Investment PhilosophyJoel GreenblattSuperinvestorsWarren Buffett

Thoughts On Return On Capital And Greenblatt’s Magic Formula Part 2

In part 1 of this post, I mentioned I caught a video interview with Joel Greenblatt at Morningstar. In the video, Greenblatt talks about indexing, and things that are not necessarily interesting to me and my investment strategy, but he also had some brief comments on Return on Capital. In the last post, I discussed the basic method that Joel Greenblatt uses to define Return on Capital. I also discuss some of the fundamentals and the importance of this key business metric, so check out that post first, if you haven’t yet.

The interesting thing was when Greenblatt specifically said he looks to fill his portfolio with businesses that have historically produced 50% returns on capital.

Greenblatt uses the often repeated example from his Magic Formula book:

  • Suppose you have a gum shop that costs you $400,000 to buy the land, build the store, stock the shelves, etc…
  • Suppose this $400,000 investment gets you a pretax earnings of $200,000

So even though I’ve probably heard him reference this example 10 times or more (and I’ve also read the Magic Formula book, which is basic—not nearly as good as his Genius book, but still quite good), his comment where he actually quantified the amount of return that he looks for (50% ROIC) was something I hadn’t heard him say previously. This really isn’t groundbreaking… we all know that high ROIC is better than low ROIC, valuation and everything else being equal… but it got me thinking about the example from a business perspective.

That hypothetical gum shop is a very good business (50% return on capital). It takes $400k to build out a store that produces $200k in pretax earnings. In fact, it’s a business that we’d all probably enjoy owning. And over time, if the business can be scaled and replicated into more locations that produce similar returns, the owner of that gum shop business will likely become quite wealthy over time.

Pretty logical, right?

It’s just interesting because we often lose sight of what we are looking for as value investors… we are owners of businesses, and all else equal, we want good businesses over bad businesses (i.e Greenblatt says 50% ROIC is obviously much better than say 10% ROIC).

The 50% comment also got me thinking about a Munger comment I heard once where he basically said that over time, the owner of a business will likely see investment results that (over very long periods of time) begin to approximate the internal returns that the business produces. Now, it’s absolutely crucial to understand this in the context of valuation.  Much of the early shareholder returns depend on valuation (the price paid), but over time, this “gap” between valuation and quality closes.

Munger used an example: a shareholder of a business that produces 6% returns on capital over time cannot expect to see investment results much better than 6%. Conversely, a shareholder of a business that makes 18% returns on capital over time will likely see superb investment results that begin to approximate this result as time goes on.

Of course, the hard part is determining durability, competitive positioning, etc… and that’s why valuation (paying a low price) gives us an all important safety net that Graham taught us about. But the math is easy to follow, over time, a stable 18% ROIC over a multidecade period will dominate a stable 6% ROIC in terms of shareholder returns.

I once looked at numerous examples of this in real life, and it’s certainly true over very long periods (15-20 years+), but valuation is the most important factor (in my opinion) in the early years and even the first decade. So valuation must not be forgotten. And of course Greenblatt factors in valuation into his investment decisions. But over time, Munger is right… imagine buying Walmart in the 70’s, Fastenal in the 80’s, etc… You could have paid 50 times earnings, and your long term results (if you held from then until now) would likely come close to the returns on capital that those businesses have produced over time (near 20% over time).

Now I’m not interested in predicting the next Walmart (well I guess I’d be interested in predicting the next Walmart, but I don’t think I have the foresight to do that on a consistent basis), so I choose to place a heavy Ben Graham emphasis (valuation) on my investment ideas.

But within that context, it is interesting to think about looking for businesses that produce 50% or more returns on capital. Those are basically the “Gum shop” businesses that Greenblatt likes. They produce $200K or more of earnings for every $400K of capital invested… and for whatever reason, they are cheap at the moment. I took a look at Greenblatt’s latest 13-F, and indeed, there are a lot of businesses that are producing very high returns on capital (but note: his new investment approach is not very helpful to follow, because he owns so many businesses, but it would be interesting to use it as a list to read annual reports from).

Another idea is to build a very simple screen that looks for businesses with high returns on capital, or alternatively high ROA or high ROE without much leverage. I wrote a post that summarizes one such screen I like to occasionally glance at from time to time. 

My hunch is that one could do better by using these concepts and studying the businesses, focusing on the more predictable businesses that can be easily understood. But it’s worth noting and respecting the results of the Magic Formula over time:

  • 23.8% annualized returns from 1988-2009 (according to the book)

Last year, I started tracking two “mock” magic formula portfolios (both have 30 stocks that I take from Greenblatt’s Magic Formula on 1/1/13—one is composed of stocks with market caps over $250 million, and the other has market caps over $2 billion) and both have done incredibly well, although one year isn’t really a good indicator in my mind.

So who wouldn’t want to use this formula to invest? Well, although I think over time the results will be difficult to beat, I feel that risk management (i.e. understanding what I own and quantifying the downside with each stock) is the most important aspect of my investment philosophy. I find it hard to do that with a computer screen telling me what to buy, so I don’t personally invest using this method. This is despite the fact I agree with the system’s concept (cheap and good). I just don’t feel comfortable allocating capital to businesses that I don’t understand and haven’t researched.

But to each his own, and hopefully for Greenblatt, he’ll be able to do very well using his mechanical stock picking system over time.

For me, I’ll take his concepts (basically Graham, Buffett, and many other value investors) and keep plugging along looking for low risk 50 cent dollars.

