The Berkshire Special + India's Fastest Growing Companies

"Never put any moat on a pedestal"

Pat Dorsey, Founder, Dorsey Asset Management

|
Published 5 years ago on Jun 07, 2019 15 minutes Read

Pat Dorsey is the guy who is at least partly responsible for democratising the use of Warren Buffett’s philosophy of moat in evaluating businesses. How, you may ask. He was instrumental in developing Morningstar’s economic moat ratings, as well as the methodology behind its framework for analysing competitive advantage. From 2000 to 2011, Dorsey was Director of Equity Research for Morningstar, where he led the growth of Morningstar’s equity research team from 20 to 90 analysts. Having authored two books — The Five Rules for Successful Stock Investing and The Little Book that Builds Wealth, Dorsey now manages money as the founder of Dorsey Asset Management, based out of Chicago. Who better than him to ask the tough questions around moats?

Where do you think moats are becoming irrelevant or fast receding?

There are a couple of interesting areas to think about. What we are seeing in more developed economies is that the value of consumer brands is perhaps not as strong a moat as it once was. If you think about it, 20 years ago, if I wanted to reach 40 million buyers of breakfast cereal, I would need to buy advertisements on TV, and that was very expensive. So the barrier to entry to building a big consumer brand was very high. Today, I can spend a relatively small amount of money with targeted digital advertising on Google or Facebook and I can reach a very targeted group of consumers at a much lower cost. The ability to make people aware of your product is much cheaper, at least in Europe and the US, given the advent of digital media. 

In the same vein, that breakfast-cereal company would need to have a massive factory, would need to have its own distribution network, all of which were very expensive. Today, if I find some consumers with targeted digital advertising that like my product, I can rent capacity at a factory and I can use fulfilment options from Amazon or others to deliver my product. A lot of former fixed costs have become variable costs, which reduces the barrier to entry. I think that the next 20 years is going to be difficult for a lot of big consumer brands. The brands are not going away, but their advantage is diminished relative to what it was, because it is easier for competitors to make consumers aware of their product as well as deliver them. So that scenario where some moats which Buffett might have once called “The Inevitables” are perhaps receding.

American Express is a different but similar example. If I have a credit card, it communicates status and, if I have an American Express card, it communicates a bigger status and because of that, American Express has always charged higher fees than Visa and Mastercard. They argued they could do that because their consumers spend more money, they are wealthier, so if you want to have them, you want to be able to accept American Express. But that’s not the case anymore. American Express brings much less status and value than it used to. And as you saw, couple of years ago, Amex lost the Costco relationship, because they didn’t realise that the world had changed and their moat had shrunk. So you have to recognise when the world is changing. A good business adapts with change and it doesn’t try to fight change.

Could you just list out the various moats in order of how valuable they are to a company?

The four types of moats that I categorise businesses into are intangible assets such as brands, patents or regulatory approvals; scale advantages; network effects; and switching costs. It’s very hard to say that one of those is better than the other, because it really depends on the context that you are in. Network effects can be very valuable, but if you think about, let’s take Uber and Lyft for example, both have strong network effects. As a consumer, I want the service that has the most cars, that is likely to be near me at any point in time, so that I don’t have to wait very long. But they both are essentially commodities because they provide the same service. They take me and get me to a place. And so in the US, when some of the news about Uber’s unpleasant internal culture and Travis Kalanick came out in early 2017, Uber lost almost 15 to 20% market share very quickly, because there is no switching cost. As a consumer, if I don’t like Uber, I go use Lyft. And imagine the same with Visa and Mastercard. Like right now, they have enormous moats, no question. But if tomorrow, Mastercard suffered a massive data breach, and people thought it could not protect their information, they could start using Visa with no problem.

I think it is important to never put any kind of moat on a pedestal. Network effects might be the best kind of moat ever, but as these examples show, the switching cost sometimes can be fairly low. 

You say management is not a moat, it is what they do that becomes a moat. Why do you say so? 

Because that’s not sustainable. I don’t know if that manager is going to be there in three years. Management has an impact on the business. Their decisions can enhance the moat or destroy the moat. The definition of a moat, for me, is something that is structural. It’s inherent to the business. You cannot separate the Coca-Cola brand from Coca-Cola; it is part of the business. Whereas, when a great manager joins the business, I don’t know if he will get hit by a bus. Even great managers make mistakes. It’s not to say managers are irrelevant but an advantage derived from talented management is a more fragile thing, than an advantage derived from something structural that’s inherent to the business. 

If the company is owned and run by the same person, there is a certain predictability or longevity. Would you view it differently then?

No, for a couple of reasons. One, founding a business requires a different skill set than growing a business. The skills that enable someone to grow a business from zero to 100 are very different from the skills you need to take it from 100 to 1,000. There are some entrepreneurs who are able to make that transition, but there are many who are not. Running a bigger business requires more rules and processes. Starting a business is more of a risk-taking, seat of the pants endeavour. 

