Someone emailed me this question:
“How do you take the industry cycle into your consideration? For example, you say CRMT shouldn't trade below receivables per share. I believe it does now and people may be worried that subprime auto loans are in the beginning of down trend (or frothy). I think it may not be a concern considering the company’s track record and in the long-term view, but I'm curious how you incorporate the industry cycle into your investment decision.”
One way to think of this question is to divide up what I think about “a cyclical industry” versus what I think about “a certain point in the industry cycle.” Some businesses are very cyclical, and for this reason tend to be poor long-term stock holdings. A good example would be oilfield service providers. In good times, these businesses can look good. These good times can last for years. During the oil boom, many oilfield service provider stocks looked like they had high returns on capital, had grown earnings per share year after year and were all around good businesses. This impression, however, was contradicted by the very long-term record. In the early 2010s, you could look back and see a good 10-year record at oilfield service providers. Often, if you looked at the compound annual return in the stock going back more like 30 to 40 years (so taking the record back to the 1970s or 1980s), these stocks actually underperformed the market. For that to happen, the record had to be truly awful at the bad points in the cycle.
Oilfield service providers have several of the features you see in cyclical businesses with bad long-term stock returns. One, they are in a long-cycle business. The price of oil fluctuates in a way that sometimes results in not just five to 15-month periods where prices are particularly good – but five to 15-year periods where prices are particularly bad. This is a problem for investors trying to evaluate the stock’s normal earning power in good times. Often, the price-earnings (P/E) ratio will be at its lowest when the stock is actually overpriced and at its highest when the stock is actually underpriced. That is because the “E” used in that P/E ratio is last year’s earnings rather than cyclically normalized earnings. So that is the danger for a value investor. A value investor may think a cyclical stock is cheap when it is actually expensive. The danger for a buy-and-hold investor is different.
A buy-and-hold investor may look at 10 years of results and think he has found an above-average compounder. What he has really found is a business whose compound returns – especially in the way returns actually “compound” – may be average to below average. When you multiply together a series of returns on equity – years with negative results or a series of years with very low growth rates, it will bring the compound annual return in the stock below the median return on equity. At very consistent companies, all types of averages – median, arithmetic mean, geometric mean and harmonic mean – will be close to each other. They will also be close to the compound annual return in the stock. This is not true of extremely variable series of returns on equity. Imagine a company retains all its earnings. It is going to grow book value by about its average return on equity. If the return on equity continues to be stable, the new – higher – book value will produce a similar ratio of earnings to equity. This means if a company retains most of its earnings with a high degree of consistency in its return on equity – that company is going to compound book value and earnings per share at a rate close to its average ROE. The stock price will track ROE fairly reasonably. This is not true at cyclical companies.
The other problem an oilfield service provider has is it is very, very far from the consumer demand that drives the entire industry. The further a business is from the ultimate consumption of a product – the longer and more uncertain the cycle can be. The cycle for a company that builds cruise ships is going to be longer and more dramatic than the cycle for the companies that use those ships. The cycle for jumbo jets is going to be longer than the cycle for airlines. So you have to be careful when looking at companies in industries with long cycles. For example, someone was recently asking me about a cement company in a developing country. It could be a good deal, but the problem is knowing what the capacity situation is in that country in terms of cement plants and what the level of building activity has been. The less you know about a country’s macro cycle and an industry’s cycle, the more careful you have to be.
Frankly, cycles have been the best source of my investment ideas. Warren Buffett (Trades, Portfolio) says you should invest in great businesses with temporary and solvable problems. Well, the most solvable problems are those that solve themselves simply through a self-correcting process in the industry. There are two reasons I was able to buy stocks like Fair Isaac (FICO, Financial) and Omnicom (OMC, Financial). One reason was the stock market crashed in 2008, so everything was cheap in 2009. The other reason is Fair Isaac and Omnicom are companies with stable customer bases and stable competitive positions, but which vary in the level of activity that clients are interested in. In 2009, there was less need for running credit scores. Likewise, there were more subdued “animal spirits” among big brands that advertise a lot. Those are both problems that are temporary and solvable. They are not competitive pressures.
