Masterspeak 2017

"The more pain a strategy extracts from you, the less likely you are to stick with it"

Tobias Carlisle reveals his deep value investing magic formula 

Published 7 years ago on Sep 01, 2017 16 minutes Read

Los Angeles/ July 11, 2017

There are very few lawyers who end up as full-time investors. Maybe because they have figured out that the grass is indeed greener on their side. Investing is not a sure thing but litigation or liquidation sure is. Tobias Carlisle then is an exception of sorts. He not only moved from Australia where he was a mergers and acquisitions lawyer to the United States, he also has authored multiple books on investing, the most prominent being Deep Value. The strategies underlined in his books also drive his investments at Carbon Beach Asset Management. Besides his books, Carlisle’s work has a wide following through his websites The Acquirer’s Multiple and Greenbackd. Like many others, Carlisle too caught the investing bug after reading Warren Buffett’s Berkshire Hathaway shareholder letters. Carlisle believes those letters are the best business education anybody can get from the greatest investing mind of several generations. As we interview him on a pleasant Tuesday morning at the Playa Vista neighbourhood of Los Angeles, Carlisle like many deep value investors can’t help but smile over the fact that for him the turning point in the market was supposed to be five years ago.

First off, what is magical about the magic formula… 

It definitely beats the market, so that is magical, and its simplicity that leads to market-beating returns is extraordinary. Joel Greenblatt had this idea that Warren Buffett’s investment strategy could be reduced to simple ideas. One of them is value and he uses Ebit on enterprise value to determine that, and the other is profitability or quality and he uses return on invested capital to calculate that. This essentially measures how much money a company makes for each dollar invested in it. He ranks all the companies on valuation and then again on the profitability and quality. The cheapest of the combined ranking leads to market-beating returns. That is an amazing proposition because Buffett is the greatest business analyst alive and probably the greatest of several generations. That the quantitative part of what he does can be reduced to such a simple strategy is magical, I think. 

And what’s the reality? You have come up with a superior version of the magic formula in some sense, and you say that if only you focused only on cheapness, it is good enough for you to beat the market. 

I saw those returns and I looked at the companies that it chose, and I thought that a lot of those companies tended to be companies at the top of their business cycle. So, if you’re looking at something at the pinnacle of the business cycle, it might tend to look profitable, but that is sort of a historical measure. All the energy companies looked very profitable just before they went through the 2008 recession, but their earnings weren’t representative of their typical earning power. I had this idea that if we got rid of that — it might be introducing some sort of error, but if we got rid of that error — then we could generate better results. When you test the proposition, what you find is that just looking at value alone leads to better returns. The question often asked is: what does quality or the return on invested capital bring to the party? I think what it does is at periods of the very top of the business cycle, when there is a sort of mania for stocks, the very profitable companies do tend to keep up with the market. It does help you for short periods of time at the tail-end of manias to do better. But it does not work well over the full period, over a full-period business cycle from peak to trough to peak.

But even when you look at cheapness alone during periods of exaggerated earnings, the stocks will look cheap if you ignored the name and if you ignored the fact that it is a commodities company. Then, on the measures such as earnings yield, it will look extraordinarily cheap, won’t it? 

The difference is that you are not purposefully selecting for these very high profitability companies, so it just removes that error. So now you have a basket of companies, some of which are very profitable and others are not so profitable, and that just tends to balance up the portfolio. What happens is that some of those profitable companies recover quite quickly. The phenomenon that I look for is depressed earnings and depressed valuation, then you look for that return to normalised earnings and the closing of the discount. The two things together give you market-beating returns.

Does this have limited application outside the US? If you look at Brazil or Russia, commodity businesses will be the bulk. This means that the counterforce may just be 20-30% of the market. In this case, will your version of the magic formula really work?

It is something that I was concerned about too. But there has been some research done in the last two or three years. It seems to be a phenomenon that persists in every single market and I think the reasons are pretty simple to understand. Often these companies are depressed because there is some short-term issue with them and all you’re doing is buying them at the time when people are worried about the business. As that shock moves away, the business seems to improve, or the underlying commodity starts doing a little bit better. Basic materials, energy, iron ore are depressed at the moment. I think it’s a pretty good time to buy those sort of companies but that remains to be seen. 

Does the Acquirer’s Multiple work for the same reason?

It relies primarily on mean reversion. You buy a company at its scariest, at its worst point in the business cycle. Every time you look at these companies, you know exactly why they are so cheap. I could give you all the reasons why these particular companies are going out of business, and in many cases the reasons will be right. But in enough cases in a portfolio, the market has simply overreacted. Given two or three years to these companies to recover, they just get back. But because you’re starting from such a depressed valuation, you make good returns. That is the nature of deep value. 

There was a period during Graham’s time or during the Great Depression, when the statistical approach really worked. Then there is a period defined by the rise of Buffett that also coincides with the period when America went through a big boom, quality companies came up, brands thrived. Now, are these winning strategies really defined by the economy at a particular point in time? For example, if you applied Graham’s approach during the 80s and 90s, would it have yielded similar results? 

