Other things being equal, good quality investing will work because the combination of free cash flow and future growth will lead to higher returns than suggested by the P/E you bought the stock at. This is because of the ability of the company to convert high amounts of earnings into free cash flow and still grow. A good current example of this is John Wiley (JW.A, Financial). The book business (which is still a substantial part of their business) has poor future prospects. However, I think the Journal Publishing business (which in my view, now makes up the majority of the company's value) can grow by anywhere from 2% to almost 5% a year without much, if any reinvestment in the business. In fact, it didn't really shrink even with tight budgets in the U.S. and especially Europe this year. It seems to have pricing power specifically and bargaining power generally. For this reason, I think that when JW.A trades at about 11 times earnings - which we'll call a 9% earnings yield, it is actually (the journal business at least) priced at a level that may deliver returns that meet my expectations (11% to 14%).
This is interesting because my estimate of 2% to 5% potential growth in that business is a pure quality approach. I do not assume the academic publishing business becomes any bigger relative to nominal GDP of U.S., EU, etc. I just assume the company can increase cash profits at least as fast as it increases the top line and it can increase sales anywhere from the rate of inflation to the rate of nominal GDP growth (but no faster). In no sense is that a growth company, but in my view, it is still likely to beat the market when it trades under 12 times earnings. In fact, I think JW.A's academic publishing business would be "fairly priced" in the sense it would offer market matching returns at the "normal" P/E of 15 to 16. Again, this is interesting, because JW.A is clearly a slow to no grower. It can not be seen as a growth company. And yet I think - purely because its business quality allows high free cash flow and pricing power - it is able to be worth as much as businesses of lower quality but better growth prospects. In this sense - and this is a very Donald Yacktman approach - quality can, if it leads to higher FCF/Net Income and lower Reinvestment/Growth - basically be a substitute for growth. I believe this is what happened at See's Candies. It is most obvious in situations where companies with 0% volume growth over a decade raise prices by 3% a year and use all free cash flow to buy back stock. In such situations, EPS can grow almost 10% a year (depending on what P/E the buybacks were done at and whether margins expanded with price increases) even while volume did not increase at all. Other than unregulated intangibles based monopolies - IMS Health, Moody's (MCO, Financial), Dun & Bradstreet (DNB, Financial), etc. - this is a situation most people will never see. Logic suggests that if the market now trades around 16 times last year's earnings and I do not expect the S&P 500 to grow its earnings by more than 10% a year, that such a company - despite being no growth - will actually beat the market if it trades at the same (16) P/E.
In this sense, we can say that at the extremes - where free cash flow always exceeds net income, returns on capital are nearly infinite, and pricing power allows increases equal to or above inflation even at 0% demand growth and the reverse (where free cash flow is always below net income, returns on capital are value destroying, and prices are deflationary at 0% volume growth) the P/Es should be very different. In this sense - although many value investors may disagree - a "bad" quality business may truly be worth a P/E of 8 and a "good" quality business may be worth a P/E of 24. I am not saying you should ever pay 24 times earnings for a great "no-growth" business anymore than I am saying you should buy a random established company at 16 times earnings. But I am saying that if you take the stereotype of the worst profitable business you can imagine, the stereotype of the most average profitable business you can imagine, and the stereotype of the best business you can imagine that at P/Es of 8, 16, and 24 respectively - they may be priced today to end up giving you roughly the same returns over the next 15 years. This does not require a true "growth" explanation. A pure quality explanation is sufficient. Volume growth is not needed to create that kind of pricing difference if it is pure price growth in the best business and it is actually value destroying in the worst business.
Graham's use of an 8.5 P/E is interesting. It is roughly equivalent to a 12% earnings yield. If you look at historical returns in stocks, etc. I think Graham got it about right. If a company has literally no opportunity to grow it should still be an acceptable investment at a 12% earnings yield. Today, you could argue the number is closer to 8% or so (but this is purely because of low interest rates that will not last forever). If we look before and after Graham's death we see this 12% yield as a pretty good estimate of what no growth should be valued at. It also works pretty well if we imagine a growth business trades around 16 times earnings. We can think of a true growth business (not a fast grower, but certainly an economy speed type grower) trading at around 16 times earnings and a no grower trading around 8 times earnings. In this situation, it is unclear to me which will win a 15 year investment race. A true no grower at 8 times earnings could beat a true grower at 16 times earnings. But the grower could win too. If the race is between a traditional value stock at 8 times earnings (with absolutely nothing else to recommend it) and the S&P 500 at a Shiller P/E of 16 - I think it is a fair race. In fact, I think the only reason that - over holding periods as long as 15 years - the "value" stock still tends to win the race is that investors aren't great handicappers. It is very easy for investors to think a 3% to 6% grower is a 0% grower. If they price a 6% grower at a P/E of 8, the race will not even be close. A 6% grower at a P/E of 8 will - over a holding period of 15 years - absolutely crush the S&P 500. In fact, a 6% grower trading at a P/E of 8 would make a good long-term investment.
Your last point about "solving for growth" is a great way of inverting. In fact, although people frequently cite Graham's growth formula, they forget this is what he was doing. What Graham was saying is that if you see a stock with a P/E of 28.5 - you can view that stock as having an expected return over the medium term future (he used 7 years as I remember) of 10% a year. That's because 10 times 2 equals 20; 20 + 8.5 = 28.5. Therefore, the market is assuming 10% growth for the company over the next 7 years or so.
I think that would be about right in Graham's days. Today's stock prices are a little different. The Shiller P/E is higher than at all points during Graham's career. Interest rates are lower than at most - but not all - of the time Graham was investing. The big difference is that when interest rates were low in Graham's day stock prices were not high. Graham did not have quite the experience we are having now of simultaneously low interest rates and earnings yield. And Graham was both a stock and bond investor. So, he didn't face quite as hostile an investing environment as we do in searching for any high returning alternatives.
There's another way we can tackle the situation. Assume you knew what the market P/E was. I think it is usually best to use the Shiller P/E for the entire market. Especially since I've seen no evidence that single year backwards looking P/Es of the vanilla flavor - and certainly not forward P/Es - have any real predictive value for future returns. People talk about whether the Shiller P/E adds anything over the traditional P/E. To me, the problem is that it's unclear the traditional P/E ever had much power as a value indicator on its own.
But, for now, let's say we know the P/E is 16 on the S&P 500. Or, better yet, let's just stick to yield numbers. So, let's say the P/E is 16.67x. That's an earnings yield of 6%. The historical growth rate in stocks' earnings per share has been about 6% in the U.S. There are reasons to think it will be a bit lower in the future (mostly because population growth in the U.S. and markets where U.S. companies export are all expected to be lower in the future than they were in the distant past - this might chop 1% off growth, not much more than that). We are left with a situation where we assume the usual 5% to 6% growth. Now, if we take a hardline reversion to the mean approach we would say that if you bought the S&P 500 today and held it for 15 years you would earn only the 5% to 6% a year plus dividends (Value Line tells me the median dividend yield on the 1,700 stocks they cover is now 2.3%). In other words, if today's P/E ratio is truly normal - and your portfolio really is getting a 2.3% dividend yield - you can earn 7.3% to 8.3% over the next 15 years in big American stocks.