The point is to just evaluate everything on its own terms, and not put any attribute on a pedestal. Somebody who owns 30% of a business and has done well may decide — I have grown the business to a certain point. Now I want to enjoy the good life and I want to pay myself a lot of dividend. But if that business still has a lot of market share to take and a lot of reinvestment runway, that may not be the best way to maximise the long-term value of the business. Taking dividends and building a big house may be the best personal decision for the owner but sometimes those things are not aligned. They often are, but not always.

Capital allocation decisions are perhaps the most important driver of shareholder return and those are driven singularly by the management. In which case, is management not the single biggest factor in delivering return? 

Capital allocation is essentially the link between business value and shareholder value. If capital allocation is poor, then the shareholder does not get the benefit of growth in business value. There is leakage in between the two, because the company raises equity to pay dividend or makes poor acquisitions or whatever it might be. Conversely, you can have a case where the value of the business is only growing by x, but shareholder value is growing by 1.2x, because the company is buying back shares, or making very intelligent underpriced acquisitions. The key to capital allocation is, first and foremost, to ensure that it’s not leaking or destroying value, so that you as a shareholder are going to get at least the incremental return of the business. Superior capital allocation is pretty rare. 

How do you value companies that have moats but don’t really make money?

The same way you evaluate any other business, which is trying to think about the present value of future cash flows. This is an area where the world has changed pretty significantly over the past couple of decades because, 30 years ago, most investments were done via the balance sheet. They were investments in buildings, in factories, in railroads, in locomotives and all those came out of the balance sheet. Today, a lot of investment happens out of the income statement. If you are a software company, and you are acquiring new customers, who might have a nine to 10-year lifespan with the business, that comes out of sales and marketing, and so that depresses your current margins.

But it seems insensible to me to argue that I should not invest in a customer who could be with me for 10 years and who will pay me 3% more every year as I raise prices. Why is that not just as valuable an investment as a machine that will wear out in 10 years? One is an appreciating asset and the other is a depreciating asset. The former — the customer — comes by way of investing through the income statement and depresses current margins. As for buying the machine, it is just a capital expenditure. If you have a business that is re-investing heavily today, a software company or an Amazon for example, you have to think about the incremental unit economics. How much does it cost to acquire each customer and how much value do they deliver over some span of time, and then try to think about what does this business look like at steady state, say in a five or 10-year timeframe. You know what margins it will have once the investment slows down and then you discount those cash flows back to the present.

Is there an ideal time frame you look at while evaluating a firm from a profit perspective? 

The purpose of business is not to generate a profit today, but generate future value and, if a management team overemphasizes making a profit today, which will get valued by the market versus investing for the future that will generate more value, they are making the wrong decision. Just because a business “has never made money” does not make it a bad business. Amazon did not turn a profit until seven years ago. Was it a bad business? A near trillion market cap would disagree with that. 

What you have to ask yourself is ‘why is the business not making any money?’ Is it not making any money because the product is not successful or because it cannot raise prices or because it is taking all of its potential profit and reinvesting them back into the business for future growth. Those are two very different scenarios. You have business A, which builds a steel mill, because it thinks demand will be very high and its costs will be low. But its costs are higher than expected and demand is not high and it loses money because it made a poor business decision. Then, there is business B, which has a huge market opportunity, people are buying its products very rapidly and it is reinvesting in additional capacity. Those are two very different situations. 

How would you value something such as Uber? 

If you are a US investor, you have to ask yourself, if the market is likely to remain a duopoly, with Uber and Lyft. And, if it is likely to be a duopoly, is there likely to be rational pricing? Sometimes, with a duopoly, you have very rational pricing. Visa and Mastercard do not compete on price. They stay out of each other’s way and they make a huge profit. But another duopoly, Boeing and Airbus, they fight pretty hard on price. They are very competitive with each other. And so, does Uber-Lyft look more like Visa-Mastercard or does it look more like Boeing-Airbus? That’s a really important question that you have to answer as a potential investor in Uber. Then, it comes down to unit economics. How much does it cost for a driver to drive a car for Uber and how much value does Uber extract? There is a potentially valid argument, that once Uber or Lyft reduce their level of subsidies for drivers, then capacity will slow. There will be fewer cars available because the drivers don’t find it economically attractive to drive an Uber. If there’s less capacity available, that might lead to a price equilibrium. 

So far, Uber and Lyft have competed very heavily on price. That was evident in both of their IPO filings, they have been trying to undercut each other on price, which is not the sign of a healthy competitive dynamic that’s going to result in great return for shareholders. Maybe that will change, I don’t know. But, when I see two big companies trying to basically undercut each other on price and, it’s not really clear who is going to win, I’d rather just stay on the sidelines and watch. One of the most important things for an investor to do is to maximise return on time. By analysing Uber and Lyft, we probably aren’t going to get a lot of advantage, because everybody and their mother is trying to have an opinion on these things, and it’s just not clear how the competitive dynamics will pan out long term. So we’ve spent literally zero time on them!

Is there a framework to think about what kind of duopolies will compete with each other and what will co-exist? Or is it just led by management behaviour? 