Many years ago, I picked a stock called Posco (PKX, Financial) for a newsletter I wrote. This was probably during 2006. The big concern with Posco was it produced steel. Steel is not a good industry because it is cyclical and has relatively poor product economics. The gross profitability of the business is not strong. It is very easy for competitors – because this is a commodity product – to do harm to even the best producer in the region. Posco has plants that – through transportation of steel – can be in competition with Chinese steelmakers. That worried me a lot. Steel plants are a sunk cost. The economics of producing steel in any region that has an overcapacity in terms of maximum possible output of steel if all plants are producing at full tilt is usually very, very bad.
This is true in other industries as well. For example, the economics of cruise lines are truly abysmal during periods of oversupply. Passengers actually spend money on board beyond their tickets. So it is in a cruise line’s interest to sell a ticket at a loss if it means filling a bunk with someone who will shop, gamble and go on shore excursions once aboard the ship. But the cruise industry is very, very concentrated, with the top three companies holding most of the meaningful supply in the areas I care about. The competitors are also often a lot more rational than steelmakers. So I was very worried about the danger of Chinese steelmakers producing steel at zero profit for years and years. The point is, I did not invest in Posco long term.
Around the same time, a company called Hanesbrands (HBI, Financial) was spun off from Sara Lee. The stock was highly leveraged, but I liked its competitive position. Hanesbrands and Fruit of the Loom (owned by Berkshire Hathaway (BRK.A, Financial)(BRK.B, Financial)) are really the only underwear companies in the U.S. with the scale to supply the big retailers. They are also the two companies with recognizable enough brands that buyers will not feel like they are buying cheap, knock-off underwear. So it was potentially a fine business. I do not necessarily invest for 11 years (that is about the time since I looked at Posco and HBI), but if I did – I would only consider Hanes. I would never buy and hold Posco because competition can hurt it even if it is run well. Competition is not going to do a ton of damage to Hanes. There is competition and some of the company’s customers have a lot of bargaining power. That constrains profitability, but is never going to undermine the long-term health of Hanes as a business. Posco is always at the mercy of a highly cyclical industry. Hanes is not.
Cyclical industries tend to be cyclical because they are irrational. Not enough investors really understand this. Here is the truth: cyclicality is always the result of a “mistake.” A cycle has to be caused by an intertemporal miscalculation of some kind. If everyone in an industry made the correct choice, there would be no cycle. A cycle is a mistake followed by a correction, which then tends to go too far in the other extreme and needs its own correction, and so on. Some cycles are super short. If you have too many egg-laying hens in the egg industry – you can fix that fast. If you order more toys than you should have starting sometime in November – you are going to be marking them down and getting them off your shelves by February. There are so many little miscalculations of this sort that we do not think of them as cycles. For instance, almost all clothes are subject to economic miscalculations very similar to long-term cycles, but limited to individual fashions and styles and “corrected” for very quickly. At the company level, this is not really relevant unless you are looking at some sort of closeout retailer or something that takes big chunks of someone else’s bad inventory and sells it at a really low price.
What worries me at Car-Mart (CRMT, Financial) is not that the company is in a bad point in the industry cycle right now. It is that irrational competitive decisions by rivals are contagious. As you know, Car-Mart’s customers are deeply subprime (they really have no credit at all), so they are in a sort of perpetual recession. For this reason, Car-Mart’s customers cannot have higher payments when the economy is better. So if cars become more expensive (and used cars had been increasing in price faster than inflation for a while), you have to extend the number of months of the loan to keep the regular payment at the same level. That is risky. This particular cycle is really dangerous because we have seen an increase in all the bad credit metrics at Car-Mart despite the job market getting better.
Frankly, the labor market right now is super tight. You can see this in the tremendously high wage increases people who are quitting one job to take another are getting. That should not happen when there is enough idle labor to fill the regular churn of positions. It means we have not seen a lot of inflation in wages, but that people who want jobs can get jobs. So now should be a really good time to keep current in your car loans, yet some people are missing payments and falling behind. Why? Because in the last few years a lot of people took car loans they cannoy repay even if they experience no bad surprises.