I think this is the middle of the road - moderate center - consensus view right now in investing. I do not entirely agree with it. I think it is a possible scenario. But I think it is on the very sunniest edge of the possible. As I see it, there is no reason to believe today's earnings are at a cyclical low of any sort. Wages are not high. Taxes are not high (relative to what the government is spending). And interest rates are not high.
Companies profit from low wages, low taxes, and cheap loans. At the moment - with plenty of unemployment, a government interested in increasing demand, and banks who have more deposits than they need to fund their loans - there are offsetting factors for corporate profits that mean economic recovery may not have to fall straight to the bottom line. This is why I tend to favor the Shiller P/E. Because, without it, investors may say they expect GDP to grow 6% a year over the next 15 years and they expect corporate earnings to follow in lockstep. They may not ask - what about taxes, interest, etc.? - they may just assume an increase in sales is an increase in operating income is an increase in after-tax income.
I am not smart enough - even Warren Buffett's article on ROE recently republished in Tap Dancing to Work turned out to be totally wrong - to consider each of these factors carefully. I can, however, notice when they seem unusual favorable or unfavorable. I think wages, interest, and taxes are on the favorable side of things right now for U.S. companies. Now, maybe demand isn't, maybe energy costs aren't, maybe currency isn't. I don't know enough to know that. But I think I know enough to be extra cautious. And so I think I'd like to use the Shiller P/E.
GuruFocus tells me the Shiller P/E is 23 and the historical mean is 17. In that case, we would assume that your return over the next 15 years in the S&P 500 will be determined by today's Shiller P/E (the one you buy at), the growth rate of EPS in the S&P 500, and the 2028 Shiller P/E (the one you sell at).
In this scenario, the 15-year price appreciation would be 4% a year if we assume 6% a year earnings per share growth. That's because the other 2% a year is needed to eat the multiple contraction from 23 to 17 over a 15 year adjustment period. Of course, we still get the dividend yield. I'll assume it's the 2.3% a year Value Line gives as the median. So, assuming 5% to 6% growth in earnings per share over the next 15 years we get expected returns ranging from a low of 5.2% on the low end (2.3% dividend yield plus 2.9% price appreciation under 5% EPS growth and Shiller P/E reversion over 15 years) to a high end estimate of 8.3% a year.
That, to me, sounds about right. Truly long-term returns (like 15 years or more) in big American stocks are probably in the 5% to 8.5% range.
Can investors like Donald Yacktman, Warren Buffett, etc. do better?
I think they can. I think durability, quality, value, capital allocation, and growth are all sometimes mispriced in the market.
Right now, a particularly interesting situation is the compression (or is it confusion) of companies with above average durability and quality and companies with below average quality and durability. Many seem to be trading at close to the same P/E.
If we look at banks, we can understand why Warren Buffett keeps buying Wells Fargo (WFC, Financial). Is it that Wells Fargo is cheap relative to other banks, that all banks are cheap, or that Wells Fargo's quality and durability is higher than banks with the same P/E?
I think it is that he believes banks are probably on the cheap side. And that Wells Fargo is certainly high in terms of the durability and quality of its business. So, if over the next 15 years, he gets above average growth for a bank in terms of EPS and then he gets a 50% higher P/E ratio at the end of these 15 years - he's looking really good.
I do not see a lot of value opportunities in today's market. I have a screen where I look at all the companies in the U.S. that have been profitable for 10 straight years and now trade at less than 8 times EBITDA. It's still a pretty long list (about 200 stocks right now). But it is all scrunched up at the top of the list. There are some for profit education stocks, etc. down in the low single digit multiples of EBITDA. But a large number of stocks of varying quality are in the 6 to 8 times EBITDA zone. This is not an especially great price. It's an okay price for an okay business. It's only a really good price for a really good business. This is where I think the opportunity is in American stocks right now. It's in fairly priced quality companies.
I like to look at five return possibilities when considering these stocks:
1. The return you could get - right now - if the company sold to a control buyer
2. The return you could get if - over 3 years and in a normal business climate - the company's price rose to match its "peers"
3. The return you could get if you held the stock for 5 years and sold it in a normal business climate
4. The return you could get if you held the stock for 15 years and sold it in a normal business climate
5. The return you could get if you held the stock forever
Sometimes, there is a #0 which is a floor. I like those best. But buying below hard floors - made up of cash, net current assets, real estate, etc. - is rare at most stocks. It is very, very rare at companies I would be willing to buy and hold for the long haul.
I've adopted a several "point-to-point" return calculation approach over time. Mostly this is based on practical experience. I realized over time that when selecting for asset quality it was best to focus on forever return potential. Basically, to buy like Warren Buffett. However, I found that because a significant amount of my returns over time came from just two sources - corporate takeovers and price-to-something mean reversion - it also made sense to look at the immediate upside (what can this company be sold for now) and short-term reversion to sanity by the industry, economy, stock market, etc. That is what the 3-year return number is for.
I think it is possible to earn a good return (10% to 15% a year) thinking only of returns beyond a 5 year holding period. You can be a truly long-term investor like the modern day Warren Buffett, Phil Fisher, etc.
However, I think all returns of the 20% to 30% a year varieties tend to come from the potential in #1 and #2 considerations. You are either buying "deep value" or "being greedy when others are fearful" or getting "a private owner value" in these situations. You are not buying and holding Coca-Cola etc. Donald Yacktman's approach is purely based on #3 through #5 type consideration.
Over time, I have learned a tough lesson. In moments of fear, panic, contempt, etc. - at companies, in industries, in countries, in markets, etc. - it is considerations #1 and #2 (what would a control buyer pay for this today and what - when the dust settles in 3 years - will the average investor pay for this stock) that makes you all your money. In "normal" times it is considerations #4 and #5 that will make you all your money. In other words, if you have to invest equal amounts of money across time, you never need to focus on a return expectation for a holding period of less than 15 years. However, if you have the opportunity, the stomach, etc. to dive into a situation where there is true psychological distress on the part of buyers, sellers, etc. you can make as much money in 1 to 3 years as it will take some people to make in a decade.
The problem is that these moments of panic rarely coincide with your circle of competence. So, you often have to learn about new industries, new countries, etc. while they are already making gruesome headlines.
The tough lesson I learned is what to do the rest of the time. It is fairly easy to know how to figure out returns are "good enough" in a stock if you buy it when stocks generally, the country you are investing in, the industry you are investing in, and the specific stock you are investing in are out of favor but otherwise sound places to put money. The hard part is knowing what to do if you - say - know the most about American stocks and expect they will only return 5% to 8.5% a year over the next 15 years. What do you do then? What do you do when stock prices are high?
Most value investors do the wrong thing. They look for the last cheap dregs. They see that the bargain bin has been emptied of 900 of the original 1,000 pieces in there - and they decide they will sort through the remaining 100 trinkets looking for that one good deal that's left.
When future potential returns are low this is probably a bad idea. You are likely to end up trading a lower price (value) for a lower quality company (quality). These two factors offset. Value investing is like "value" in shopping. It doesn't just mean a low price. It should mean a higher quality of merchandise than seems reasonable to be sold at that price.
Over time, I have realized it is much better to slide toward considerations #4 and #5 when prices are high. It is better - when the market is high - to think of buying and holding the best companies you can find at acceptable prices than to try to find the last bargains left. The biggest reason for this is that - as long as you don't confuse insanely high prices with acceptable prices - you will tend to lose only time in high quality companies where you are paying a price value investors might not like. When you buy lower quality merchandise - especially cyclically, operationally, and financially leveraged low quality merchandise at a seeming bargain price - you actually run the risk of losing more than time. You risk losing money. You risk a permanent impairment of capital. That risk is always highest when investors are jolliest.