A lot of it comes down to the unit economics of the business. Boeing and Airbus need to absorb a lot of fixed costs. Building an aircraft factory, investing and designing a new aircraft, requires a lot of very high fixed costs, and so they need to absorb that. And so, each incremental plane sold is very important to both companies. So they need to take market share from each other. Whereas for Visa and Mastercard, their fixed cost for the payment networks, those costs were sunk decades ago. Their network is there. It exists. So there’s no incentive to compete on price, because they don’t have the same economics of cost absorption.

Is there a time frame over which you evaluate value creation? Even over a medium time-frame of five years, there can be a disconnect between business value and shareholder value depending on the market conditions…

You begin by looking at the track record. Has the company made overpriced acquisitions? Has the company issued equity and paid dividend in the same year, which to my mind is just unconscionable? But on a forward basis, you have to ask whether the capital allocation decisions today are likely to deliver value in the future. The real value in analysing a business comes from making a rational judgement about future value before it is created and in the price.

Nintendo is a good example. It is a company we own and at this point the share price is beginning to reflect the popularity of its new device called Switch. But before Switch was launched, people did not think it would succeed. But if you talked to potential consumers, if you looked at the level of interest that it was generating, like on internet message boards and places where videogame consumers hang out, the feedback was this is going to be pretty popular. For us, it was an investment that made sense. So you have to make decisions that kind of get ahead of the market.

How do you view unrelated diversification? India’s largest company Reliance ventured into telecom and retail, but only after 10 years is one seeing a payoff. When cash gets allocated into absolutely unrelated areas where the management has no expertise, how do you view those capital allocation moves?

With great skepticism. You just have to evaluate the venture on its own terms.  The history of companies getting into unrelated areas is generally very poor. There are always exceptions to the rule but it is not a strategy that has usually worked out very well.

How do you view conglomerates? Do they have a moat because of their very structure?

Individual parts may have, you look at each piece of the business individually.Sowith GE, we will say, let’s evaluate the aircraft leasing business, the aircraft engine business, the lighting business and the power turbine business because they are all pretty independent. Arguably, they benefited from a lower cost of capital because of GE Capital. So, for a capital intensive business such as aircraft leasing, being able to source capital more cheaply, by virtue of being inside GE was a benefit, no doubt about it. In the same way at Berkshire, if you think about MidAmerican Energy, they havearguably a lower cost of capital than many utilities do. Many utilities pay high dividend because that is what their shareholders want. MidAmerican does not have to pay a dividend. MidAmerican can source capital cheaply from Berkshire and invest very heavily. It has a more flexible balance sheet than other utilities and that is how MidAmerican benefitsfrom being inside Berkshire. But for their agricultural equipment business or a picture frame maker like Larson-Juhl, do they benefit by being inside Berkshire? May be or maybe not. 

Would you say cheap capital is Berkshire’s moat? Or is it the culture?

Cheap and dependable capital sure helps for capital-intensive businesses such asrailroads and utilities. Buffet has often talked about the advantage of float and that is essentially cheap capital. Berkshire will always be there providing capital, if MidAmerican or BNSF needs it, whereas Mr. Market sometimes is not very happy to provide capital. It is much easier to put your finger on, and value the advantage that cheap capital brings, versus a culture. There is no doubt that the culture is there but, for most businesses, relying on culture as a competitive advantageis easier said than done. It takes decades to build a really good culture.

Can you give us leads on how to evaluate network effect and whether the company is doing the right thing to maintain the network effect?

One good way to maintain the network effect is by subsidising one side of the network. Good examples of this are Adobe and Uber. It’s like the chicken and the egg problem, which comes first? If I’m Adobe and I want people to use my software to create a PDF document, well, no one’s going to buy the software if nobody can read the PDF document. So, you make a reader available for free, you subsidise. In the case of Uber, users are not going to request a car unless there are lots of cars available and it’s convenient for them. So, when you enter a new market, if you’re Uber, you subsidise the drivers to generate lots and lots of capacity. So there are lots of cars on the road, so that the users will, well, use your service. Then you suddenly withdraw the capacity by reducing the subsidy to drivers.

The contrary example to that is Bloomberg. If your position as a business is strong enough, you are integral to the way somebody conducts their business, you are a necessary tool, but you abuse that by raising prices too aggressively, what will happen is, you create pricing umbrella for competition. If you raise prices at 7-8% per year for 10 years, you basically create an umbrella and a huge profit pool for the competitor to show up and say, ‘Hmmm, maybe I can attract some consumers and generate a profit, because the going price for this service is so high, I can afford to do it’. Whereas if, as a business, you raise prices at a more modest pace and deliver consumer surplus, you’re making the product 5% better, but you’re only raising prices by 3%, that’s a lot better than if you’re making the product 0% better and raising the price by 5% per year.

When do you know when a moat is weakening for good? Is that the right time to sell a stock?

When to sell and when a moat is weakening are really two different questions. But I would say, the biggest signal that a moat is weakening is the lack of pricing power. If a business historically had been able to raise prices and is no longer able to raise prices, that generally indicates that its competitive advantage is weakening or disappearing.