Hilton Food (HFG): A Super Predictable Meat Packer with Long-Term “Cost Plus” Contracts and Extreme Customer Concentration at an Expensive – But Actually Not Quite Too Expensive – Price
Hilton Food Group (HFG) trades on the London Stock Exchange. It qualifies as an “overlooked stock” because it has low share turnover (17% per year) and a low beta (0.28) despite having a pretty high market cap (greater than $1 billion in USD terms).
On a purely statistical basis, Hilton Food is one of the most predictable – in fact, in one respect, literally THE most predictable – companies I’m aware of. There’s a reason for this I’ll get into in a second. But, first let me explain what I mean about the predictability here. Over the last 11-13 years, Hilton Food has shown very, very little variation in its operating margin. EBIT margin variation can be measured in terms of ranges (this would be 2.2% to 2.9% in the case of Hilton Food), standard deviations (this would be like 0.2% in the case of Hilton Food) or the coefficient of variation (0.08 in the case of Hilton Food). When talking about margin variation – I almost always am talking about this coefficient of variation, which is the standard deviation scaled to the mean. So, if a company had 27% EBIT margins on average and a 2% standard deviation or a 2.7% average margin and a 0.2% standard deviation – I’d talk about those two situations as if they were equally stable or unstable margins. Another way to look at it would be to think about standard deviations. If you own a stock for any meaningful length of time, you’re going to see one standard deviation and probably two standard deviation moves in margins. You may not see a three standard deviation move. And it’s entirely possible – unless something major changes with the business, which of course, it often does – you won’t see a 4 standard deviation move. With Hilton Food, a move of 4-5 standard deviations to the downside would only be a 1% of sales move in EBIT. Now, margins at Hilton Food are so low that 1% of sales is like 35% of EBIT and awfully close to 50% of earnings. So, it’s a big move. But, 4-5 standard deviation moves in margins are obviously unusual. And you’d be surprised how common 35% of EBIT to 50% of after-tax earnings swings are for many public companies. They happen all the time. I don’t want to go too far into the statistical weeds here – but, I will say that margin fluctuations that literally happen every 1-3 years for a normal company might only happen once in like 1-3 decades for Hilton Food. Now, Hilton Food has not been around for 3 decades (the original plant started operating in 1994 and the company only went public like 12 or so years ago). And what I’ve been discussing here is the predictability of earnings in cases where sales are fixed. Sales fluctuate. So, Hilton Food’s earnings will move around. It’s just that earnings will move around very, very little relative to sales compared to almost any other business on the planet.
For that reason, I’m just going to price Hilton Food off its sales. I think it makes no sense to pay attention to the “noise” of any fluctuations from year-to-year in margins. These are unlikely to last. It makes the most sense to just decide on the right price-to-sales ratio (or EV/Sales – at Hilton Food these two are often close to the same thing) to pay for the stock. I’m going to completely ignore topics like price-to-free-cash-flow, EV/EBITDA, P/E, etc. I don’t believe any of these numbers are as good signals to buy, sell, or hold Hilton Food as the simple EV/Sales ratio. So, that’s the only ratio I’ll be using in this write-up.
Before even discussing the business Hilton Food is in – I’m going to talk multiples. One easy way to value Hilton Food is to think of the stock as just providing a total return – for however long you own it – of dividend yield plus dividend growth. So, if you own the stock for 30 years and pay 50 times dividends (a 2% dividend yield this year) and the stock grows dividends per share by 6% a year for those 30 years, you’d have an 8% return over 30 years. This will only happen if you enter and exit the stock at the same multiple. Obviously, the influence of entry and exit multiples is amplified or muted depending on how short or long your holding period is. I don’t actually analyze a stock like Hilton Food on a 30-year holding period myself (I assume a 10-year holding period), but just saying “30 years” makes it a lot easier for you to see what a theoretically indefinite holding period would look like. Basically, a stock you hold forever that pays dividends to you will return dividend yield plus dividend growth.