Even Ben Graham knew this. And he said as much. But I have been slow to recognize this fact. Even after I understood the upside of quality merchandise, I didn't pay enough attention to the special risk of compromising on quality because stock prices generally are just too high. And yet that has been the ruin of many otherwise intelligent value investors.
I think it's hard to solve for growth and price quantitatively in a way that makes practical sense. It is very easy at the extremes. In other words, if we avoid using a spectrum of values to deal with and instead focus on a few static points useful as thought experiments - especially pricing increases only, value neutral growth, etc. - I think we can understand growth and price better. I still have trouble with the inbetween situations. Practically, how do you deal with a company retaining 50% of earnings and paying out a stable and slightly rising dividend plus buying back stock in fits and starts over time while growing in a way that growth probably creates some value?
That's a pretty common scenario. I know it is certainly a difficult to deal with scenario in terms of finding a practical solution to the problem. It's not easy to value that growth. The company is growing, but is the return on marginal capital employed stable. How likely is it that stock repurchases and growth in the future are likely to actually add value? It's very easy in such situations to see that a little growth right now will encourage both more stock repurchases and more expansion and yet both of these may therefore be ill-timed because they lean into a boom. Looking backwards, we can value what the company did in terms of creating or destroying value through share repurchases and through organic growth. But we invest looking forward.
My biggest problem is that return on the next dollar of retained earnings is hard to know at most companies. I believe it becomes harder to know as a company grows faster than GDP. I know investors love industries that grow faster than GDP. I know they love sectors that grow faster than GDP. I know they love locations in cities and states that grow faster than GDP. I know they love a business with customers who are growing faster than the overall population.
I have never been able to translate that into a situation where I was confident that a business catering to fat, old, diabetics in Texas will actually grow faster than a business catering to 30 year olds in New Jersey. That would only be true if some government were issuing a monopoly on serving fat, old, diabetic Texans. In my experience, the reverse tends to be true. While no one has a monopoly on a static populations they do - once they are well established - have less to fear from hot growth hunters. They tend to be left alone. And if they tend to their niche very carefully they tend to do as well over the truly long term - like my 15 year look into the future - than the company that should have all the tailwinds.
Of course, like any Warren Buffett wannabe I dream of finding that business with sticky customers, a great brand name, etc. and all the demographics going for it.
Unfortunately, in those situations where I think I understand both that growth is likely and that there is some sort of durable competitive advantage, the market more than agrees with me.
See Copart (CPRT, Financial), Boston Beer (SAM), and Under Armour (UA) for details. I wouldn't call those fair prices.
Now see John Wiley (JW.A), Weight Watchers (WTW), and Dun & Bradstreet (DNB). I wouldn't call those growth business. One (Weight Watchers) is debatable.
Even in cases where I think there is potential for future growth - but it hasn't even shown up at all in the financial results yet - the market is pricing the stock too high for me. See DreamWorks (DWA). At $13 a share, you'd be getting in under book value. Nothing in their 5-year GAAP ROE results: 14% in 2008, 14% in 2009, 14% in 2010, 7% in 2011, LOSS in 2012 suggest this is a company investors would be eager to pay a big premium to book value for. But the stock - which I'd be eager to buy at $13 - is now trading around $19. The stock is trading around a 16 P/E of normalized earnings if earnings since it went public are basically normal. I think it has real growth opportunities.
If I was sure of that growth, maybe today's price would be a fine price to pay. That is usually the problem with trying to apply a quantitative approach to the right price and right growth rate problem. A good model is one that is theoretically correct, that you can understand and feel comfortable using, and that gets good practical results.
The problem with an investing model is that for investors - unlike economists - it has to be a model that tells you something different from the market. It is only in being right in a way that others are not right that you win in investing.
Bill Miller is the person I associate most with the idea that complete DCF type approaches and value investing can work well together. Basically, you can use a DCF to get an intrinsic value and then you can buy at a steep discount to that intrinsic value to get your margin of safety. I've never been able to use that approach. And I'm not sure it's a good one.
I think some sort of comparative approach makes more sense. An approach based on what P/E ratios, Shiller P/Es, Price-to-free-cash flow, EV/EBITDA, etc. the market and peers now trade at and what they are expected to grow the bottom line by over the next 7 years or 15 years or whatever makes sense to me. Likewise, a historical approached based on what P/E, Shiller P/E, Price-to-free-cash flow, EV/EBITDA the company, peers, the market, etc. has traded for in the past makes sense to me.
Always using both a "normal" P/E and a "normal" growth rate based on comparative and historical data as reference points seems logical. It seems like you should be able to look for companies where you are sure they will grow closer to 10% than 5% and then try to always buy them at 5% growth type prices and you should be able to find 5% growers at no-growth prices and buy those too. In other words, there should be a good way to know you can pay 8 times earnings for a stock because its future growth is really just as good as the market and the market often trades at 16 times earnings. And you should be able to pay 16 times earnings for a business you think will grow 10% a year rather than 5% a year like many 16 times earnings stocks.
But here is the real question: Can I buy Copart at a 24 P/E?
On a trailing 5-year average basis, there is no year in the last 10 - and the last 10 for cars on the road growth in the U.S. were as bad as ever - where Copart grew EPS by less than 10% annually. Let's assume I think they can grow EPS by 10% a year. A P/E of 24 is higher than an earnings yield of 4%. And a 4% earnings yield growth plus 10% is 14%. That's definitely an acceptable return for me personally given today's stock prices. At any point in time, I'll take 15%. I'd like 10% at a minimum. And, as we discussed, I think the next 15 years won't see much better than 5% to 8.5% returns in U.S. stocks. Personally, I've often given 7% a year returns as the high end of future returns expectations for the S&P 500. So, 14% a year certainly clears those hurdles.
You mentioned the retention rate approach. Copart - I'm using their Value Line page here - has essentially retained all earnings all the time since going public. Return on total capital has ranged from 11% to 19%. Return on equity has ranged from 11% to 33%. The two only decoupled when Copart bought back stock very recently. I believe they are likely to have more free cash flow than they can use in the future and therefore must decide on buybacks, dividends, etc. going forward. Previously, they could always buy enough small competitors out in the U.S. as a roll up. They and their biggest competitor now account for too much of the market (almost three-quarters) to make that approach a possibility in the U.S. any longer. Obviously, free cash flow has been rising while the number of potential competitors to buy out has been shrinking. They plan to expand internationally. There are hurdles to clear. But I expect U.S. companies will go from the U.S. market to other countries in this business. I think the reverse is unlikely. Someone is going to take the American business model in this industry from country to country around the globe over the next few decades. This is especially true because already a meaningful amount of sales are cross border. Before Copart and IAA rolled up competitors and went online there were very few cross state border sales in the U.S. Now, there are plenty of cross country sales. These are networks. So growth seems likely to be foreign. But I think it's still possible. This is on top of baseline growth equal to U.S. car population plus car price inflation (basically I view annual percentage nominal dollar growth in the salvage value of cars in the U.S. as a given for CPRT because of their competitive position).
