Graham's "Unpopular Large Caps" Part 2: Thoughts on Diversification

Follow up to my post on Graham's still-effective strategy on buying out of favor large caps; Graham's view on diversification and my own thoughts on portfolio management

In the last post, I talked about Graham's categories of investments, and how one category specifically caught my eye: Unpopular Large Caps. It's an investment strategy that worked in Graham's day and it's one that works as well today. Perhaps even better today.

For today's post, I wanted to finish publishing my notes on last weekend's journal entries that stemmed from rereading Graham's thoughts on diversification.

First, it's important to remember that Graham's strategy was built on the foundation of investing in bargain stocks trading below the value of the company's net cash (and quick assets like receivables and inventory that could be turned into cash) after deducting all liabilities, without giving any value to any real estate or other long-term assets. He referred to these as net current asset bargains, or net nets. The idea is you were buying a company for less than the cash you could pull out of it if you shut down operations, with the added bonus of any value from the fixed assets or real estate that you might be able to sell later.

But Graham was not a liquidator, and so he never actually intended to make a claim for this cash. He just felt that buying stocks below this level implied a price that was undervalued — there was excess value well above the stock price to some private owner, and Graham's simple view was that the market would eventually realize this view.

Over time, Graham earned around 20% annually for many decades investing in net nets.

One of the keys to this approach was he viewed it like an insurance operation: as an insurance underwriter, you don't know which policy will lead to a claim, but you do know that if you write many policies with similar odds of success, that you'll profit on that group overall. Like insurance, Ben Graham's net net strategy utilized the law of large numbers, and so diversification ensured that his result would meet the average expected value of the group.

And the expected value for those stocks were quite high.

Net Nets

In the course of researching a number of companies outside the US (mostly Japan), I’ve been thinking a lot about net nets and how they've consistently performed so well.

Graham has a great quote: "It always seemed, and still seems, ridiculously simple to say that if one can acquire a diversified group of common stocks at a price less than the applicable net current assets alone -- after deducting all prior claims, and counting as zero the fixed and other assets -- the results should be quite satisfactory."

It makes sense that a collection of assets that you can buy for less than working capital (in theory the cash you could take home if you closed the business), without even including any value to the real estate or other assets a company owns, that this would work out well.

Margin of Safety and Diversification

Graham says diversification is a central part of margin of safety concept. You need to try and achieve a margin of safety in each investment, but the reality is that each investment, no matter how carefully studied, can go wrong. That is the reality of business: unexpected things occur, advantages can erode, people can make mistakes, etc… So, you need a second layer of diversification beneath the individual stock layer. That is, you need an adequate (though not excessive) layer of diversification.

How much is adequate? Everyone here has a different answer. Charlie Munger said you don't need more than 3 stocks. Buffett has said that if he were running a partnership today with $250 million in assets, he would "probably have 5 or 6 investments and half of the capital in what he liked best". Joel Greenblatt earned 40%+ returns for many years, and said he typically had most of his money in 5 to 8 stocks.

At the other end of the spectrum, Walter Schloss produced incredible results (21% over 50 years, or 16% net to LP’s after Schloss performance allocation). He achieve these incredible results through owning a widely diversified portfolio of deep value stocks (he owned around 100 stocks at at time. Peter Lynch was famous for owning hundreds of stocks (at a very high turnover rate; see The Misunderstanding of Peter Lynch’s Investment Strategy). Ben Graham owned over 100 stocks most of the time in his partnership.

All the investors mentioned above have earned excess returns for decades, proving that there is not just one correct way to manage a portfolio.

Finally, I thought it was interesting that Ben Graham himself had some recommendations on diversification which run counter to his own portfolio practice: in Chapter 4 (The Defensive Investor and Common Stocks), Graham says the following:

"There should be adequate though not excessive diversification. This might mean a minimum of 10 different stocks and a maximum of about 30".

Contrast that with Graham's portfolio of "usually more than 100" different net nets, and his advice seems to be more concentrated. However, his advice of 10-30 consist of stocks of large, well-capitalized companies at low prices, not the net net portfolio. He didn't recommend net nets for the "Defensive" investor.

Two Types of Investments; Differing Levels of Diversification

Graham was very diversified in the net net category because he viewed that group as a collection of assets. He might have even viewed it as one position: a collection trading below the amount of cash and assets that could be quickly turned into cash, without giving any value to the real estate or other assets. Ben Graham viewed this net net group as a separate business line, almost like a merchant retailer would view bargain products that can be resold at a profit, or like an insurance underwriter would view a collection of well priced risks. The more you can find in this group, the better.

I would segment these ideas into two groups: core operating investments and bargain assets. In the former, you want to be very selective in picking a relatively small number of companies you intend to own for the long term. In the latter, you'd want to think like the insurance underwriter, buying as many as you can to ensure that the law of large numbers is on your side.

