Someone emailed me this question:
“How do you estimate the reasonable growth rate for a company? I believe one way is to see the penetration rate of the product or service in each country and where they are in the current cycle, also considering the population growth and real spend per household as you did for Hunter Douglas. What is the general way of thinking about the growth?”
I am not very good at estimating future growth. I try to err on the side of being conservative. I focus on simple, repeatable growth that seems backed up by the past record. For example, I was recently trying to figure out the future growth rate of Howden Joinery Group (LSE:HWDN, Financial). The company had – at the time – about 630 depots in the U.K. It also had about 20 depots in France. There was a depot or two in some other countries as well. I treated the depots outside the U.K. as irrelevant to this estimate of future growth because those depots had different economics than the core concept of the depots in the U.K. Howden is – at the depot level – essentially a local business. It has a local manager of each depot who is in charge of most of the important operational decisions.
So, Howden – although it does not sell to the public – can be treated as a single retail chain. The easiest companies in the world to value the growth of are those who do the same volume of business each year and simply raise their price. The second easiest companies in the world to value are those who have the same clients every year and do a different volume of business with those same customers. The third easiest companies to value are those who do a different volume of business each year with a different set of customers – but who simply replicate a local model over and over again. Basically, we are talking about a chain of retail stores or a chain of restaurants. It is important this chain be just one concept.
If I was looking a L Brands (LB, Financial), which owns both Victoria’s Secret and Bath & Body Works – I would need to consider the growth of each chain separately. This can get complicated even at relatively small companies. Urban Outfitters (URBN, Financial) is not some mega-cap stock. It does have offline and online operations however. It also has three important chains: Urban Outfitters, Anthropologie and Free People. Those are not the only concepts. Those are just the ones I might bother to model the growth prospects for. Since Urban Outfitters is a company with three or more different chains, I have to assess the growth of each. Yet, the market cap of the stock is under $3 billion.
Meanwhile, Cullen/Frost Bankers Inc. (CFR, Financial) is a single bank in a single state (Texas) that seeks deposits from two groups (Texan households and Texan businesses). Yet, Frost has a nearly $6 billion market cap. Although Frost is twice the size of Urban Outfitters, it is also simpler. It has one “concept” (Frost bank) in one state (Texas). There is no doubt Urban Outfitters – because it can quickly win over new customers by opening stores in different states and by selling online – can grow much faster than a business like Frost. It is easier for me to estimate the long-term growth rate of a company like Frost however. You can look at its past growth in market share (whether they have grown faster or slower than other banks in the cities they are in) and at the branch level. I try to keep things simple. Say I feel the nominal gross domestic product of Texas will grow no slower than about 6% a year. Say I also think the bank will not lose market share and might grow it. In that case, I would simply estimate Frost will grow deposits by 6% a year.
Let’s look at Howden. The company is effectively a chain. So I would break down growth by “concept” and by “geography.” In Howden’s case, this means its depots in the U.K. There is no other concept at the company. And there is no other geography that is particularly important. So I would simply set aside all the depots outside of the U.K. and focus on Howden depots in that country. Very often, the management of a company that runs a chain will tell you how many stores, branches, depots, restaurant locations and so on it eventually hopes to put in that country. This is the store count level at which a concept has “fully saturated” a country. Howden’s management has set that figure at 800 depots in the U.K. They had about 630 depots last year and have been opening new ones at a rate of about 30 per year.
I looked at that situation and said I could estimate the company’s growth from now until the point of “full saturation” in the U.K. market. I have no clue if Howden the corporation will ever come up with another concept. I also do not know if it will successfully export this concept to another country. Like I said, they have about 20 depots in France. The company has disclosed those depots are profitable at a local level. They are a lot less profitable, however, than its depots in the U.K. And – because of the higher cost of doing business in France – a chain of depots in France will never be as profitable (on a return on capital basis) as a chain of depots in the U.K. So I really do not know where Howden will grow outside the U.K., how big those growth opportunities are and – most importantly – how profitable that growth will be. For those reasons, I decided to limit my entire estimate of Howden’s growth to its U.K. depots and to make this calculation extend out to only the end of 2022.
By the start of 2023, Howden would – if it keeps opening 30 depots a year – reach the point where it has 800 depots in the U.K. Management has said they see room for 800 depots in that market. Obviously, they might one day change their mind and say the opportunity is bigger than they thought, but that is speculative. So let’s limit our estimate of future growth to a pace of 30 new depot openings a year from now through 2022. We will then cut off our estimate there and say the company is as big as it is ever going to be. That is a lie, but it is a useful one for the calculation we are going to do.