Let's be pessimistic and exclude the recent ROE boost. This is due to leverage. For the record, Copart uses less leverage than most companies in a similar position would. They own a lot of their land. Because the core business is good and uses the land for its operations, this land should not be assigned its own value. However, it is worth noting that Copart operates with a ton less borrowed money relative to the collateral and stable cash flow they have then most of the retailers, restaurants, etc. value investors look at everyday. For these reasons, I think it shouldn't be hard for return on equity in the future to be as good or better than return on total capital has been in the past. This is also helped by goodwill. The likely leveraged return on tangible equity someone would get out of this business is certainly at least in the 11% to 20% range Copart has achieved on its total capital.
Let's take 15% as the midpoint. Let's assume Copart can return 15% on its equity and grow its book value by 10% a year through retained earnings. They will have to do something else with the rest of that money.
Earnings are now $1.43 a share. They will grow earnings by 10% a year. That means earnings will grow by 14 cents next year. If they earn 15% on incremental equity, they will need another 93 cents of book value ($0.14/0.15 = $0.93). That would leave 50 cents of earnings they don't need to retain. Copart doesn't pay a dividend. They can buyback stock. For simplicity here we'll assume they pay out all earnings not retained as a dividend. We would then have two series of cash flows over the next 15 years. One would go back into the business to increase earnings. The other would be paid out to shareholders. That would leave earnings and dividends looking something like this:
EPS
2013: $1.57
2014: $1.73
2015: $1.90
2016: $2.09
...
2028: $5.97
The stock at the end of 2028 - if it sold at 16 times earnings as a "normal" P/E for a stock - could then be sold for $95.52 a share.
Meanwhile, the dividend payments would look like this:
DPS
2013: $0.50
2014: $0.55
2015: $0.61
2016: $0.67
...
2028: $2.09
So, in 2028 we would own a stock earning $5.97 a share and paying $2.09 a share in dividends. We assume this stock would be trading at $95. This is equal to a P/E of 16 and a dividend yield of 2.2% at the time of our hypothetical 2028 sale of the stock.
If you plug just the dividend portion - the non-retained earnings - cash flow stream into a DCF using 10% growth over the next 15 years in the dividend, 6% terminal growth (basically nominal GDP), and use our high end for the S&P 500 return of an 8.3% as the discount rate you will get a value of $37.62 a share just on the dividend portion of Copart's hypothetical future. As I'm writing this, the stock is trading at $34.36 a share. So, we can just (a tad conservatively) cancel those numbers out. Copart is - under my 10% a year growth for the next 15 years, 15% ROE, 6% terminal growth, 8.3% discount rate, etc. - assumptions trading at the value of its future dividends alone. The stock is worth more than its future dividends. This is because the company retains earnings and the retained earnings add value because they are assumed to earn a higher return than we could earn on our dividends. Our estimate of a $95 market price for a stock earning $5.97 a share in 2028 supports this.
For the record, the DCF thinks the retained earnings stream is worth $107.60. That's higher than our 16 P/E in 2028 so I'll ignore the DCF just like I canceled out the dividend and the share price. In both cases, the DCF wants me to be a bit more aggressive. It's safe to ignore that.
So, the only question becomes how much profit is there in Copart? How willing should I be to trade $34.36 today for $95 in 2028?
On a CAGR basis this is 7% a year of pure profit. Simply put, the DCF thinks Copart will run 7% a year ahead of the market over the next 15 years.
I have to say that when we use my carefully cherry picked, oddly idiosyncratic estimates for Copart, the S&P 500, etc. I think the DCF looks pretty smart. Or at least it sounds a lot like me. I'm not sure that's the same thing. But, yes, I actually agree with the DCF here. The earnings Copart won't be able to find a place for inside the business are probably worth about what the stock trades for. The really speculative component - beyond the mere assumption of growth - is the issue of what growth and ROE will look like together. This is always the tricky part.
I'm not sure I can guarantee 10% growth at Copart. I think if we knocked the estimates down to 6% a year growth combined with 15% a year ROE, I'd be real comfortable Copart is worth at least that much. This is unusual in that I think there will be 6% growth, it will return 15%, etc. and downside surprises will be much less likely than with stocks generally. In other words, I see a wide moat around a good and growing business. There are very few companies aside from Copart where I could say that.
But can we quantify this? Can we then say 24 times earnings is a fine - even good - price to pay for Copart? Even using my confident estimates of 6% earnings growth forever and $6.65 of added equity needed for every $1 increase to earnings (in other words, a 15% marginal ROE), we'd get values for the stock using a DCF that are close to double today's price. Bill Miller would be fine with those. Would Warren Buffett?
What's interesting here - in a practical rather than theoretical sense - is that I'm really quite confident in that 6% growth number and 15% ROE number and I'm really quite confident in them both being perpetual figures. This is not just a thought experiment. I really think Copart can grow 6% a year and return 15% on equity literally forever.
What is a company like that worth?
I crunched the numbers separately on price appreciation and dividends over the next 15 years. In other words, what would just 15 years of price appreciation be worth in Copart if it grows 6% a year and ends up at 16 times earnings in 2028? You would have to buy the stock at $16.60 a share for that to make sense. (Again, we're using an 8.3% discount rate - my high end for the S&P 500). Meanwhile, the dividend stream on a company that pays 86 cents this year (that's what Copart would have left to distribute in earnings if it grew only 6% next year and earned a 15% return on incremental equity) and growing by 6% forever would be worth $39.63 a share.
In other words, the quality/growth part of the company - the growing free cash flow that can't be put back in the business - would be worth more than the stock trades for today ($39.63 vs. $34.35). However, the value part of the equation actually takes away value. On a valuation only basis - looking just for stock price appreciation - Copart is extraordinarily overvalued if it only grows 6% a year and you only hold the stock for 15 years. The price appreciation potential of a 6% a year grower over 15 years that already trades at 24 times earnings today is exactly what you'd expect - it's worse than the S&P 500.
This brings us to the Warren Buffett / Ben Graham paradox. Buying and selling Copart is no way to make money. But holding Copart is theoretically going to make you money. In fact, you could argue that Berkshire Hathaway could afford to pay the very high looking price of 24 times earnings and still make money in Copart. On the other hand, if investor sentiment sours on Copart just enough so the company is seen as an "average" stock it will have a 33% drop ahead of it. The company is overvalued if the future for the company is the same as for the average public company. It's only undervalued if we count the extent to which Copart's future creates more value than the future of other public companies.
I think it's a near certainty that will prove true. Copart's future will be more valuable than the S&P 500's future. So should we count it?
I’m not sure.
Warren Buffett can count quality to its fullest because Berkshire is never going to sell what it buys. I have a hard time recommending a stock you need to hold a long time to overcome the fancy price you pay upfront.
Why?
Because I know anyone who acts on that recommendation is unlikely to keep the stock forever.
It’s hard to lose money holding a quality company forever. But it’s harder to find investors willing to hold a company forever.
The numbers on quality add up. Paying up for quality – in the sense of a company that needs to retain less earnings to grow at an equal rate to other companies – makes every bit as much sense as paying up for growth.
The problem with both the growth approach and the quality approach for most investors is that it has short-term risks and long-term rewards. If you pay 24 times earnings for a stock – and you are right to do it – that stock could easily drop 33% if market sentiment on the stock becomes just “blah”. Not “ick” but “blah”.
Unless you are the kind of investor who will see the market say “blah” and still believe in the business – which really requires the same stomach as buying an “ick” stock does for the value investor – than I’m afraid quality investors will enter a stock with the best of intentions but exit it too soon.
