One of the best illustrations of the many ways a business valuation can go wrong is the pre-bankruptcy story of mobile satellite service Iridium. I knew Iridium well since I did potential acquisition due diligence on Iridium several times and because people who worked for Craig McCaw had a long history with Motorola. We (Eagle River) always passed on buying Iridium no matter how low the price dropped. The potential upside of buying Iridium even for $1 was tiny, but the potential downside (especially the opportunity cost) was massive. We already had enough wholesale transfer pricing and other problems with Motorola related to Nextel.
There is a recent book about Iridium, but that account does not identify why such colossal mistakes were made in assessing the potential value of the business. The common narrative is:
“John Bloom, an investigative journalist and author of the book Eccentric Orbits: The Iridium Story. “It did what it was supposed to do. It was an engineering miracle. It’s just that not enough people needed a phone at all longitudes and latitudes.”
That’s true only in that not enough people valued the product at the price point. But there are deeper explanations of why this happened. The common explanation is that “cellular spread faster than people imagined.” That wasn’t the root cause.
The root cause of Iridium’s valuation failure was what I call goal seek bias which is a special case of confirmation bias and incentive caused bias. The goal seek function in a spreadsheet allows a modeler to use the desired result of a formula to find the possible input value necessary to achieve that result. To understand how the biases played out in the case of Iridium you must know a few background facts.
Motorola was a satellite subcontractor. It wanted to be a prime contractor, but it did not have a customer. The best way around that problem was to create the customer for Motorola satellites from scratch. Iridium was born for that reason. In order to raise the billions of dollars needed to pay for the system a credible business model was required. Naturally spreadsheets were created and it was necessary to goal seek a total addressable market (TAM) to financially justify building and operating the system to investors. The outcome of that financial modeling was a case of what Warren Buffett calls the institutional imperative at work: “Any business craving, however foolish, will be quickly supported by detailed rate-of-return and strategic studies.”
The result of the goal seek bias in the case of Iridium was preposterous on its face, if the assumptions were carefully examined.
The phone did not work indoors
The phones did not work in a car
The phones did not work without line of sight to the satellite (buildings and even trees are a problem)
The phones were very big, heavy and expensive
The service was expensive
What did the market researchers ask potential consumers about the Iridium service? A classic leading question of course: “Would you value a mobile phone that you could use anywhere?” Who wouldn’t say yes to that question? But it wasn’t even close to the right question to ask.
The reality was “anywhere” meant: to use the Iridium service a person would need to find an open field and lug out a heavy and expensive phone and pay expensive rates. That isn’t anywhere. Anywhere means inside buildings and cars or in the shadow of a buildings. On a mountainside is a place but is not enough to mean anywhere and instead is a niche market. My friend and wireless expert Tim Farrar sent this Tweet on the topic:
Goal seek bias ignored this reality and that resulted in a number of silly estimates from forecasters about the size of the mobile satellite market. For example:
The way confirmation bias works inside a company like Motorola meant that the problems with the addressable market I just discussed were rendered invisible and the momentum enabled by social proof (“Hey lots of big companies are behind this’”) enabled the system to be built.
“In order to secure bank loans in early 1999, the following subscriber targets (“covenants”) were set for Iridium: 52,000 by March 31, 1999; 213,000 by June 30, 1999 and 454,000 by September 30, 1999. Even though never publicly disclosed, the break-even point for Iridium was expected to be between 500,000 and 600,000. In order to reach this level, Iridium would have had to add roughly 50,000 subscribers per month in 1999. In reality, in March 1999, Iridium only had 10,000 subscribers. The number of three to six million subscribers expected by 2001 was never reached. A comparison between the predicted subscriptions and reality is shown in Figure 12”
The end result was both tragic and relatively swift:
On November 1, 1998, after launching a $180MM advertising campaign and an opening ceremony where VP Al Gore made the first phone call using Iridium, the company launched its satellite phone service, charging $3000 for a hand-set and $3 – $8 per minute for calls. The results were devastating. By April 1999, the company had only ten thousand subscribers. Facing negligible revenues and a debt interest of $40MM per month, the company came under tremendous pressure. In April, two days before Iridium was to announce quarterly results, CEO Staiano quit, citing a disagreement with the board over strategy. John Richardson, an experienced insider, immediately replaced Staiano as interim CEO, but the die was cast. In June 1999, Iridium fired 15% of its staff, including several managers who had been involved in designing the company’s marketing strategy. By August, Iridium’s subscriber base had grown to only 20,000 customers, significantly less than the 52,000 necessary to meet loan covenants. Two days after defaulting on $1.5 Billion in loans – on Friday, August 13, 1999 – Iridium filed Chapter 11 bankruptcy protection.