The problem with cigar-butts is that they are always going to be small companies. It’s the nature of the way it is calculated. There has to be a very significant portion of liquid assets on the balance sheet — cash, inventory or receivables. They have to have fewer liabilities than those liquid assets. That’s cutting out a lot. Big companies don’t qualify often on that measure. Furthermore, of the residue, you’re only supposed to buy two-thirds of that number. Thus, by definition they must be very small companies. That is not future proof, and as the search methods get better, those sort of opportunities have disappeared, unequivocally so. 

The Acquirer’s Multiple doesn’t have that problem because it scales, and we can apply it in different ways. We can apply it on an absolute basis, so we’re not going to pay more than a certain multiple. Or we can just find the cheapest 10% stocks in the market whose valuation make it valuable overtime. But we’re always looking at the cheapest portion of stocks in the market by that metric. So, it should continue to work, although it goes through periods when it doesn’t work as well. The last seven years have been one of those periods in the States. It’s a momentum market at the moment and those periods come around every ten years or so.

But is there any logical reason why this has happened in the last seven years? It’s a fairly large time period to justify a stock not reflecting its true value. 

I think these are new paradigm periods where some new technology comes to the fore and that seems to spur investments. In the boom times it tends to be the more glamorous companies that do better. Tesla is a good example of that at the moment. Even Amazon. Although it is a fantastic business, you pay a lot for it. So people wouldn’t have capital for smaller, undervalued, boring businesses. These boring and undervalued businesses seem to be the ones that will lose to whatever is the flavour of the moment. It’s really only dedicated deep value investors who are looking at those kinds of businesses. Long periods of underperformance come around about once a generation for businesses like these. We had one in the late 1990s — this is the worst one since the late 1990s, there will be more in the future. There is probably a reasonably good period for value coming up though.

While there is proof that the magic formula and several other quantitative parameters work, isn’t it natural that if something becomes popular, its efficacy should diminish over a period of time?  

The reason that value in general and as the subset of the magic formula in the Acquirer’s Multiple will continue to work is because they go through periods of extensive underperformance, and people have a very low threshold for pain when it comes from underperforming. If you are underperforming, and the market does well, you tend to shift into what the market is doing. It is the constant renewal of people coming in and going out that creates the big discount for deep value. When that big discount exists, it does tend to start working again. They call it a pain trait — the more pain a strategy extracts from you, the less likely you are to stick with it and the more likely it is to outperform. 

But it has led to some real pain for value investors. A lot of deep value investors have seen fund flows and redemptions. So how do you see this playing out? 

This happened during the dotcom boom when lots of value investors who had very good long-term reputation, including Buffett, were damned. Like I said, underperformance causes people to leave and join the momentum group, and eventually, that will turn around. I thought it would turn around a little bit sooner, but it hasn’t and I don’t see any signs of it turning around anytime soon. But they say it’s darkest before dawn. Often that’s an indication that there is some sort of change coming. Probably, it would require a crash because in a crash, momentum will be hit the most. That’s because if you are a new investor, you see something that has gone up a lot, and you think that’s what will continue to do well. And when things crash, they’re very quick to leave, they go the other way. On the contrary, people who are buying undervalued stocks are buying on yield basis, buying the assets and the earnings. If you’re buying them cheaply, then you’re probably prepared to hold them until they reach their fair value at least. I think there’s some sort of turn coming along but if you ask me when it would have happened, I would have said, five years ago.

Michael van Biema says that value investors at times don’t seem to realise that if the value realisation or unlocking doesn’t happen, the business itself is undergoing a degeneration. Since you’re a quantitative investor, do you take that into account? 

I have quantitative accounts, and we have a special situations fund as well, which is a totally different approach. Special situations are more Graham-like in the sense that you have to find things that are mispriced in the market. There are lots of different strategies covering a broad range of things that we think about. The biggest one we have at the moment is long Yahoo and short Alibaba, because Yahoo owns a chunk of Alibaba. If you take out Alibaba from the valuation of Yahoo, there is no value to describe the rest of Yahoo. It definitely has some value — $16 billion in cash, some real estate, Yahoo business has Yahoo Japan and those businesses are worth something. For a $50 Yahoo stock, you get another $10-20 in value. We definitely think about the qualitative aspects in those sorts of transactions. Alibaba is very expensive. I think their accounting is opaque. Yahoo is much more transparent, and much cheaper. And if the market goes down, I want to be short Alibaba and long Yahoo. We look at other things too. In the quantitative side, we want to make sure the cash flows are in sync with the accounting earnings and that we are not buying at the top of the cycle. We’re trying to find things where valuation matches the true earning power of the company, which is more of a qualitative kind of estimation.

