How We Think

Avoiding the Loser and the “Too Hard” Pile

Short-term quotational changes in the value of a business do not concern us.  We do, however, spend a significant amount of time thinking about how we might suffer permanent capital loss in a holding, as this type of loss is a major obstacle to long-term investment success.  How can we foresee and avoid such occurrences?  We are constantly thinking about what event or confluence thereof will cause the business results of a company to be significantly worse than our expectation.  Bruce Berkowitz summarizes this concept succinctly by asking, “How do we die in an investment?”

For each business in which we invest, we think about the inevitable march of certain trends like the “Wal-Mart effect.”  We define the Wal-Mart effect as the increasing domination of retail by a few large, sophisticated operators like Wal-Mart, Costco, Home Depot, Tesco, etc.  What effect does this phenomenon have on vendors, customers, real estate developers, and competitors?  Many vendors do either the majority of or the plurality of their sales through Wal-Mart.  Do the increased sales volumes generated through this channel offset the lower margins created by Wal-Mart’s purchasing power?  Does a Procter & Gamble have an advantage versus a smaller consumer products company with Wal-Mart, or does Wal-Mart and/or the end consumer take all of the benefits regardless of the vendor size?  Does the Wal-Mart effect aid the vendor’s business in the short run only to damage it in the long run?  This trend is but one of many ideas that must be considered in order to understand the future of a wide variety of businesses.

We have a long list of such trends, many of which are interrelated, that we constantly think about because they impact the competitive dynamics of industries and markets all over the world.  The most important of these trends currently is the Internet (we know this is not exactly a ground-breaking insight).  We suspect the creative destruction caused by the Internet is just in its infancy; what has happened to newspapers is just the tip of the iceberg.  Pricing for all manner of goods and services has become almost completely transparent.  The incorporation of this trend into our thought processes may or may not enable us to predict a winner, but it increases our ability to avoid the losers.

Compared to most investors, we put many more businesses in the “too-hard pile.”  We do this for a couple of reasons.  First, we try to be disciplined in admitting our ignorance about many things as we respect the boundaries of our circle of competence.  Second, we frequently see more uncertainty in the future than the typical investor does.  You cannot be successful as an investor if you uncritically accept the status quo.  We think our focus on recognizing the potential for disruptive change will help us reduce our exposure to permanent capital losses in a variety of businesses and industries.  Some of our ideas about the disruptions emerging in education illustrate how we find uncertainty where conventional wisdom sees stability.

The current lecture/test model used since the Middle Ages was the best method of teaching pre-Internet, but its weaknesses include:

  • It presupposes student homogeneity. Students are expected to have similar prerequisite backgrounds, to have similar aptitudes, and to move through the pertinent subject matter at the same pace.  Testing occurs only on material covered during fairly long intervals of time (six weeks or so in many cases), and regardless of whatever mastery students demonstrate over the tested material, everyone moves forward.  As the teacher attempts to adjust for the variances in the class, the pace inevitably becomes too slow for some students and too fast for others.  The result is a poor fit for the majority of students – certain students lose interest and become bored while others fall further and further behind because they never grasp the building blocks required for the next level of material.
  • The schedule is inflexible and does not match natural learning patterns. Current teaching styles are generally based on a broadcast model that requires a physical classroom, which forces students to congregate at the same time and place.  This requirement makes one hour minimum classes a virtual requirement and, by default, teachers must move through material in a linear fashion.  The end result is a general decrease in comprehension across the board, which can worsen as a class moves forward through a semester.

Computers and the Internet currently allow a different approach than what is employed in the current model:

  • Increase the digestibility and interconnectedness of subjects
    • Break lectures into smaller pieces that match typical attention spans and allow students to move at their own pace through material
    • Cease assigning a low value to students’ time; if a student grasps a particular set of material easily, he should be freed to spend time on another set he is struggling with or should simply be allowed to advance to new material altogether
    • Centralize content so students can easily cross-reference other subjects as needed to promote improved comprehension (e.g., a 15 minute lecture on Ancient Greek History may mention The Iliad, which includes a click-through to an introductory lecture on Homer)
    • Use the scale and the low cost structure of the Internet to allow lectures to include far better visual demonstration, graphics, art, etc. (e.g., take a virtual field trip to a reactor while studying nuclear energy to bring the subject alive)
  • Let effectiveness rather than proximity determine from whom a student learns
    • Have the “best” lecturer in each subject teach the subject matter regardless of his physical location (this opens the world of “teachers” to anyone who can properly convey an idea beyond the traditional boundaries of tenured professors, teaching assistants, or education majors)
    • Test the effectiveness of and subsequently rank various teachers in real time through video viewing statistics and scores on follow-up quizzes (reward the best lecturers in subjects and eliminate the substandard ones)
    • Offer parallel tracks through material depending on the students’ identified learning skills and preferences (visual, auditory, etc.)
    • Create a mechanism whereby a student can ask for peer intervention from virtually anyone on the web (even allow students to rate the effectiveness of peer interveners and to pay these “tutors” through PayPal or another system)
  • Continually assess comprehension to reduce the number of students falling behind
    • Test more frequently for fluency in the covered subject matter to identify and then remedy budding deficiencies
    • Pinpoint when a student is stuck on a particular section or subject matter and concurrently notify a teacher who can intervene one-on-one
  • Mine all of the data that can be captured with the Internet-based system to continually improve the science of education

Not surprisingly, the changes implied by this new and different approach are already happening.  A gentleman named Salman Khan began tutoring his cousin over the phone a few years ago, and he gradually recognized that he could do better job teaching her by using a simultaneous virtual chalkboard.  From there he realized that he and his cousin would not always be able to line up their schedules, so he used a simple and readily accessible software program to pre-record his lectures.  He posted these lectures on YouTube, and now he has thousands of lectures and tests up on  We encourage you to visit this wonderful website.  His model can fundamentally change the current educational system from kindergarten through PhD level programs.  Academic Earth presents another interesting and competing vision.  Its program combines all of the free material from such esteemed institutions as MIT, Stanford and Yale.  You can even get a degree, which begs an important series of questions:

  • What if you are a self-motivated 18-year-old in Harbin, China; Santiago, Chile; or Cape Town, South Africa?  With a netbook and an internet connection you can have access to the best teaching in the world.
  • What if you are a self-motivated and intelligent 18-year-old in Boston, MA?  Would you pay a six-figure premium to get a similar education at Harvard?  Is there a smaller premium you would be willing to pay?  Depending on the ultimate answer to this question, the fixed cost structure at most higher education institutions could doom them.

Bill Gates postulates that when a disruptive technology comes along, the normal course of things is for an upstart to embrace the technology and to replace the entrenched market leader that seemingly has all of the advantages.   If that is so, we suspect Salman Khan’s model will beat the tens of thousands of PhDs he is competing against within education.  The medieval college system could fail quickly like the newspaper industry, or a hybrid model could develop that can persist for many years to come (especially at institutions blessed with a massive endowment or a taxpayer subsidy).  Regardless of the actual outcome, we feel uncomfortable projecting the income statements of many universities in 2020.  We do feel comfortable putting the higher education industry in the “too-hard pile.”  To predict a great many businesses and industries, we must analyze disruptive trends like this one while also anticipating new ones.  We believe that this type of vigilance and unusual thinking will serve to protect your capital.