The growth in a chain has two parts: 1) new store openings as a percent of the existing store base and 2) the growth in same-store sales. The compound annual growth rate to take 630 stores up to 800 stores over six years is almost exactly 4% per year, so we start with 4% annual growth at Howden. That is how fast the company would grow its sales if it did not have any same-store sales growth.
Historically, Howden has had a lot of same-store sales growth. The figures for same depot revenue growth are roughly as follows: 6% (2006), 9% (2007), 3% (2008), -5% (2009), 4% (2010), 3% (2011), 2% (2012), 6% (2013), 11% (2014), 9% (2015) and 4% (2016). The company’s sales are cyclically most related to kitchen renovations in the U.K. The customers are small builders, so they are not usually building totally new construction. To put this in perspective, I think about three-quarters of Howden's customers buy $3,000 or less a month from the company. Howden is often the biggest supplier to these builders and, for some items, probably their only supplier. So these are not individually very big builders. They are getting credit extended to them by Howden to buy the stuff they need, install it and get paid by the customer. They then pay off their balance with Howden and repeat the process with the next job. So in terms of normalizing things we would need to consider: nominal GDP growth in the U.K. and anything that would cause big swings in kitchen renovations that are not also swings in nominal GDP.
Rising housing prices in parts of the United States can sometimes cause people to take out home equity loans and re-do their kitchens and bathrooms. Therefore, a housing bubble and loose credit can encourage renovation activity. We have a pretty good sample of time here (2006 to 2016) that includes some not very good periods in housing in the U.K. I am OK with this not being a cyclically misleading same-store sales record. Now, Howden depots take a while to mature. It is often the case with a chain that once a store matures, it no longer grows sales much at all. Howden says stores are unprofitable for the first one to two years and generally grow until they are seven years old. It does not actually say there is no growth after that time. The numbers seem to suggest there actually is growth even at truly mature depots. But for the purpose of estimating Howden’s future same depot growth, I did classify the depots by age.
The company gives the depot count from year to year. So you can divide the existing base of depots into age groups. Many years ago, Howden’s same-store growth benefited from the fact it was opening more stores each year relative to its existing store base than it does now. If you look at the record however, relatively young depots have accounted for about the same percentage of Howden’s total depot count from about 2012 to 2016. Same-store sales growth in those years was 2%, 6%, 11%, 9% and 4%. The U.K. population growth is about the same as U.S. population growth. In the short term, the U.K. might have some economic headwinds from leaving the European Union. In the long run, it is not clear why the U.K. and U.S. would have drastically different futures looking at the way their economies are set up, their population demographics and things like that. I would estimate that nominal GDP growth in the U.K. should not be much below 4% a year. Even relatively mature Howden locations may be able to roughly match the nominal GDP growth of their local area. Beyond that, we are only looking out six years. The way Howden is opening its depots, the proportion of new depots to old depots will only very gradually shrink over this period. As a result, I feel confident projecting same-store sales growth of 4% a year through 2022.
It is worth mentioning here what really matters is earnings growth – not sales growth. Part of what makes me confident enough to take an estimate of 4% same depot sales growth and apply it out through 2022 is profit at the depot level should actually outpace sales growth a little. You can see margin expansion in Howden’s past. Because the company is vertically integrated, adding new depots should – if they have exactly the same-store level economics as the old depots – cause margin expansion at the corporate level. Anyway, because of the possibility of margin expansion at the store level as same-store sales grow and the possibility of margin expansion at the corporate level as the same supply chain serves a larger store base – I am confident adding the same-store sales growth rate to the new store growth rate and assuming that profit will grow no slower than sales. For Howden, my estimate over the next six years is the company will grow the size of its depot network by 4% a year and each depot will grow sales by 4% a year. This will drive company-wide sales growth of 8% a year for the next six years. That means net income will not grow slower than 8% a year.
In addition, the company has enough free cash flow – at today’s stock price – to pay a dividend of about 2.5% a year and still have 1.5% of its market cap left over to buy back stock. As a result, I would actually expect Howden to grow sales per share by 4% (new depot growth), plus 4% same depot sales growth, plus 1.5% in share buyback rates to equal 9.5% a year. Again, because there are economies of scale in a vertically integrated store network operated in a single country, I would expect something like a 9.5% growth in sales per share to drive EPS growth of 10% or better.