Iridium was purchased by an investment group for very little cash and today is both an operating business and a listed company. The business is very different than originally envisioned (r.g., they are chasing non voice markets) and the link to Motorola that would have created huge wholesale transfer pricing problems is gone (Iridium has multiple suppliers). The prospects of that new business is not a subject covered in this post since this is about what happened pre-bankruptcy.
What are “the dozen lessons” in the case of Iridium?
- “An unresolved contradiction exists: to perform present value analysis, you must predict the future, yet the future is not reliably predictable.” “A perfect business in terms of the simplicity of valuation would be an annuity; an annuity generates an annual stream of cash that either remains constant or grows at a steady rate every year. Real businesses, even the best ones, are unfortunately not annuities.” Seth Klarman
When valuing an asset it is wise to focus of the simplest system possible, preferably one that has a stable and well known operating history (e.g., an individual business like See’s Candies). Iridium was about as far from an annuity as was possible. When valuing a business like Iridium a massive margin of safety must be built into the model to account for risk (risk meaning the potential for a permanent loss of capital). How are risky investment intelligently made? Well, very carefully. A portfolio approach is used in venture capital since only a very small number of outsize winners determine the financial outcome of a given fund. Warren Buffett describes the strategy adopted:
“If significant risk exists in a single transaction, overall risk should be reduced by making that purchase one of many mutually-independent commitments. Thus, you may consciously purchase a risky investment – one that indeed has a significant possibility of causing loss or injury – if you believe that your gain, weighted for probabilities, considerably exceeds your loss, comparably weighted, and if you can commit to a number of similar, but unrelated opportunities. Most venture capitalists employ this strategy. Should you choose to pursue this course, you should adopt the outlook of the casino that owns a roulette wheel, which will want to see lots of action because it is favored by probabilities, but will refuse to accept a single, huge bet.”
“When future cash flows are reasonably predictable and an appropriate discount rate can be chosen, NPV analysis is one of the most accurate and precise methods of valuation. Unfortunately, future cash flows are usually uncertain, often highly so. Moreover, the choice of a discount rate can be somewhat arbitrary. These factors together typically make present-value analysis an imprecise and difficult task. Although some businesses are more stable than others and therefore more predictable, estimating future cash flow for a business is usually a guessing game. A recurring theme in this book [Margin of Safety] is that the future is not predictable, except within fairly wide boundaries.” Seth Klarman.
Klarman is pointing out a fact that will be made repeatedly in this post. A NPV calculation of the value of a business is not precise. This means that having a margin of safety is wise. Everyone makes errors and mistakes and so having insurance against those mistakes is wise. With a margin of safety you can be somewhat wrong and still make a profit. And when you are right you will make even more profit than you thought. I have written posts before about the discount rate issue.
- “Using precise numbers is, in fact, foolish: working with a range of possibilities is the better approach.” “Charlie and I admit that we feel confident in estimating intrinsic value for only a portion of traded equities and then only when we employ a range of values, rather than some pseudo-precise figure.” Warren Buffett “Warren talks about these discounted cash flows. I’ve never seen him do one.” Charlie Munger.
Warren Buffett’s response when Charlie said this was to say: “It’s true. “If the value of a company doesn’t just scream out at you, it’s too close.” “You just want to estimate a company’s cash flows over time, discount them back and buy for less than that.” Warren Buffett “In the Theory of Investment Value, written over 50 years ago, John Burr Williams set forth the equation for value, which we condense here: The value of any stock, bond or business today is determined by the cash inflows and outflows – discounted at an appropriate interest rate – that can be expected to occur during the remaining life of the asset.” Warren Buffett in the Berkshire 2000 annual report.
“In 44 years of Wall Street experience and study I have never seen dependable calculations made about common stock values, or related investment policies, that went beyond simple arithmetic or the most elementary algebra. Whenever calculus is brought in, or higher algebra, you could take it as a warning signal that the operator was trying to substitute theory for experience.” Ben Graham
Some people are seduced by Greek letters in valuation formulas. Seth Klarman has said: “It is easy to confuse the capability to make precise forecasts with the ability to make accurate ones.”