This last quarter, about 31% of Car-Mart’s receivables were dedicated to a provision for credit losses. That is a high number compared to a past that often had Car-Mart taking around a 25% provision. Now at the worst point in a recession, you might understand that, but we are at about the best point in a job boom. So you know that some of the car loans made during these past few years were the worst, riskiest loans the industry has ever made and is likely to ever make for a long time in the future. The catalysts for this were obviously a low federal funds rate for a long time and lower than normal new car production. These two things happened as a result of the financial crisis. Keeping the federal funds rate low led to an overly aggressive desire for assets with yield, especially short-term assets like car loans. Additionally, lower new car production in the aftermath of the financial crisis obviously led to higher used car prices. Used cars are the collateral for loans made to purchase Car-Mart-type vehicles. So you had loose credit and inflated prices on the collateral. If this is a one-time deal, we can analyze the situation and decide if it can ruin an investment in Car-Mart or not. That is usually not a hard calculation to make.
For example, I looked at what it would take for the oil price bust to cause unacceptable losses in energy loans at Cullen/Frost Bankers (CFR, Financial). The stocks of some Texas banks dropped like investors expected the oil price bust to cause meaningful problems for these banks. There was no justification for that. Frost was only putting about 16% of its loans into energy and loans were only about half its assets – so people were worried about 8% of the company’s balance sheet going bad. The bank earns a couple percent on its assets before the provision for credit losses, so you would need losses in the 25% to 50% range on the energy loan portfolio to threaten the bank with an insufficient capital position. It just was not a serious problem. But in addition to that, energy lending is not that important to Frost. While I do not think Frost does energy lending as well as BOK Financial (BOKF, Financial) does– they both do OK compared to how other banks lend in other areas. As a result, I am not worried about energy lending being a permanently insufficient return business because of lenders' mistakes caused by the oil boom and bust cycle.
Am I worried about that happening at Car-Mart? Absolutely. The entire rationale for that investment was based on the rate at which Car-Mart could compound net receivables per share. In other words, I looked at what the company had in receivables after the provision for credit losses in each year and asked how quickly it could grow that number per share over the next 5, 10 and 15 years. The rate of growth in net receivables per share is the gain in intrinsic value at the company. That rate is directly threatened by the behavior of rivals. Overzealous lending by Car-Mart’s competitors can seriously ding the long-term rate of growth in net receivables per share. Obviously, the long-term return in Car-Mart’s stock should almost perfectly match the long-term growth rate in receivables per share. This is much like the way the long-term growth rate in an insurer's book value should drive the long-term growth in that stock. There are a lot of areas in insurance where even a very good underwriter and investor cannot drive strong book value growth if rivals are repetitively and cyclically stupid in their behavior. Most insurers earn bad returns on equity and underperform as stocks because they cannot insulate themselves from the cycle caused by the miscalculations of their rivals. I am worried Car-Mart could end up in the same situation.
On the other hand, now is obviously a good point in the cycle to consider Car-Mart stock. The stock is priced “normally” on an earnings basis. Earnings are cyclically depressed however. That tends to be when you should buy a great company facing cyclical problems. The P/E looks normal to investors, but price-book, price-sales, price-premiums and price-deposits are clearly wrong. Frost is a good example of that. The P/E was a perfectly normal 15 or 16 when I bought the stock. The price to deposits was extraordinarily low however. It was a great bank priced at what even a good bank should not be priced at. Car-Mart is much less insulated from competitor actions than Frost is however. Therefore, I cannot recommend Car-Mart with the kind of confidence I can recommend Frost. It is worth mentioning I have 40% of my portfolio in Frost and 0% of my portfolio in Car-Mart, and there is probably a reason for this. But yes, at this point in the cycle, now would be the time to start seriously considering Car-Mart.
Start a free 7-day trial of Premium Membership to GuruFocus.
