This article is authored by MOI Global instructor David Sather, president of Sather Financial Group, based in Victoria, Texas.
Late in November, I received a call from my nephew, Ryan. He graduates in May with a finance degree from a respected mega-university.
He was neck-deep in a “financial modeling” class and asked if I’d review his semester project. It was a spreadsheet more than 750 lines deep. It was full of macros, pivot tables and limitless assumptions.
As I perused my nephew’s work, I remembered a well-known value investing quote: “If your answer takes until line 350 of a spreadsheet, you are making it too hard.”
Although financial models can be very useful, they can also be quite deceptive. Often, the longer the spreadsheet, the more variables which must be properly assessed. A small error rate compounds geometrically over time resulting in false precision.
This can also be true with discounted cash flow analysis. It is very seductive to move the growth rate up a tick or the discount rate down to justify a given position.
With DCF analysis, we have struggled to find an intellectually honest position given what has happened with interest rates. When the 10 Year US Treasury was closer to 6%, we routinely used a 12% discount rate in our initial ten-year growth phase. With the 10 Year US Treasury at 1.9%, we will often use a discount rate of 9% to 10%, if a business is deemed to be highly consistent. Even then, you must use tremendous discretion and caution as there is a world of difference in running DCF analysis on a company like Tesla verses a boring, yet highly predictable company, like Brown-Forman.
We also believe that the quantitative justification of a moat is also coming under attack. Twenty or thirty years ago, it was easy to look at Proctor & Gamble, Gannett, Gillette or Kraft and simply model out consistent growth as far as the eye can see. However, rapid-fire innovation and technological revolutions are compressing the span of competitive moats. For this reason, we have also had to scrutinize the logic and application of terminal growth rates in our DCF analysis.
A standard two stage DCF might use a 4% terminal growth rate for years 11 through 20. I think I can justify a five or six percent terminal rate given the stability of Brown-Forman. Although it doesn’t sound like much, increasing the terminal growth rate to 6% increases the current fair value on Brown-Forman by 15%.
Although I might get lucky with a higher terminal growth rate on Brown-Forman, the smart investor needs to be intellectually honest when building in a margin of safety. Fancy spreadsheets or cash-flow analysis can be manipulated to say whatever you want. That doesn’t mean that your math is wrong….but the logic and assumptions used to populate your formula’s can be wildly off.
All of this gets back to building in conservative assumptions.
At past Berkshire annual meetings, Warren Buffett and Charlie Munger have been asked why academic programs don’t teach finance and investing the way they do. Munger has tersely replied that if they focused on things that truly matter, universities and textbook companies would not have the volume of material to teach or books to sell.
Financial motivation leads me to believe that many “experts” feel compelled to deliver reams of data and books of analysis to justify fees—regardless of the results. Can you imagine the look on a big client’s face when presented with a $100,000 fee but only one sheet of paper with a concise and effective answer? Rather, volumes of pages, graphs and charts typically accompany the six-figure fee, even if the answer is less effective. Complexity and opaqueness sells consulting services and promotes academic literature.
When I was in graduate school it was common for consultants and academics to develop incredibly complex models. They would triumph that their particular model had a 92% accuracy rate, or some similar claim.
What they later found was that in chasing money or bragging rights the experts added so many variables to a given model it was hard to know which variables truly mattered. As such, efficacy was lost.
A good example of this complexity was in 1998 when Long Term Capital Management imploded. With two Nobel prize winners on their team, they certainly had the smarts. However, their models were incredibly complex with a long line of inter-related variables. Making matters worse, they used tremendous amounts of leverage.
LTCM had over $120 billion of assets when it cratered. However, due to borrowed money its credit exposure was so enormous that the Federal Reserve was forced into action to prevent a broader collapse.
Buffett has often said if you have a business to sell, he can tell you whether he is interested in five minutes. That statement used to confuse me. However, in studying his methods one realizes Buffett has filtered every possible question that could be asked, into the few that truly matter.
Buffett knows that in providing downside protection a simpler model allows you to build a wider margin of safety. This allows him to be “approximately right as opposed to precisely wrong.” Again, the greater the complexity, the more variables. This makes it geometrically harder to build a true safety cushion.
In addition to knowing which variables matter, Buffett puts the odds in his favor by extending time frames. Even if your models and assumptions are 100% accurate, it does not mean the markets agree with you. However, the longer you extend your time frame, the higher the odds are that the market will come to appreciate the wisdom in your analysis.
When I look at our best investments this year, it is easy to puff out my chest and look smart. However, truth be told, those same “best” investments were our worst investments last year. Being patient allowed them to perform as expected.
As you build your own financial models, budgets or projections, know the top five or ten variables or questions that really matter.
Summarize your investment thesis in one page or less and then look in the mirror. Are you making it too hard? Are you being intellectually honest? Are you making the math say what you want it to say? Or, is your math building in a sizable margin of safety that offers a buffer if your assumptions are wrong?
Be careful with leverage. Borrowed money makes us look smart when the market agrees with us. However, it is wickedly painful when the market disagrees. Given some of the rumors swirling around the repo market, it gives me pause to wonder how much leverage is in the system, propping up a variety of hedge funds. Leverage is incredibly cheap right now. However, as Buffett has famously said, “Only when the tide goes out do you know who is swimming naked.”
Give your investments enough time to succeed. The dividing line between bum and hero can be mere days, weeks or months. Why leave it to chance? We are very blunt with our clients. Do not give us any of your money for stock market investments, unless you are committed to a ten-year time frame. Otherwise, we set our clients (and ourselves) up for failure.