Keeping that in mind, we can look at the long-term average EBIT margin for Hilton Food (about 2.7%) and apply the current tax rate in many of the countries Hilton Food is in (about 20%, but some of this may change a bit soon), and then round the resulting number (2.7 * 0.8 = 2.2%) down to just 2%. This makes my math very easy and it borders on slight conservatism. Once I’ve done that, I can find various equivalent levels between EV/Sales (which is how we’ll be pricing the stock today) and P/E ratios at no net cash or debt. I’ll be rounding these numbers. I’m doing all this to help frame the problem of how much is too much to pay for Hilton Food. I’m going to be talking EV/Sales now. But, I want you to see what this would require in terms of an exit P/E (I figure most readers are more comfortable working with P/E ratios of growth stocks than EV/Sales ratios) however many years down the line to leave the total return equation as simply dividend yield plus dividend growth.
Here’s the table…
EV/Sales of 0.6x = P/E of 30x
EV/Sales of 0.5x = P/E of 25x
EV/Sales of 0.4x = P/E of 20x
EV/Sales of 0.3x = P/E of 15x
EV/Sales of 0.2x = P/E of 10x
For me, some of these numbers are silly. We’ll get into the economics of Hilton Food soon – but, the business model is basically this: plant level unleveraged cash returns on equity of like 20-40% a year after-tax secured by 5-15 year “cost plus” type supply agreements with the biggest supermarket retailer of meat in a given Western European (U.K., Netherlands, Sweden, Denmark, Ireland, etc.) country. These are high returns on capital. These are long supply agreement. These are extremely fixed agreements. And these are the bluest of blue chip customers they are selling too. Retail packed meat is also a pretty stable demand sort of product – especially in richer countries. Households will switch from beef to lamb to chicken to pork depending on pricing differences in those commodities. They may trade down at economically stressed times a bit. This isn’t quite like supplying chewing gum or cola or toothpaste or tissues. But, it’s a lot more like supplying chewing gum, cola, toothpaste, or tissues than it’s like supplying cars, washing machines, or computers.
So, personally, I think the 10x P/E is silly. Hilton Food might trade at 10 times earnings (with no net debt) when everyone is worried that a major customer contract will be allowed to expire or that a major customer is rapidly losing market share. But, as long as Hilton Food is expected to grow sales over time – a P/E of 10 is not the level at which you’ll ever be selling this stock no matter when you get out. Personally, I also think an EXIT P/E of 30 is silly. But, let me explain this. I don’t think that the stock trading at 30 times earnings today would be silly. I think that’d make perfect sense. There is some Australian business just now coming online. And the company has shown a very good history of growth even without a big new expansion like this Australian plant opening in Brisbane. For example, in every year since Hilton Food went public it has grown the physical weight of meat packed for customers. Sales have dipped a couple times in money terms. And earnings have dropped year over year at times too. But, not for long. And, actually, the real output of this company hasn’t declined since it went public. There are strong indications that the company’s “share of wallet” from its customers has grown, because we can sometimes check sales figures for the company to a specific customer, country, etc. versus supermarket market share data for that customer. Hilton sometimes confirms this fact too. For example, there’s no doubt that Hilton has grown its sales of red meat to Tesco faster than Tesco itself has grown red meat sales.