“DCF is an unruly valuation tool for young companies. This is not because it is a bad theoretical framework; it is because we don’t have accurate inputs. Garbage in, garbage out.” Bill Gurley
Whenever you receive a spreadsheet valuing a business, it is wise to focus on the assumptions. And as Gurley points out that is particularly true of a young company. But people make emotional mistakes:
“Ed Staiano, formerly a senior executive at Motorola and the CEO of Iridium at the time of commercial activation, knew in intimate detail how the Iridium system actually functioned and was well aware of its various technical limitations, including the fact that the Iridium telephone would not work dependably indoors or in the urban canyons of central business districts, but he made the decision, nonetheless, to invest $500,000 of his own money in Iridium securities in March 1999”
“Multiple witnesses indicated that they were familiar with the “line-of-sight” nature of the system, and, despite that, they appear to have assumed that the service still would be acceptable to users.”
Iridium is a particularly extreme example of garbage in and garbage out. Sometimes people suspend disbelief:
NEW YORK-Iridium World Communications Ltd., which plans to launch a satellite-based global telephone and paging service, went public June 10, raising $240 million in an offering of 12 million shares of Class A common stock at $20 per share. The initial public offering, lead managed by Merrill Lynch & Co., New York, was oversubscribed. Consequently, the issuer, which had planned a 10 million share IPO, increased its initial offering by 2 million shares.
“Buffett’s reluctance to invest in technology businesses “is not a statement that technology stocks are unanalyzable.” Robert Hagstrom in his book The Essential Buffett: Timeless Principles for the New Economy.
Buffett says: “We have no religious belief that we will not invest in tech, just can’t find one where we think we know what the bush will look like in ten years or how many birds will come out of the bush.” Both Munger and Buffett have said that if they were young today they would acquire a technology circle of competence. But they are not young today. What they are looking for is an unfair advantage when they invest and in high-technology they have no such advantage.
“Some of the worst business decisions I’ve seen came with detailed analysis. The higher math was false precision. They do that in business schools, because they’ve got to do something.” Charlie Munger.
That a spreadsheet generates precise numbers can create an illusion of precision. It is one of several biases that can impact a valuation. Aswath Damodaran:
Multiple bias and illusions were involved in this result:
“The Global Arrangers for a $1 billion credit facility that closed in December 1997 engaged Coopers & Lybrand (now known as PriceWaterhouseCoopers) and Arthur D. Little to conduct extensive independent diligence with respect to the Iridium business plan as a condition to extending credit to Iridium. These consultants to the lenders “stress tested” the Iridium projections, prepared a more conservative set of projections that assumed a downside case for future company performance after commercial activation and gave a green light to the loan which was oversubscribed.”
- “I have seen so many cases where there is a complex model that is exactly wrong. This focus on a model may cause you to move away from thinking about the competitive advantages of the business. Then you are making decisions based on all these numbers rather than thinking about whether this is one of the ten businesses that you would like to own.” Glenn Greenberg.
Numbers can seduce even the most rational investor. It is qualitative factors that generate quantitative results in a business. Does the business have a sustainable competitive advantage? This can be tested for existence using numbers but it cannot be analyzed by just using numbers. Competitive advantage is determined qualitatively. Mohnish Pabrai make the point in this way: “It’s not about the numbers. For most investments the factors that will drive long term success don’t have much to do with spreadsheets. They have to do with something other, either understanding human nature or understanding nuances about how certain aspects of how things work rather than running spreadsheets.”
“So, if one can’t use DCF how should one think about valuation? Well, one solution that I have long favoured is the use of reverse engineered DCFs. Instead of trying to estimate the growth ten years into the future, this method takes the current share price and backs out what is currently implied. The resulting implied growth estimate can then be assessed either by an analyst or by comparing the estimate with an empirical distribution of the growth rates that have been achieved over time, such as the one shown below. This allows one to assess how likely or otherwise the implied growth rate actually is.” James Montier.