The company also pays a 2.5% dividend yield, so I would expect EPS growth to be about 10% a year over the next six years while you also get a 2.5% dividend yield. Let’s round that down and call it a 12% annual return projection in the stock. That projection only works if the company’s price-earnings (P/E) multiple does not contract. The P/E right now is around 15, which is normal for a stock. The company has a pension deficit and leases its stores. It also has more than 200 million pounds ($251.3 million) in net cash, however, that I have excluded from consideration here. I do not expect the company to use that cash to pay dividends or buy back stock (it will do all that from the free cash flow it generates each year), but I do expect it to be enough to plug the pension deficit from time to time. The company could also borrow money if it really needed to. As a result, I think it is not aggressive here to treat market cap as if it were enterprise value. In other words, I am just going to ignore the fact Howden has some liabilities like a pension deficit – but I am also going to ignore the fact it has net cash when most companies as predictable as it is would actually carry some net debt. So I would expect the stock to still trade at a P/E of 15 at the start of 2023. Simply put, I would expect a roughly 12% annual return on the stock over the next six years.
I used Howden as my example here because I am currently considering buying the stock. I do not own it, but I expect I will at some point. It is unusual for me in the sense that most of the return on this stock will come from growth. If you look at my 12% annual return estimate over the next six years, a full two-thirds (8%) of that return comes from growth. Growth is always a speculation, so Howden gets 8% of its expected annual return from a speculation on my part.
I do not think it is an especially aggressive speculation considering the company’s past and management’s belief it can one day have 800 depots in the U.K. I am willing to speculate on this kind of growth. Note, however, the speculation is limited to one concept (that has already been successful year after year) and one country (the U.K.).
I have no interest in speculating at all on any possible value in other concepts or other countries. The downside of this for me is Howden will be all used up as an investment idea in just six years. It will not be a growth stock anymore. At least, it will not be a growth stock I would bet on anymore. I think the company would still – even if it stopped opening depots – deserve a P/E of around 15 in 2023. That is because the free cash flow yield plus the same-store sales growth should still add up to a satisfactory return at a P/E of 15. If it stopped opening stores, it would have close to 6% of its market cap (at a P/E of 15) to distribute in dividends or use as stock buybacks. When you combine that with even a low level (like 2%) of annual same-store sales growth, you get a decent return that would match the market. Therefore, I am not worried about multiple contraction.
When a value investor buys a growth stock, his biggest worry should be a contraction in the multiple. For example, if you have a company that is growing 10% a year now but will only be growing 4% a year in about five years, that stock’s P/E multiple could easily collapse from 30 to 13. In fact, that would be a fairly normal multiple contraction for a stock with growth that slowed that dramatically.
That is why I included a discussion of valuation here. It is not enough to know how much a company will grow sales over the next five years. You also need to have some idea how slowly it will be growing after that to have an idea of what the multiple contraction risk in the stock is. I am comfortable with that risk at Howden. As a result, I would be comfortable buying the stock and holding it for five to six years.
I should also mention my hurdle rate. I do not make investments unless I expect a 10% annual return over the time I hold the stock. I would rather hold cash than lock myself in something I expect will earn 9% a year or less. So Howden’s growth rate does not really need to be 4% new depot growth plus 4% same depot sales growth. It can fall a bit short of that – like 3% new depot growth and 3% same depot sales growth - and still make me about 10% a year when you add in dividends and buybacks. That is important. I would never invest in a company where a 1% change in my estimate of some key variable meant I should no longer buy the stock. Here, there is enough room for me to be wrong about some of the speculative estimate I am making and still do OK in the stock.
The most critical part though is, of course, the possibility of margin expansion. I have intentionally left any possible growth in earnings above the growth in sales (that is, an expansion in the profit margin) out of my math here. That is not because I did not think about it. I thought a lot about it, but I really thought of it as an added defense against something like a contraction in the P/E. I cannot model margin expansion well here. It is more complicated than the four variables I laid out , which include new depot growth, same depot sales growth, dividend yield and share buyback rate. I really do not want to speculate beyond those four things. Margin expansion might provide upside, but P/E contraction might threaten me with downside.
The biggest speculation here is the one I did not mention. My cash is in U.S. dollars. Howden’s stock trades in pounds. So if I hold Howden for five to six years, I am making a multiyear bet the pound will not fall relative to the dollar in a big enough way to matter to me. All I know right now is the pound is not overvalued versus the dollar. But that does not mean it will not get more undervalued over the next several years. Growth is not the only thing that is speculative in a stock. Making an investment in a stock that trades in a different currency is also a kind of speculation. We can never eliminate all speculation from our investments.
Disclosure: Long CFR.
Start a free 7-day trial of Premium Membership to GuruFocus.