This leaves the reasonable exit P/E range for me as 15x to 25x. That would suggest that at an EV/Sales ratio of 0.4x to 0.6x, my return in Hilton Food will end up being just the current dividend yield plus the CAGR in dividends per share while I own the stock. The company currently trades a bit above 0.5x EV/Sales. However, by my math – which involves a ton of estimates and outright guesswork based on various triangulation of statements by management about how big the Brisbane plant is and how big I think the company’s existing plants are and so on – I wouldn’t be surprised if “sales capacity” or “already achievable earnings power” or whatever you want to call it is 10-15% higher than last year’s sales. In reality, I think the Australia and New Zealand business might be as much as 20-30% of what Hilton is now doing in Western Europe (Western Europe accounts for all of Hilton’s profit, so I’m ignoring other regions). However, this is a 50/50 joint venture. So, I’m cutting that number in half. The Brisbane plant is running already. New Zealand is under construction. The customer (Woolworths – one of the two biggest supermarkets in Australia) had been working with Hilton Food for over 5 years already and then agreed to hand over operational control of the joint venture to Hilton, shift production to a much bigger facility (cap-ex was probably like $100 million USD on this thing) and sign a 15-year supply agreement with Hilton. Also, the annual report specifically mentions that the retiring CEO was asked to stay on as executive chairman instead of non-executive chairman (which is the much more common U.K. practice) till the Australian operations had transitioned to this new setup. So, while I don’t usually like to put a lot of weight on company guidance, my near term predictions, “forward” earnings, etc. I really do believe here – and these numbers are my own made up figures, not management’s – that we need to price Hilton Food as if it’s already doing 10-15% more business than it did last year. This is a $100 million dedicated factory built for a customer they have a joint venture with and have been working with for several years already. It’s a 15-year supply agreement. This isn’t speculative. The business has already been won. We need to count it as part of future earnings power even though it didn’t show up in last year’s sales figures.
This doesn’t make a huge change to EV/Sales. It lowers today’s price to somewhere almost midway between 0.4x EV/Sales and 0.5x EV/Sales. To put it another way, that means our return in Hilton Food will be dividend yield plus dividend per share CAGR over the years we own the stock IF we sell out at a P/E multiple of something like 20-25x. A lot of value investors are reluctant to assume you can ever count on a sale price that high. So, they’d need to assume the return in this stock will be LOWER than dividend yield plus dividend CAGR.
For the rest of us, the total return equation is simple. Today’s dividend yield is 2%. So, we should only considering buying the stock to the extent that:
Hurdle Rate – 2% = Expected Growth Rate
Imagine your hurdle rate is a 10% annual return. Then, the equation becomes
10% – 2% = 8%
Do you expect Hilton Food to grow sales, earnings, and (thus) dividends per share by 8% or more for however long you own the stock? Yes. Then, you can buy it and expect 10%+ return. No. Then, you have to expect worse than 10% annual returns if you buy in at today’s price.
Which do I think is more likely? Dividend per share growth greater than or less than 8%?
Historically, the company’s lowest sales growth period ending today is if we look back 5-9 years. Measuring from 5-9 years ago to today (longer than 9 years and shorter than 5 years would both give higher CAGRs) would give us an 8% sales growth rate. However, there has sometimes been share dilution of a little over half a percent per year here. Let’s round that up and deduct it from sales growth to get sales per share growth of as low as 7%. Based on past experience, we wouldn’t expect Hilton Food to grow much less than 7% a year. So, we wouldn’t expect a total return much less than 2% plus 7% equals 9% a year. That might beat the market. But, there are risks. Of course, over time, I’d expect dividends (or buybacks if the company ever starts doing those in earnest) to grow faster than sales. This is because of ROE type considerations. If Hilton only grew 7% a year for the next 10 years, it’d get seriously overcapitalized and need to use the cash on something. As a result, it’d probably increase dividends faster than it grew sales. So, it’s possible you’d still get a 10% or better return in this stock even if sales per share only grew 7% a year.
A different approach would be to measure the longest CAGR period instead of the lowest. Going back like 12 years, sales have compounded at 10% a year. Assume as much as 1% dilution. That leaves 9% sales per share growth long-term at Hilton historically. This would give you an expected return of like 11% a year (2% dividend yield plus 9% dividend CAGR).
So, I’m coming up with expected returns – if there’s no multiple expansion or contraction while you own this stock – of like 9-11% a year based on Hilton’s past growth rates.
But, should we be using past growth rates?