I don’t think there is an intellectual cure for this problem. What you need is the right personality rather than the right rationality.
In that sense, sticking with a quality company for the long-term is not really that different from buying a value stock. The difficulty in value investing is often having the courage to make the initial buy. The difficulty in quality investing is having the patience not to sell.
Ask Geoff a Question
This is interesting because my estimate of 2% to 5% potential growth in that business is a pure quality approach. I do not assume the academic publishing business becomes any bigger relative to nominal GDP of U.S., EU, etc. I just assume the company can increase cash profits at least as fast as it increases the top line and it can increase sales anywhere from the rate of inflation to the rate of nominal GDP growth (but no faster). In no sense is that a growth company, but in my view, it is still likely to beat the market when it trades under 12 times earnings. In fact, I think JW.A's academic publishing business would be "fairly priced" in the sense it would offer market matching returns at the "normal" P/E of 15 to 16. Again, this is interesting, because JW.A is clearly a slow to no grower. It can not be seen as a growth company. And yet I think - purely because its business quality allows high free cash flow and pricing power - it is able to be worth as much as businesses of lower quality but better growth prospects. In this sense - and this is a very Donald Yacktman approach - quality can, if it leads to higher FCF/Net Income and lower Reinvestment/Growth - basically be a substitute for growth. I believe this is what happened at See's Candies. It is most obvious in situations where companies with 0% volume growth over a decade raise prices by 3% a year and use all free cash flow to buy back stock. In such situations, EPS can grow almost 10% a year (depending on what P/E the buybacks were done at and whether margins expanded with price increases) even while volume did not increase at all. Other than unregulated intangibles based monopolies - IMS Health, Moody's (MCO, Financial), Dun & Bradstreet (DNB, Financial), etc. - this is a situation most people will never see. Logic suggests that if the market now trades around 16 times last year's earnings and I do not expect the S&P 500 to grow its earnings by more than 10% a year, that such a company - despite being no growth - will actually beat the market if it trades at the same (16) P/E.
In this sense, we can say that at the extremes - where free cash flow always exceeds net income, returns on capital are nearly infinite, and pricing power allows increases equal to or above inflation even at 0% demand growth and the reverse (where free cash flow is always below net income, returns on capital are value destroying, and prices are deflationary at 0% volume growth) the P/Es should be very different. In this sense - although many value investors may disagree - a "bad" quality business may truly be worth a P/E of 8 and a "good" quality business may be worth a P/E of 24. I am not saying you should ever pay 24 times earnings for a great "no-growth" business anymore than I am saying you should buy a random established company at 16 times earnings. But I am saying that if you take the stereotype of the worst profitable business you can imagine, the stereotype of the most average profitable business you can imagine, and the stereotype of the best business you can imagine that at P/Es of 8, 16, and 24 respectively - they may be priced today to end up giving you roughly the same returns over the next 15 years. This does not require a true "growth" explanation. A pure quality explanation is sufficient. Volume growth is not needed to create that kind of pricing difference if it is pure price growth in the best business and it is actually value destroying in the worst business.
Graham's use of an 8.5 P/E is interesting. It is roughly equivalent to a 12% earnings yield. If you look at historical returns in stocks, etc. I think Graham got it about right. If a company has literally no opportunity to grow it should still be an acceptable investment at a 12% earnings yield. Today, you could argue the number is closer to 8% or so (but this is purely because of low interest rates that will not last forever). If we look before and after Graham's death we see this 12% yield as a pretty good estimate of what no growth should be valued at. It also works pretty well if we imagine a growth business trades around 16 times earnings. We can think of a true growth business (not a fast grower, but certainly an economy speed type grower) trading at around 16 times earnings and a no grower trading around 8 times earnings. In this situation, it is unclear to me which will win a 15 year investment race. A true no grower at 8 times earnings could beat a true grower at 16 times earnings. But the grower could win too. If the race is between a traditional value stock at 8 times earnings (with absolutely nothing else to recommend it) and the S&P 500 at a Shiller P/E of 16 - I think it is a fair race. In fact, I think the only reason that - over holding periods as long as 15 years - the "value" stock still tends to win the race is that investors aren't great handicappers. It is very easy for investors to think a 3% to 6% grower is a 0% grower. If they price a 6% grower at a P/E of 8, the race will not even be close. A 6% grower at a P/E of 8 will - over a holding period of 15 years - absolutely crush the S&P 500. In fact, a 6% grower trading at a P/E of 8 would make a good long-term investment.
Your last point about "solving for growth" is a great way of inverting. In fact, although people frequently cite Graham's growth formula, they forget this is what he was doing. What Graham was saying is that if you see a stock with a P/E of 28.5 - you can view that stock as having an expected return over the medium term future (he used 7 years as I remember) of 10% a year. That's because 10 times 2 equals 20; 20 + 8.5 = 28.5. Therefore, the market is assuming 10% growth for the company over the next 7 years or so.
I think that would be about right in Graham's days. Today's stock prices are a little different. The Shiller P/E is higher than at all points during Graham's career. Interest rates are lower than at most - but not all - of the time Graham was investing. The big difference is that when interest rates were low in Graham's day stock prices were not high. Graham did not have quite the experience we are having now of simultaneously low interest rates and earnings yield. And Graham was both a stock and bond investor. So, he didn't face quite as hostile an investing environment as we do in searching for any high returning alternatives.
There's another way we can tackle the situation. Assume you knew what the market P/E was. I think it is usually best to use the Shiller P/E for the entire market. Especially since I've seen no evidence that single year backwards looking P/Es of the vanilla flavor - and certainly not forward P/Es - have any real predictive value for future returns. People talk about whether the Shiller P/E adds anything over the traditional P/E. To me, the problem is that it's unclear the traditional P/E ever had much power as a value indicator on its own.
But, for now, let's say we know the P/E is 16 on the S&P 500. Or, better yet, let's just stick to yield numbers. So, let's say the P/E is 16.67x. That's an earnings yield of 6%. The historical growth rate in stocks' earnings per share has been about 6% in the U.S. There are reasons to think it will be a bit lower in the future (mostly because population growth in the U.S. and markets where U.S. companies export are all expected to be lower in the future than they were in the distant past - this might chop 1% off growth, not much more than that). We are left with a situation where we assume the usual 5% to 6% growth. Now, if we take a hardline reversion to the mean approach we would say that if you bought the S&P 500 today and held it for 15 years you would earn only the 5% to 6% a year plus dividends (Value Line tells me the median dividend yield on the 1,700 stocks they cover is now 2.3%). In other words, if today's P/E ratio is truly normal - and your portfolio really is getting a 2.3% dividend yield - you can earn 7.3% to 8.3% over the next 15 years in big American stocks.
I think this is the middle of the road - moderate center - consensus view right now in investing. I do not entirely agree with it. I think it is a possible scenario. But I think it is on the very sunniest edge of the possible. As I see it, there is no reason to believe today's earnings are at a cyclical low of any sort. Wages are not high. Taxes are not high (relative to what the government is spending). And interest rates are not high.
