higher dividend yield than the FTSE 100.
Ben Graham suggested 20 to 30 holdings in a portfolio of large and leading companies and as I’m gradually warming to the idea that more diversification is better than less I’ll be changing my target from 20 to 30 companies.
Although some investors call diversification ‘di-worse-ification’, I think that in most cases that’s wrong. Diversification is an essential tool in an investor’s arsenal and is a logical defence against the fact that nobody knows what the future will bring.
I would suggest: diversify your asset classes! As a value investor, you are at the volatile end of the investor spectrum.
Moneyman, that’s an excellent point. It does raise some interesting questions about what diversification is trying to achieve.
If you are diversifying in order to reduce systematic (market) risk (volatility) then diversifying among uncorrelated asset classes is a good idea. It allows the portfolio to gain additional returns and reduce volatility when the portfolio is rebalanced.
But if you’re using diversification to reduce the fundamental risks inherent in the underlying companies then I don’t think asset class diversification is needed. In this case you’re not looking to reduce market induced volatility, you’re simply hedging your bets against the failure of any single company or the long term decline of a particular industry.
So in my case I have two portfolios. One is an index tracking stocks/bonds portfolio which is diversified in terms of the underlying companies (100 companies in the FTSE 100) and it’s also diversified to reduce market volatility by holding equities and government and corporate bond ETFs.
In my other stock picking portfolio I have no intention of holding any bonds or cash or REITs or whatever as I’m not trying to avoid market volatility.
I’d also agree that typically value investors are at the volatile end. I guess it just depends on how affected you are by volatility. If an investor was bothered by volatility then it can be reduced with high yield blue chips which can also be bought as value investments.