Robert Kirby, a professional manager whose heyday was in the 1970s and 1980s, once wrote an article titled “The Coffee Can Portfolio” which focused on the importance of letting winners run as the most important yet chronically neglected component of finding investment success.

When Kirby presented the concept, he told the story about a lady whose money he managed. Specifically, after her husband died, she transferred the assets from her deceased husband to Kirby. 

When Kirby reviewed the husband’s portfolio, he noticed that he had tracked the same exact investments that were made in the wife’s account, except there was one exception: the husband never sold any of the stocks when Kirby did–he just let them all run and achieve whatever compounding it was that they generated. 

Upon comparing the results, Kirby had two observations: (1) the portfolio was dramatically skewed, with just four or five stocks making up the bulk of the holdings, and (2) the husband’s portfolio, which was identical to Kirby’s portfolio on the buy side but never featured any sales, outperformed Kirby.

As part of his self-reflection on why a portfolio that mimicked Kirby’s buys outperformed a portfolio that included his sales as well, Kirby wrote as follows: 

“The investment counsel business, as it is traditionally practiced, and probably as it should be practiced, is a simple process of making sure that clients never have so much risk exposure that their capital or standard of living can be impaired by some specific negative surprise. In other words, as your most successful investments grow in value, you make partial sales and transfer the capital involved to your less successful investments that have gotten cheaper. The process results in a stream of capital being transferred from the most dynamic companies, which usually appear somewhat overvalued, to the least dynamic companies, which usually appear somewhat undervalued.”

When you look at the best-performing stocks of the past twenty years, such as Apple, Amazon, Alphabet, Facebook, Domino’s Pizza, Fair Isaac, and United Rentals, the price usually appears overvalued in the range of 30-50x earnings. The good news is that, with the exception of Amazon, each of those stocks traded below 20x earnings at some point in the past ten years so an opportunity to purchase them using conventional valuation metrics has existed. But there is also needs to be a recognition that the fastest growing companies often have a tendency to appear somewhat overvalued. 

In Kirby’s case, he sold off shares of Hewlett-Packard, Xerox, Exxon (which seems to always show up in these historical reviews), and IBM whereas the husband kept those holdings in his estate. Kirby observed that the husband’s $20,000 investment into HP, Xerox, Exxon, and IBM at a rate of approximately $5,000 each grew into $1.3 million over a twenty-something year period and accounted for the bulk of the husband’s estate. 

As a hypothetical, Kirby surmised that investors should buy great companies and “stick them in the proverbial coffee can and bury them” as the investor’s neglect of a portfolio of great businesses would lead to the greatest amount of gains.

My own observation is that, for most people who have put together a net worth that seems disproportionate to their career ownings, there are a few outsized investments that make up most of the net worth. The Pareto principle applies. Not everything you buy will compound along nicely at a 8% to 12% rate. You will own assets that go nowhere for a decade and you will own assets that compound at 20%. 

The key is to tolerate the former and maintain the benefits of the latter. If you own twenty different stocks that are currently in the S&P 500, the odds are fair that one of them will increase 100x over the next thirty years. That’s the advantage of being a buy and holder–you get to reap all the benefits from the investments that you maintain (that other people would have sold) but also tolerating the subpar compounding and bankruptcies of a few holdings (that maybe others would have sold). But you end up coming out ahead because the magnitude of the gains from the former dwarf the losses from the latter. 

Even Warren Buffett, the most famous investor of the age, isn’t immune from this. Charlie Munger once said the bulk of Berkshire’s fortune is the result of a dozen decisions. No doubt it’s true. The decision to buy National Indemnity, See’s Candies, the Washington Post, Coca-Cola, Reynolds Tobacco, Gillette, Gen Re, GEICO, Burlington Northern Santa Fe, and Apple is largely responsible for creating Berkshire Hathaway’s record of 18% compounding for over half a century. A billionaire who has made hundreds (thousands?) of investments over a career owes his billionaire status to under a dozen decisions. 

Most money managers sell their stocks after a stock doubles. Heck, some of them do it after a stock rises 30% or 50%. Could you imagine if Warren Buffett sold Coca-Cola after it rose in value from $1 billion to $1.3 billion instead of holding on to watch it rise about $20 billion and pay out over $7 billion in dividends? Or See’s Candies, which was purchased for just under $30 million. Imagine if Berkshire discarded it upon receiving a bid for $45 million. A few years ago, Warren Buffett estimated that See’s Candies provided him with $1 billion in dividends over the course of a lifetime. 

To benefit from compounding requires that you step back and allow compounding to occur. Remarkable careers are not made by selling stocks that gain 30-50% and then selling them and trying to repeat the process over and over. The wealth gets made by identifying a few of the best businesses in the world, adding them to your portfolio one at a time based upon which ones trade at the most attractive valuation at a given time, and then getting out of the way to allow them to grow in value. 

Like this general content? Join The Conservative Income Investor on Patreon for discussion of specific stocks!