Companies profit from low wages, low taxes, and cheap loans. At the moment - with plenty of unemployment, a government interested in increasing demand, and banks who have more deposits than they need to fund their loans - there are offsetting factors for corporate profits that mean economic recovery may not have to fall straight to the bottom line. This is why I tend to favor the Shiller P/E. Because, without it, investors may say they expect GDP to grow 6% a year over the next 15 years and they expect corporate earnings to follow in lockstep. They may not ask - what about taxes, interest, etc.? - they may just assume an increase in sales is an increase in operating income is an increase in after-tax income.
I am not smart enough - even Warren Buffett's article on ROE recently republished in Tap Dancing to Work turned out to be totally wrong - to consider each of these factors carefully. I can, however, notice when they seem unusual favorable or unfavorable. I think wages, interest, and taxes are on the favorable side of things right now for U.S. companies. Now, maybe demand isn't, maybe energy costs aren't, maybe currency isn't. I don't know enough to know that. But I think I know enough to be extra cautious. And so I think I'd like to use the Shiller P/E.
GuruFocus tells me the Shiller P/E is 23 and the historical mean is 17. In that case, we would assume that your return over the next 15 years in the S&P 500 will be determined by today's Shiller P/E (the one you buy at), the growth rate of EPS in the S&P 500, and the 2028 Shiller P/E (the one you sell at).
In this scenario, the 15-year price appreciation would be 4% a year if we assume 6% a year earnings per share growth. That's because the other 2% a year is needed to eat the multiple contraction from 23 to 17 over a 15 year adjustment period. Of course, we still get the dividend yield. I'll assume it's the 2.3% a year Value Line gives as the median. So, assuming 5% to 6% growth in earnings per share over the next 15 years we get expected returns ranging from a low of 5.2% on the low end (2.3% dividend yield plus 2.9% price appreciation under 5% EPS growth and Shiller P/E reversion over 15 years) to a high end estimate of 8.3% a year.
That, to me, sounds about right. Truly long-term returns (like 15 years or more) in big American stocks are probably in the 5% to 8.5% range.
Can investors like Donald Yacktman, Warren Buffett, etc. do better?
I think they can. I think durability, quality, value, capital allocation, and growth are all sometimes mispriced in the market.
Right now, a particularly interesting situation is the compression (or is it confusion) of companies with above average durability and quality and companies with below average quality and durability. Many seem to be trading at close to the same P/E.
If we look at banks, we can understand why Warren Buffett keeps buying Wells Fargo (WFC, Financial). Is it that Wells Fargo is cheap relative to other banks, that all banks are cheap, or that Wells Fargo's quality and durability is higher than banks with the same P/E?
I think it is that he believes banks are probably on the cheap side. And that Wells Fargo is certainly high in terms of the durability and quality of its business. So, if over the next 15 years, he gets above average growth for a bank in terms of EPS and then he gets a 50% higher P/E ratio at the end of these 15 years - he's looking really good.
I do not see a lot of value opportunities in today's market. I have a screen where I look at all the companies in the U.S. that have been profitable for 10 straight years and now trade at less than 8 times EBITDA. It's still a pretty long list (about 200 stocks right now). But it is all scrunched up at the top of the list. There are some for profit education stocks, etc. down in the low single digit multiples of EBITDA. But a large number of stocks of varying quality are in the 6 to 8 times EBITDA zone. This is not an especially great price. It's an okay price for an okay business. It's only a really good price for a really good business. This is where I think the opportunity is in American stocks right now. It's in fairly priced quality companies.
I like to look at five return possibilities when considering these stocks:
1. The return you could get - right now - if the company sold to a control buyer
2. The return you could get if - over 3 years and in a normal business climate - the company's price rose to match its "peers"
3. The return you could get if you held the stock for 5 years and sold it in a normal business climate
4. The return you could get if you held the stock for 15 years and sold it in a normal business climate
5. The return you could get if you held the stock forever
Sometimes, there is a #0 which is a floor. I like those best. But buying below hard floors - made up of cash, net current assets, real estate, etc. - is rare at most stocks. It is very, very rare at companies I would be willing to buy and hold for the long haul.
I've adopted a several "point-to-point" return calculation approach over time. Mostly this is based on practical experience. I realized over time that when selecting for asset quality it was best to focus on forever return potential. Basically, to buy like Warren Buffett. However, I found that because a significant amount of my returns over time came from just two sources - corporate takeovers and price-to-something mean reversion - it also made sense to look at the immediate upside (what can this company be sold for now) and short-term reversion to sanity by the industry, economy, stock market, etc. That is what the 3-year return number is for.
I think it is possible to earn a good return (10% to 15% a year) thinking only of returns beyond a 5 year holding period. You can be a truly long-term investor like the modern day Warren Buffett, Phil Fisher, etc.
However, I think all returns of the 20% to 30% a year varieties tend to come from the potential in #1 and #2 considerations. You are either buying "deep value" or "being greedy when others are fearful" or getting "a private owner value" in these situations. You are not buying and holding Coca-Cola etc. Donald Yacktman's approach is purely based on #3 through #5 type consideration.
Over time, I have learned a tough lesson. In moments of fear, panic, contempt, etc. - at companies, in industries, in countries, in markets, etc. - it is considerations #1 and #2 (what would a control buyer pay for this today and what - when the dust settles in 3 years - will the average investor pay for this stock) that makes you all your money. In "normal" times it is considerations #4 and #5 that will make you all your money. In other words, if you have to invest equal amounts of money across time, you never need to focus on a return expectation for a holding period of less than 15 years. However, if you have the opportunity, the stomach, etc. to dive into a situation where there is true psychological distress on the part of buyers, sellers, etc. you can make as much money in 1 to 3 years as it will take some people to make in a decade.
The problem is that these moments of panic rarely coincide with your circle of competence. So, you often have to learn about new industries, new countries, etc. while they are already making gruesome headlines.
The tough lesson I learned is what to do the rest of the time. It is fairly easy to know how to figure out returns are "good enough" in a stock if you buy it when stocks generally, the country you are investing in, the industry you are investing in, and the specific stock you are investing in are out of favor but otherwise sound places to put money. The hard part is knowing what to do if you - say - know the most about American stocks and expect they will only return 5% to 8.5% a year over the next 15 years. What do you do then? What do you do when stock prices are high?
Most value investors do the wrong thing. They look for the last cheap dregs. They see that the bargain bin has been emptied of 900 of the original 1,000 pieces in there - and they decide they will sort through the remaining 100 trinkets looking for that one good deal that's left.
When future potential returns are low this is probably a bad idea. You are likely to end up trading a lower price (value) for a lower quality company (quality). These two factors offset. Value investing is like "value" in shopping. It doesn't just mean a low price. It should mean a higher quality of merchandise than seems reasonable to be sold at that price.
Over time, I have realized it is much better to slide toward considerations #4 and #5 when prices are high. It is better - when the market is high - to think of buying and holding the best companies you can find at acceptable prices than to try to find the last bargains left. The biggest reason for this is that - as long as you don't confuse insanely high prices with acceptable prices - you will tend to lose only time in high quality companies where you are paying a price value investors might not like. When you buy lower quality merchandise - especially cyclically, operationally, and financially leveraged low quality merchandise at a seeming bargain price - you actually run the risk of losing more than time. You risk losing money. You risk a permanent impairment of capital. That risk is always highest when investors are jolliest.
Even Ben Graham knew this. And he said as much. But I have been slow to recognize this fact. Even after I understood the upside of quality merchandise, I didn't pay enough attention to the special risk of compromising on quality because stock prices generally are just too high. And yet that has been the ruin of many otherwise intelligent value investors.
