Print this page.


We do not have a typical investment management industry hierarchical structure of portfolio managers supported by a layer of junior analysts that do most of the actual research work, visiting the companies and reporting back to headquarters. Too many things are lost in the handoff in such a structure, in which the portfolio manager never gets his hands dirty. Such structures are frequently seen as necessary in large or even medium sized investment management companies. How else can they try to stay on top of 50-100 companies they seem to think are necessary for proper diversification? It is not the quantity of researchers that lead to good performance; otherwise the large investment managers would have the best numbers. Clearly, that is not the case. There is no correlation between size of an investment organization and results.

Over the last several decades, the portfolio managers at Underhill Investment Management have examined hundreds of companies. We make sure we visit most of the companies we own, many on a regular basis and, of course, we supplement this coverage with phone calls, conference calls and by attending investment conferences and various trade shows. Visiting companies and getting to know their people gives us a better understanding of the company and allows the management of the company to gradually become more comfortable with us. We also develop our own detailed financial models on each holding. Over the years we have built extensive Rolodexes of contacts across various industries as well as long relationships with a number of especially outstanding analysts on Wall Street who we utilize as "sanity checks".

We concentrate our clients' investment monies and do not overdiversify. It has been shown that once a portfolio of common stocks gets much beyond 15 or so companies, without having more than one stock in any industry, adding one position reduces risk from concentration by less than 1%. Focusing on three or four handfuls of stocks allows us to know our companies better. We believe over-diversification is a major cause of mediocre performance.

When we find a truly great company we tend to hold the stock for a very long time, thereby giving management the years to fully exploit its competitive advantages and allowing the magic of compounding to impact the portfolio for several years or more. Some of our holdings we started to buy over 10 years ago.


Our #1 and #2 investment criteria in importance are high barriers to entry and a durable competitive advantage.

Usually you don't get one without the other. A distinct competitive leg up may come in many forms, from proprietary one-of-a-kind products to a superior distribution set up, to low cost producer status, but it is nearly always accompanied by, or leads to, a high barrier to entry and the ideal of most businesses --- limited competition. Such a situation will not last long if the barriers to entry fall and the competitive advantages erode.

Other important criteria include:

      1. Open-ended growing markets.
      2. A leadership brand franchise like Heinz in catsup, Coca Cola in soft drinks or Frito-Lay in snack foods.
      3. Being first into a market, like Federal Express.
      4. A long time proven track record.
      5. Cash flow well in excess of capital spending.
      6. The ability to self-finance growth.
      7. Management that allocates capital well and that owns goodly amounts of the stock, or whose compensation is highly dependent on growing earnings.
      8. Consistently above average returns on invested capital and operating margins.
      9. Pricing power of the company's products.
    10. Little or no debt.
    11. A rising dividend.


Our pricing discipline is based on the standard "margin of safety" concept, which Ben Graham and Warren Buffett, among many other successful investors have called amongst the most important ideas in investing. Margin of safety is our key to pricing discipline in "buying low" and also, in reducing risk. By margin of safety, we mean a cushion or discount should exist between our assessment of fair or intrinsic value and what is paid for a stock.

We believe a major cause of mediocre performance is not just poor stock selection but investing in otherwise terrific companies at high or non-bargain prices.

Gains from holding some very good companies thus become average unless the stocks are bought at those rare times, such as the mid-1970's, 1981-2, 1987, 1990, 1994-5, etc., when great company stocks become a bargain. Anyone knows that PepsiCo, Proctor & Gamble, and Schlumberger are fine companies but the question is when to buy them.

The traditional stock pricing mechanism in our business is to project earnings 5-10 years into the future (good luck!), and then discount those earnings by a chosen discount percentage to arrive at a valuation one is willing to pay for a stock. A terminal value is also assumed. This approach involves too many estimates and too many guesses, in our opinion.

Historically, earnings multiples tend to eventually approximate the return on invested capital (ROIC) of a particular company. The purchase of a high-grade stock at a discount to the ROIC thus tends to greatly reduce the risk of common stock investing via taking advantage of a temporary mispricing. This approach takes historic ROIC and requires an earnings multiple at a discount, the larger the better. Return on invested capital is merely net income after taxes divided by the total of long term debt plus common and preferred equity. This methodology uses primarily historic and audited numbers, not much in the way of guesswork. Such a margin of safety doesn't guarantee success but, we believe, it greatly reduces market risk and enlarges appreciation potential.