It's interesting that Graham did not find success in picking stocks in the first category: core operating businesses, and he said he actually abandoned this approach in 1939. This is despite him writing about “unpopular large caps” (see last post), which is a strategy I’ve personally found to be very effective, and perhaps even more effective today than in decades past, given the ever-shortening time frames that investors focus on.

But, while picking operating businesses didn’t work as well for Graham, the net nets remained consistently profitable for him throughout his 3 decades of running the fund, returning an average of more than 20% per year. Side note: there is a false narrative out there that Graham’s fund underperformed had it not been for Geico; this is simply not true; I’ve calculated the returns of his fund pre-Geico and the results are sizably better than the market (and near the 20% number he references for his net net basket’s long term).

I think the net net strategy continues to work today. There are a variety of studies that have tracked the results of various net net portfolios and every one that I've seen have shown outstanding results. The issue is very few (of reasonable quality) exist in the US so I've never been attracted to buying these. But there are net nets in other countries that are of excellent quality with safe balance sheets, a history of profits, excess cash flow for dividends and buybacks, and often a net cash (not just net working capital) that is greater than the current market value of the stock, meaning you pay nothing for the business's earnings or its other hard assets. There are a variety of reasons why these negative enterprise value stocks exist, but I think as a group, they'll likely perform quite well, especially when considering the capital allocation improvements that are beginning to surface in certain markets.

It's possible that a Ben Graham approach could still be effectively applied to situations like this.

Conclusion; My Thoughts on Diversification

Each investor has their own "correct" level of diversification that matches their investment style and perhaps more importantly, their personality. My view on diversification is this: the minimum level of diversification you need is the level that allows you to "zoom out", meaning the level that allows you to behave in a rational manner at all times.

You never want to be overly reactive to short-term results. Being too concentrated might lead you to extrapolate too much into a short-term earnings result, or make a hasty decision based on emotion and not reason. When problems arise (as they inevitably will), you need to be able to look at those problems with a clear and detached mindset, which gives you the best chance of making a rational decision. Emotional decisions are almost always bad decisions and will often lead to poor results.

The reason there is no right or wrong answer on diversification is that each investor will have their own level of stocks that allows them to "zoom out", to behave rationally. A 6 stock portfolio might in fact be the optimal level of stocks, but if this level is below the minimum "zoom out" threshold of the portfolio manager, then it will prove to be suboptimal. It would be better for this manager to have 10 or 12 or whatever level he or she needs in order to remain rational.

In theory, it's easy to say "volatility is not risk" or "I always will act rationally", but I've found that there is a certain level of diversification that allows you to implement this required behavior at all times, and that level varies and depends on one’s personality.

Charlie Munger said he needed just 3 stocks; Warren Buffett suggested 5-6; Greenblatt had fewer than 10. I’ve noticed in the last decade, concentrated investing in high quality companies has become much more “normal”, but I’m not sure most investors using this approach have given their portfolio construction as much thought as I think they should. I also don’t know if these portfolios have truly been stress tested yet. That said, these portfolios certainly could — depending on the industry makeup of the individual stocks — be adequately diversified. I think the behavioral tendencies of the person managing the portfolio are just as critical as the stock selection.

I know of very few deep value investors running a Walter Schloss or Ben Graham style portfolio (though there are a few of them), but I think this also could be very successful, despite seemingly being out of style with current accepted practices among active stock pickers.

I use a simple method that involves opportunity costs to determine which stocks enter my portfolio; the basic goal is to rank each stock against every other stock in the portfolio (and every new prospective stock that I’m looking at). It’s not a perfect or precise method, but I do believe it helps me keep the capital invested in the stocks offering the best prospective returns. See this podcast for more details.

I categorize my stocks into three groups, each which has different levels of required diversification levels: 1) core operating businesses I hope to own for many years (these are few and far between, and often represent my largest positions); 2) out of favor (but very high quality and in my view very safe) large caps (see last post for details: Apple in 2016 at 10 P/E is an example here; sometimes even the largest and most well-followed companies get mispriced); and 3) bargains and/or special situations that are mispriced (occasionally these situations can offer such great risk/reward that they warrant a large position, but this category can also contain a small basket of stocks that I would view as one position; e.g. thrift conversions trading at 70% or less of tangible book.

All three categories play a role in our portfolio, and I view all 3 categories as equally high quality in terms of the returns I expect and safety of principle I demand.

One of the things about investing that has always fascinated me is that there is no one way to win. It's a true liberal art and not a science, even when implementing a more quantitative type approach. Each of these investors mentioned above beat the market by a wide margin over a long period of time. They all found the level of diversification that worked for their strategies and their personalities. That is a critical aspect of portfolio management, and one that probably deserves more thought than it gets.

---

John Huber is the founder of Saber Capital Management, LLC. Saber manages separate accounts for clients and also is the general partner and manager of an investment fund modeled after the original Buffett partnerships.

John can be reached at john@sabercapitalmgt.com.