That’s a tricky question here. Normally, I would always want to assume a lower growth rate over the next decade than a company had this past decade. That’s a basic “compounding gets harder as you get bigger” way of thinking about it. The weird thing here is that Hilton’s growth has obviously accelerated in the last few years, they obviously increased cap-ex, they keep talking about much faster growth for the medium term than they had in much of the past half decade to a decade. I don’t want to assume more growth over the next 10 years than they had over the last 10 years. But, I’m not as confident as I often am in assuming growth can’t be much faster.
There are geographic reasons for this too. Hilton Food has zero presence in the U.S. It just ramped up its presence in Australia in a big way. The truth is that all of Hilton’s earnings really come from like 5 European countries where it serves like 4 major supermarkets. One of Hilton’s customers does have a meaningful presence in the U.S. The U.S. is potentially a bigger market than the combined European countries they are in now. And I can think of a handful of U.S. supermarkets that would be viable targets for a company like Hilton. In addition to the national or nearly national chains familiar to most readers there are the leaders in specific states. I don’t know enough about the industry and Hilton’s competitors to know if it’d ever be possible for Hilton to win a contract like Publix in Florida or H-E-B in Texas or something like that. Yes, those are just two states out of 50. But, it’s easy to forget that the economy of Florida on its own or the economy of Texas on its own is roughly the same as the economy of Australia on its own. There are probably like 5 individual U.S. states that would be on par with medium sized developed national economies. And there are 45 other states that are on par with small developed economies. I have no idea if Hilton Food will ever break into countries like the U.S. or other continents besides Europe and Australia. But, I do know that Hilton Food does not yet have a presence in at least half – and probably a lot more than half – of what the addressable market for retail packed meat will be 10 or 20 years down the road. So, I can’t rule out future growth rates being similar to past growth rates.
We can also look at two other indicators of possible growth. Historical dividend growth was at its lowest over the past 6-10 years. It compounded at 10% a year. Over the time since Hilton Food started paying a dividend, its dividend per share CAGR has been 14% a year. So, past sales per share growth would suggest 7-9% a year type growth rates while past DIVIDEND per share growth would suggest more like 10-14% a year growth rates.
What about the bonus plan?
Hilton Food has an incentive plan tied to earnings per share growth. The number used is always the 3-year CAGR ending this year. The plan was changed so part of the incentive compensation is now tied to the stock’s return. Most is still tied directly to EPS growth. And historically it was all tied to EPS growth. My reading of the current incentive plan is that Hilton “expects” (that is, its goal is) a range of like 5-15% a year EPS growth. They’d have to do like 6% a year to get any incentive compensation from this plan. And they won’t get any added bonus from doing better than 15% a year. How does this compare to what the targets had been in the past?
It seems pretty much in line with their long-term goals. The company fiddles with this number a little, raising it a couple percent sometimes and dropping it a couple percent sometimes. But, the middle of the range is often around 10% give or take 2-3%.
Does the company usually hit the target?
It seems like an “honest” target to me. What I mean by this is that if we look at like 10 years, the company seems to hit more what I’d call the middle of this range around half the time and seriously miss the range (hitting the high and low edges) about a quarter of the time. This isn’t a situation where if the company sets a 5-15% range they are hitting 15% all the time. This is more like if they set a 5-15% range, you should take that as aiming for 10% a year as realistic based on what the board knows about the next 3 years. That’s how I’d take it.
So, the bonus plan would seem to suggest you can get 10% type returns in this stock with slightly below target growth of like 8% a year (quite near the bottom the EPS range – executives need 6% to get paid extra for the EPS growth).
However…
I am already counting on like 10-15% growth tied to Australia. And that’s part of what the board knows and expects. So, I wouldn’t put a lot of faith in a 3-year EPS CAGR target. You don’t want to go into a stock thinking as short-term as what EPS growth will be for just these next 3 years. You want to be thinking more like: will this company still be growing in 10 years?
I think a 10% type hurdle rate is totally hittable for Hilton Food stock given today’s price. You could buy it today and clear that hurdle based on like 8% growth in sales, earnings, dividends, etc. per share. I think the company could do that. This is a growth company. It may continue to grow at those rates for the next 10+ years. You might get a chance to sell out of the stock at a P/E of 25 or higher. Despite this looking like a very expensive stock – I think it can give you a double digit annual percentage return if you buy in today.