I think it's hard to solve for growth and price quantitatively in a way that makes practical sense. It is very easy at the extremes. In other words, if we avoid using a spectrum of values to deal with and instead focus on a few static points useful as thought experiments - especially pricing increases only, value neutral growth, etc. - I think we can understand growth and price better. I still have trouble with the inbetween situations. Practically, how do you deal with a company retaining 50% of earnings and paying out a stable and slightly rising dividend plus buying back stock in fits and starts over time while growing in a way that growth probably creates some value?
That's a pretty common scenario. I know it is certainly a difficult to deal with scenario in terms of finding a practical solution to the problem. It's not easy to value that growth. The company is growing, but is the return on marginal capital employed stable. How likely is it that stock repurchases and growth in the future are likely to actually add value? It's very easy in such situations to see that a little growth right now will encourage both more stock repurchases and more expansion and yet both of these may therefore be ill-timed because they lean into a boom. Looking backwards, we can value what the company did in terms of creating or destroying value through share repurchases and through organic growth. But we invest looking forward.
My biggest problem is that return on the next dollar of retained earnings is hard to know at most companies. I believe it becomes harder to know as a company grows faster than GDP. I know investors love industries that grow faster than GDP. I know they love sectors that grow faster than GDP. I know they love locations in cities and states that grow faster than GDP. I know they love a business with customers who are growing faster than the overall population.
I have never been able to translate that into a situation where I was confident that a business catering to fat, old, diabetics in Texas will actually grow faster than a business catering to 30 year olds in New Jersey. That would only be true if some government were issuing a monopoly on serving fat, old, diabetic Texans. In my experience, the reverse tends to be true. While no one has a monopoly on a static populations they do - once they are well established - have less to fear from hot growth hunters. They tend to be left alone. And if they tend to their niche very carefully they tend to do as well over the truly long term - like my 15 year look into the future - than the company that should have all the tailwinds.
Of course, like any Warren Buffett wannabe I dream of finding that business with sticky customers, a great brand name, etc. and all the demographics going for it.
Unfortunately, in those situations where I think I understand both that growth is likely and that there is some sort of durable competitive advantage, the market more than agrees with me.
See Copart (CPRT, Financial), Boston Beer (SAM), and Under Armour (UA) for details. I wouldn't call those fair prices.
Now see John Wiley (JW.A), Weight Watchers (WTW), and Dun & Bradstreet (DNB). I wouldn't call those growth business. One (Weight Watchers) is debatable.
Even in cases where I think there is potential for future growth - but it hasn't even shown up at all in the financial results yet - the market is pricing the stock too high for me. See DreamWorks (DWA). At $13 a share, you'd be getting in under book value. Nothing in their 5-year GAAP ROE results: 14% in 2008, 14% in 2009, 14% in 2010, 7% in 2011, LOSS in 2012 suggest this is a company investors would be eager to pay a big premium to book value for. But the stock - which I'd be eager to buy at $13 - is now trading around $19. The stock is trading around a 16 P/E of normalized earnings if earnings since it went public are basically normal. I think it has real growth opportunities.
If I was sure of that growth, maybe today's price would be a fine price to pay. That is usually the problem with trying to apply a quantitative approach to the right price and right growth rate problem. A good model is one that is theoretically correct, that you can understand and feel comfortable using, and that gets good practical results.
The problem with an investing model is that for investors - unlike economists - it has to be a model that tells you something different from the market. It is only in being right in a way that others are not right that you win in investing.
Bill Miller is the person I associate most with the idea that complete DCF type approaches and value investing can work well together. Basically, you can use a DCF to get an intrinsic value and then you can buy at a steep discount to that intrinsic value to get your margin of safety. I've never been able to use that approach. And I'm not sure it's a good one.
I think some sort of comparative approach makes more sense. An approach based on what P/E ratios, Shiller P/Es, Price-to-free-cash flow, EV/EBITDA, etc. the market and peers now trade at and what they are expected to grow the bottom line by over the next 7 years or 15 years or whatever makes sense to me. Likewise, a historical approached based on what P/E, Shiller P/E, Price-to-free-cash flow, EV/EBITDA the company, peers, the market, etc. has traded for in the past makes sense to me.
Always using both a "normal" P/E and a "normal" growth rate based on comparative and historical data as reference points seems logical. It seems like you should be able to look for companies where you are sure they will grow closer to 10% than 5% and then try to always buy them at 5% growth type prices and you should be able to find 5% growers at no-growth prices and buy those too. In other words, there should be a good way to know you can pay 8 times earnings for a stock because its future growth is really just as good as the market and the market often trades at 16 times earnings. And you should be able to pay 16 times earnings for a business you think will grow 10% a year rather than 5% a year like many 16 times earnings stocks.
But here is the real question: Can I buy Copart at a 24 P/E?
On a trailing 5-year average basis, there is no year in the last 10 - and the last 10 for cars on the road growth in the U.S. were as bad as ever - where Copart grew EPS by less than 10% annually. Let's assume I think they can grow EPS by 10% a year. A P/E of 24 is higher than an earnings yield of 4%. And a 4% earnings yield growth plus 10% is 14%. That's definitely an acceptable return for me personally given today's stock prices. At any point in time, I'll take 15%. I'd like 10% at a minimum. And, as we discussed, I think the next 15 years won't see much better than 5% to 8.5% returns in U.S. stocks. Personally, I've often given 7% a year returns as the high end of future returns expectations for the S&P 500. So, 14% a year certainly clears those hurdles.
You mentioned the retention rate approach. Copart - I'm using their Value Line page here - has essentially retained all earnings all the time since going public. Return on total capital has ranged from 11% to 19%. Return on equity has ranged from 11% to 33%. The two only decoupled when Copart bought back stock very recently. I believe they are likely to have more free cash flow than they can use in the future and therefore must decide on buybacks, dividends, etc. going forward. Previously, they could always buy enough small competitors out in the U.S. as a roll up. They and their biggest competitor now account for too much of the market (almost three-quarters) to make that approach a possibility in the U.S. any longer. Obviously, free cash flow has been rising while the number of potential competitors to buy out has been shrinking. They plan to expand internationally. There are hurdles to clear. But I expect U.S. companies will go from the U.S. market to other countries in this business. I think the reverse is unlikely. Someone is going to take the American business model in this industry from country to country around the globe over the next few decades. This is especially true because already a meaningful amount of sales are cross border. Before Copart and IAA rolled up competitors and went online there were very few cross state border sales in the U.S. Now, there are plenty of cross country sales. These are networks. So growth seems likely to be foreign. But I think it's still possible. This is on top of baseline growth equal to U.S. car population plus car price inflation (basically I view annual percentage nominal dollar growth in the salvage value of cars in the U.S. as a given for CPRT because of their competitive position).
Let's be pessimistic and exclude the recent ROE boost. This is due to leverage. For the record, Copart uses less leverage than most companies in a similar position would. They own a lot of their land. Because the core business is good and uses the land for its operations, this land should not be assigned its own value. However, it is worth noting that Copart operates with a ton less borrowed money relative to the collateral and stable cash flow they have then most of the retailers, restaurants, etc. value investors look at everyday. For these reasons, I think it shouldn't be hard for return on equity in the future to be as good or better than return on total capital has been in the past. This is also helped by goodwill. The likely leveraged return on tangible equity someone would get out of this business is certainly at least in the 11% to 20% range Copart has achieved on its total capital.