Take company X and its market cap. What sort of growth would be required to support that valuation? For Iridium:
|May 1998:||Full satellite constellation in orbit, stock peaks at $68 7/8 with Iridium’s total market capitalization near $10 billion|
Paraphrasing Buffett: “Think about a company with a market cap of $10 billion. To justify paying this price, you would have to earn $1 billion every year until perpetuity, assuming a 10% discount rate. Think about how many businesses today earn $1 billion, or $.75 billion, or $.5 billion. It would require a rather extraordinary change in profitability to justify that price.”
- “Some investors swear off the DCF model because of its myriad assumptions. Yet they readily embrace an approach that packs all of those same assumptions, without any transparency, into a single number: the multiple. Multiples are not valuation; they represent shorthand for the valuation process. Like most forms of shorthand, multiples come with blind spots and biases that few investors take the time and care to understand.” Michael Mauboussin.
“Relative valuation is much more likely to reﬂect market perceptions and moods than discounted cash ﬂow valuation. This can be an advantage when it is important that the price reﬂect these perceptions as is the case when
the objective is to sell a security at that price today (as in the case of an IPO)
investing on “momentum” based strategies
With relative valuation, there will always be a signiﬁcant proportion of securities that are under valued and over valued. Since portfolio managers are judged based upon how they perform on a relative basis (to the market and other money managers), relative valuation is more tailored to their needs. Relative valuation generally requires less information than discounted cash ﬂow valuation (especially when multiples are used as screens)”
From January 1998 to January 1999 a great Number of analysts believed that Iridium had an equity value that ranged between $4 and $14 billion. Garbage in and garbage out.
July 1997: DLJ gave Iridium a $6.2 billion private market equity valuation.
October 1997: BancAmerica Robertson Stephens gave Iridium a $9.3 billion private market equity value.
February 1998: Goldman Sachs found $10.6 billion private market equity value.
“Though many DCF models do incorporate sensitivity analysis (typically a grid of values driven by alternative cost of capital, growth, or terminal valuation assumptions), these grids provide little relevant information for anyone trying to understand the prospects of the business. Investors should look to the value drivers—sales, margins, and investment needs—as sources of variant perception. Even sensitivity analysis based on the value drivers is generally flawed because it fails to consider the interactivity between value drivers. Proper scenario analysis considers how changes in sales, costs, and investments lead to varying value driver outcome.” Michael Mauboussin
Sensitivity analysis is useful in getting a “feel” for a business model. What inputs are most important? Even with a sensitivity analysis since the systems involved are complex and adaptive, scenario analysis is important. Maboussin writes: “Scenario analysis also addresses concerns about an uncertain future. By considering “if, then” scenarios and insisting on a proper discount to expected value—or margin of safety—an investor can safely and thoughtfully weigh various outcomes.
“Generally, when companies or investors run a cash flow model they go out five or ten years. Why is that? Because that’s how many fingers you have. Literally, we live in a decimal world because that is how many fingers we have! What you really want to do is link the competitive position of your business and/or industry to an economically sound competitive advantage period.” Michael Mauboussin
It is always best to read Mauboussin in the original.
“I agree with Warren to keep it simple and not use higher mathematics in your analysis.” Walter Schloss
“In my life there are not many questions I can’t properly deal with using my $40 adding machine and dog-eared compound interest table.” Charlie Munger
“A person infatuated with measurements, who has his head stuck in the sand of the balance sheet, is not likely to succeed.” Peter Lynch
“Any attempt to value businesses with precision will yield values that are precisely inaccurate.”
“We never sit down, run the numbers out and discount them back to net present value … The decision should be obvious.” Charlie Munger
“It is better to be approximately right, than precisely wrong.” Warren Buffett
“There’s no such thing as precise intrinsic value.” Mohnish Pabrai
“There are so many factors involved that it is never wise to attempt to judge intrinsic value to the last eighth or even point.” Phil Fisher
“Given that the future is inherently uncertain, we do not believe the value of any business can be known with certainty at a given point in time, so our aim is to be generally right as opposed to precisely wrong.” Wally Weitz
“It is important to understand that intrinsic value is not an exact figure, but a range that is based on your assumptions” Jean-Marie Eveillard
“Businesses, unlike debt instruments, do not have contractual cash flows. As a result, they cannot be precisely valued as bonds” James Montier
“The essential point is that security analysis does not seek to determine exactly what is the intrinsic value of a given security. It need only establish that the value is adequate.” Ben Graham
“If modest changes in assumptions cause a substantial change in NPV, investors would be prudent to exercise caution in employing this method of valuation.” Seth Klarman