But, that’s just the return. What about the risk?
Hilton Food isn’t a government bond. It has risks.
The biggest risk is customer concentration. Customer concentration here is absolutely extreme. Here is my best guess of how much of Hilton’s business is done with each of its biggest customers – excluding Woolworths (which would appear in this list at some point, but not yet)…
Tesco: 55%
Ahold: 19%
Coop Danmark: 13%
ICA Gruppen: 6%
You should probably view this like a stock portfolio. I believe Hilton Food is so decentralized that each of these 4 customers are served on a dedicated basis as an individualized profit center run on a long-term basis (these are 5-15 year supply agreements) by a manager who is effectively like a CEO of that unit.
These are not new relationships. Hilton Food started out even more concentrated than it is to Tesco, because the company’s origin is a plant built in the mid-1990s to retail pack meat for Tesco.
In recent years, Hilton has branched out into seafood, vegetarian meals, sous vide, etc. My impression is that profit comes disproportionately from retail packing red meat. I don’t have any good insights into this. But, it’s just my impression that the highest value add least commodity thing you can do in this industry is providing like 100% of the retail packing of red meat for one supermarket.
Hilton Food does not break down contracts by customer, year they expire, whether they are done on a “packing rate” or “cost plus” basis etc. I believe the company tends to make something like the equivalent of 8 American cents per pound of meat it packs. I’m not confident in that number being stable at all from year-to-year. But, if you told me it was a lot lower than 4 American cents per pound or a lot higher than 16 American cents per pound – I’d be skeptical.
Compare this to the price of various store brand packaged meat cuts you can buy – and you can see this is potentially a good deal for the customer. The biggest cost in the entire chain of production is the commodity cost of the meat. After that, the biggest cost at the actual Hilton plant is labor. I think the labor component is greater than 3 times the amount retained as profit by Hilton (in other words, if they are marking 8 cents per pound – I think workers contributing to that cut of meat are making 25 cents per pound). From what I’ve seen, Hilton is a lot more aggressive in investing in labor saving automation (robots) than a lot of meat packers would be and certainly than the customer would be if running a plant themselves. I just can’t imagine many competitors running the plants in the U.K., Ireland, Netherlands, Australia, etc. (Central / Eastern Europe is different) with less labor than Hilton does. Also, a dedicated plant should have the lowest shipping costs of any solution I can think of. Any plant that serves multiple customers can’t have lower shipping costs than a plant that is only delivering direct to your distribution centers and your stores.
What do I think the customer cares about here?
My guess is that the customer wants continuously lower real costs per pound to the extent possible. I think this largely means removing more and more of the labor component of each pound of meat. I also think the customer cares about food safety, recalls, bad publicity, etc. (this is all sold under their store brand – nothing is Hilton branded). Again, I think reducing the labor component increases food safety. Robots don’t mind working in cold temperatures. Employees do. And employees touching meat can increase the risk of contamination. As far as sourcing the product – I believe the customer does this with Hilton or even directs Hilton how to source it. I don’t believe Hilton is making any final and independent decisions on where to get this meat from. I believe Hilton is packing meat where the supermarket has chosen or co-chosen the source of the meat.
There are risks here. And possible returns look more like 10% returns not 20% returns. But, it’s a very predictable company. And it’s very insulated from both competition and macroeconomic ups and downs. The stock doesn’t look cheap on any of the usual measures (P/E, P/B, etc.). Nonetheless, if this was a $10 billion company instead of a $1 billion company and someone offered to sell it to Berkshire Hathaway – I think Warren Buffett would definitely take that call. I don’t know if he’d buy it at today’s price. But, I’m sure he’d be very, very interested in this business model.
Geoff’s Initial Interest: 80%
Geoff’s Revisit Price: 600 pence (down 45%)