Let's take 15% as the midpoint. Let's assume Copart can return 15% on its equity and grow its book value by 10% a year through retained earnings. They will have to do something else with the rest of that money.
Earnings are now $1.43 a share. They will grow earnings by 10% a year. That means earnings will grow by 14 cents next year. If they earn 15% on incremental equity, they will need another 93 cents of book value ($0.14/0.15 = $0.93). That would leave 50 cents of earnings they don't need to retain. Copart doesn't pay a dividend. They can buyback stock. For simplicity here we'll assume they pay out all earnings not retained as a dividend. We would then have two series of cash flows over the next 15 years. One would go back into the business to increase earnings. The other would be paid out to shareholders. That would leave earnings and dividends looking something like this:
EPS
2013: $1.57
2014: $1.73
2015: $1.90
2016: $2.09
...
2028: $5.97
The stock at the end of 2028 - if it sold at 16 times earnings as a "normal" P/E for a stock - could then be sold for $95.52 a share.
Meanwhile, the dividend payments would look like this:
DPS
2013: $0.50
2014: $0.55
2015: $0.61
2016: $0.67
...
2028: $2.09
So, in 2028 we would own a stock earning $5.97 a share and paying $2.09 a share in dividends. We assume this stock would be trading at $95. This is equal to a P/E of 16 and a dividend yield of 2.2% at the time of our hypothetical 2028 sale of the stock.
If you plug just the dividend portion - the non-retained earnings - cash flow stream into a DCF using 10% growth over the next 15 years in the dividend, 6% terminal growth (basically nominal GDP), and use our high end for the S&P 500 return of an 8.3% as the discount rate you will get a value of $37.62 a share just on the dividend portion of Copart's hypothetical future. As I'm writing this, the stock is trading at $34.36 a share. So, we can just (a tad conservatively) cancel those numbers out. Copart is - under my 10% a year growth for the next 15 years, 15% ROE, 6% terminal growth, 8.3% discount rate, etc. - assumptions trading at the value of its future dividends alone. The stock is worth more than its future dividends. This is because the company retains earnings and the retained earnings add value because they are assumed to earn a higher return than we could earn on our dividends. Our estimate of a $95 market price for a stock earning $5.97 a share in 2028 supports this.
For the record, the DCF thinks the retained earnings stream is worth $107.60. That's higher than our 16 P/E in 2028 so I'll ignore the DCF just like I canceled out the dividend and the share price. In both cases, the DCF wants me to be a bit more aggressive. It's safe to ignore that.
So, the only question becomes how much profit is there in Copart? How willing should I be to trade $34.36 today for $95 in 2028?
On a CAGR basis this is 7% a year of pure profit. Simply put, the DCF thinks Copart will run 7% a year ahead of the market over the next 15 years.
I have to say that when we use my carefully cherry picked, oddly idiosyncratic estimates for Copart, the S&P 500, etc. I think the DCF looks pretty smart. Or at least it sounds a lot like me. I'm not sure that's the same thing. But, yes, I actually agree with the DCF here. The earnings Copart won't be able to find a place for inside the business are probably worth about what the stock trades for. The really speculative component - beyond the mere assumption of growth - is the issue of what growth and ROE will look like together. This is always the tricky part.
I'm not sure I can guarantee 10% growth at Copart. I think if we knocked the estimates down to 6% a year growth combined with 15% a year ROE, I'd be real comfortable Copart is worth at least that much. This is unusual in that I think there will be 6% growth, it will return 15%, etc. and downside surprises will be much less likely than with stocks generally. In other words, I see a wide moat around a good and growing business. There are very few companies aside from Copart where I could say that.
But can we quantify this? Can we then say 24 times earnings is a fine - even good - price to pay for Copart? Even using my confident estimates of 6% earnings growth forever and $6.65 of added equity needed for every $1 increase to earnings (in other words, a 15% marginal ROE), we'd get values for the stock using a DCF that are close to double today's price. Bill Miller would be fine with those. Would Warren Buffett?
What's interesting here - in a practical rather than theoretical sense - is that I'm really quite confident in that 6% growth number and 15% ROE number and I'm really quite confident in them both being perpetual figures. This is not just a thought experiment. I really think Copart can grow 6% a year and return 15% on equity literally forever.
What is a company like that worth?
I crunched the numbers separately on price appreciation and dividends over the next 15 years. In other words, what would just 15 years of price appreciation be worth in Copart if it grows 6% a year and ends up at 16 times earnings in 2028? You would have to buy the stock at $16.60 a share for that to make sense. (Again, we're using an 8.3% discount rate - my high end for the S&P 500). Meanwhile, the dividend stream on a company that pays 86 cents this year (that's what Copart would have left to distribute in earnings if it grew only 6% next year and earned a 15% return on incremental equity) and growing by 6% forever would be worth $39.63 a share.
In other words, the quality/growth part of the company - the growing free cash flow that can't be put back in the business - would be worth more than the stock trades for today ($39.63 vs. $34.35). However, the value part of the equation actually takes away value. On a valuation only basis - looking just for stock price appreciation - Copart is extraordinarily overvalued if it only grows 6% a year and you only hold the stock for 15 years. The price appreciation potential of a 6% a year grower over 15 years that already trades at 24 times earnings today is exactly what you'd expect - it's worse than the S&P 500.
This brings us to the Warren Buffett / Ben Graham paradox. Buying and selling Copart is no way to make money. But holding Copart is theoretically going to make you money. In fact, you could argue that Berkshire Hathaway could afford to pay the very high looking price of 24 times earnings and still make money in Copart. On the other hand, if investor sentiment sours on Copart just enough so the company is seen as an "average" stock it will have a 33% drop ahead of it. The company is overvalued if the future for the company is the same as for the average public company. It's only undervalued if we count the extent to which Copart's future creates more value than the future of other public companies.
I think it's a near certainty that will prove true. Copart's future will be more valuable than the S&P 500's future. So should we count it?
I’m not sure.
Warren Buffett can count quality to its fullest because Berkshire is never going to sell what it buys. I have a hard time recommending a stock you need to hold a long time to overcome the fancy price you pay upfront.
Why?
Because I know anyone who acts on that recommendation is unlikely to keep the stock forever.
It’s hard to lose money holding a quality company forever. But it’s harder to find investors willing to hold a company forever.
The numbers on quality add up. Paying up for quality – in the sense of a company that needs to retain less earnings to grow at an equal rate to other companies – makes every bit as much sense as paying up for growth.
The problem with both the growth approach and the quality approach for most investors is that it has short-term risks and long-term rewards. If you pay 24 times earnings for a stock – and you are right to do it – that stock could easily drop 33% if market sentiment on the stock becomes just “blah”. Not “ick” but “blah”.
Unless you are the kind of investor who will see the market say “blah” and still believe in the business – which really requires the same stomach as buying an “ick” stock does for the value investor – than I’m afraid quality investors will enter a stock with the best of intentions but exit it too soon.
I don’t think there is an intellectual cure for this problem. What you need is the right personality rather than the right rationality.
In that sense, sticking with a quality company for the long-term is not really that different from buying a value stock. The difficulty in value investing is often having the courage to make the initial buy. The difficulty in quality investing is having the patience not to sell.
Ask Geoff a Question