Is there any optimal portfolio size? When you use quantitative measures, the list of underperforming companies may be pretty huge…

There are different ways of constructing. It’s one of those funny occasions where the academic view based on the efficient market theory, and the value investing view tends to fall around the same number, which is 30. At 30, adding additional stocks doesn’t add much diversification. You can go down to as few as 20. Between 20-30 it moves between 90% of the benefits to 95% and then to capture the last 5%, you need to go to 100. But when people are constructing portfolios they’re trying to achieve different things. If they are trying to maximise the amount of money they can get into it, then more stocks is better. If you’re trying to maximise the performance, then fewer stocks is better. Thirty seems to give the best balance of quantity and performance. If there is fewer than that, it gets very volatile. More than that, your performance is going to get closer to the underlying index. 

Does the expectation of secular growth pose a threat to mean reversion across stocks?

There are a few interesting bits of research that I have seen, one of them is the relationship between stock market return to GDP growth, which is what you expect if the country as a whole is growing very quickly — you would expect that the stock market is growing very quickly too. It turns out that that’s not actually the case. The relationship between stock market return to GDP growth is either non-existent, or slightly negative. The reason seems to be very high growth countries have their stocks bid up very expensively, because there is a lot of optimism about future performance. They get too expensive, and then it’s like a high growth stock that is too expensive. 

So returns are front-ended?

Right. And the opposite is also true of all the slow growing countries. If you see data from 1980s to 2010, China hasn’t had a great stock market performance and China is one of the fastest growing countries in the world. Britain hasn’t had rapid growth in GDP, but it has had a very good stock market performance in that period. So GDP growth and stock market performance aren’t closely correlated. I think it is always valuation that drives return. 

Have you looked at the effectiveness of various financial parameters in driving performance? 

Valuation ratios have done well individually, but financial ratios work best in combination because each one looks at a different aspect of the business of the company. There are plenty of quantitative investors who use stock price momentum, which also work very well. It seems to be predictive. Companies with the higher stock price performance are not the most expensive ones that you might expect; they tend to be from the bottom to the middle who have done very well over the year or so, because they have just recovered. If you buy them, you do see another year or so of outperformance. It just means that the market is identifying something in those businesses that’s not obvious yet in the financial statements. It’s an interesting metric, but it’s not one that we use because we think of ourselves as value investors. I can understand value and see how it moves, but I don’t fully understand momentum. 

Are there any parameters that are particularly effective in judging quality or franchise power? 

I look at the size of the gross margin and how much money the company makes on each product that it sells. There are two ways to examine that. One of them is to see if they have a very high margin, and the other is to see if they have a growing margin. Either of those is the true test of whether you can sell your product for much more than what it costs to make. If you can, that is the sort of thing that encourages competitors to enter, to make it cheaper or sell for more. If you can sustain that, then that is the best measure of the quality and potential profitability of a company. The sort of companies that identify with these are exactly what you would expect: Procter and Gamble, Google and so on. The companies that get eliminated are commodity-based businesses because they have no pricing power, they can’t sustain their margins, neither can they grow their margins overtime. It doesn’t really matter whether the margins are growing or whether they are very high. I think if growing margins is what identifies a future star, then very high margins is identifying something that is more stable and sustained.

How do you view net-nets as a strategy especially in the context of India, where the regulatory ambience is not very conducive for companies to liquidate? Companies just lie low; they don’t go into liquidation, like in the US. 

Net-nets are a terrible reflection on the company. But you are not buying the business, you are buying the asset, the balance-sheet value, you’re hoping for some improvement in the fortunes of the business. You use it as a proxy for value. The returns to net-nets in the US haven’t traditionally been through liquidations, they have been through mean reversion. They just get too cheap. It doesn’t take much for them to get over that barrier but they’re always terrible businesses. They just don’t get that cheap if they’re not. Net-nets are not typically buy and hold sort of investments. This is why I like the Acquirer’s Multiple a little bit better, because you can find companies that have fundamental ideas with a sort of valuation to match. Given five or six years, you can see some very good returns. You have to remember that nets-nets aren’t very great businesses, when they start looking good, they need to be sold.

Is there a framework to look at value embedded in highly leveraged companies? What could be the trigger points to buy them?  

The price-to-book identifies those businesses. When people buy cheap on price-to-book basis, they think they are getting lots of machinery very cheap. But what they’re getting is a very, very leveraged balance sheet, with not very much equity capital, and they are buying a tiny little sliver of that. The way that you would look at whether it could survive or not is to make sure that it has interest coverage ratios, and hope that there is some fundamental improvement in the business. 

The problem with highly leveraged companies is that sometimes, their value is not in the equity but in the debt. If the debt is traded and if you have that mechanism for getting control of the debt, then the value might be realised if you go into liquidation — and for some sort of restructuring. It is difficult to predict where the value will end up flowing. So if you are a shareholder, you could end up valueless, even if the infrastructure project itself is very valuable, the cash flow might falter and you won’t end up with anything.

It’s not something that I do because you can do without unnecessary financial risk. You don’t need to take that risk as an investor when you can find companies that are cash rich and are still generating strong operating earnings relative to what you are paying, even though those operating earnings might not be enough to make them a high return on investment capital business. But then you are almost by definition getting cheap assets, cheap cash flow, cheap operating earnings. If they survive, they tend to start doing a little bit better. I think you can get better safer returns doing that, than by buying high leverage companies.