By the way, in the last post I mentioned an  interesting study that Buffett discussed in the 1980’s about high return on capital businesses, and I’ll have to delay that for a post some other time since this post is getting long (what else is new?)

In media, they call that a teaser. I’m sure the anticipation is unbearable…

26 thoughts on “Thoughts On Return On Capital And Greenblatt’s Magic Formula Part 2

  1. Great post! Very thought provoking.

    Like yourself, I believe Joel Greenblatt method of selecting stocks out of his screen makes logical sense but, clearly, not for everyone. His strategy does well because it can have 2 years of what is considered poor performance, to then have stellar returns thereafter (though the computer is not emotional, people are). Many people think they can withstand price volatility but, in all truthfulness, many people can’t ,especially if the do not understand the business or story. Therefore, I will have to stick to what I can quantitatively prove and qualitatively understand. It helps me sleep better at night or when I have a lot of red on my screen.

    Disclamer: I have glanced at the magic formula list from time to time and have bought 2 stocks in the last 3 years because, after doing the due diligence, I felt it was a good business with strong normalized operating metrics that was being severely undervalued. I have exited those positions at +50% in 2 years. Clearly, Joel Greenblatt is a smart guy and realizes the computer does not have emotion and sticks to a strategy that is proven and actually works. You just have to stick with the plan and the price volatility that may arise.

  2. What I find fascinating is Greenblatt’s disconnect. The gum shop example isn’t fabricated. If you trawl the local businesses for sale it’s not uncommon to see a business going for 2x earnings, or 2x ROIC. Some sell for even less. The disconnect is that he then extrapolates this to public companies.

    The problem is the gum shop company isn’t scalable. You might build the first gum shop for $400k because you know the area and can get a good deal on a store front. The second location might cost $600k because you don’t know the area well, and it’s a pricier part of town. Eventually when you have a well known brand you can’t buy locations for $400k, now they’re $1.2m. Plus there is overhead to take care of those stores, distribution centers etc.

    A 50% or 100% or 300% ROIC business is very straight forward if we’re talking small and private. But hard if not impossible to find if public.

    Munger’s comments are accurate if you hold a stock forever. But are irrelevant if one is focused on buying and selling when issues are under priced or fully valued.

    The irony of the whole situation isn’t lost on me either. Greenblatt is building his $400k gum shop. He’s built a very profitable franchise with very little invested capital. He has created his example and brought it to life. Unfortunately shareholders in his funds will never see those returns.

    1. Good points Nate, as usual. And the asset management point is good too. Very high ROIC, very scalable, and virtually no capital requirements to grow–in some ways even more attractive that the gum shop.

      As for Greenblatt’s new strategy, I’ve always been confounded by his decision. The only thing I can think of is to your point… he became less interested in absolute performance and more interested in gathering and growing assets… not necessarily bad since he might be able to help more smaller individual investors with his mutual funds than he could in his hedge fund. But certainly his results won’t come close to his old fund, and he seems to be fine with that.

      And regarding performance, I made this comment in a previous post, but basically… it seems that computer systems often times perform great on paper and in backtests, but for some reason those returns can’t be realized in real life. I’m not sure why… some of its psychology I suppose, but there seems to be a reason that the actual results fall far short of the backtested performance–and by more than could be attributed to slippage, commissions, etc…

      At some point, I’m going to write to Greenblatt and ask him about his huge strategy shift. Doubtful I’ll get a response, but I’d love to ask him some details about his thought process. I really think it will be difficult for him to add much value owning 100 or 200 stocks in his portfolio. Regardless of the metrics, it is so hard to beat the market by any material margin when you have that many stocks, especially larger stocks like those that are in his portfolio.

      As for the local business example, I think he probably used the gum shop for simplicity purposes so his readers would grasp the concept easily. But yeah, it’s hard to find businesses that are scalable that produce those types of pretax returns on capital. There are some though. I keep a spreadsheet of businesses that have produced 50% ROIC recently. To your point, some of them are not scalable, but some of them are… Most of them trade at high multiples, so they aren’t really exciting ideas, but they are businesses I like to follow. Glancing through the list, there are beverage companies, some retailers, service businesses, a couple pizza franchises…. mostly businesses that require very little capital.

      One business on the list is an example of Greenblatt’s gum shop: Francesca’s is a retailer that operates a chain of boutique stores that sell a mix of different things. It might be a fad type business, but so far, it has done extremely well. The stores have grown from 111 to 360 (as of last 10-K I checked) in the past 4 years. They are small stores, but produce excellent sales per square foot of $632.

      The economics of the boutique model are very attractive… consider this: Each store, averaging just 1385 sq ft, costs an avg of $186,000 to build out with equipment and fixtures, and an additional $45,000 to stock with inventory. These are net costs after tenant allowances. So each store requires roughly $231,000 of invested capital. In the first year, the stores average $750K of sales and a pretax cash return on investment of over 150%. So each store produces an avg pretax cash flow of around $350,000 or so.

      Using Greenblatt’s method to calculate ROIC for the firm, I come up with 68% returns.

      So FRAN kind of reminds me of Greenblatt’s gum shop, only it has proven to be quite scalable.

      I think it’s a nice business with excellent economics, high margins, etc… the question of course is whether it’s undervalued, and I’m not as certain about that at this point because I’m always somewhat skeptical of these retailers’ long term staying power. But given a cheap enough price, this might be interesting.

      In the end, I think you’re right… Greenblatt’s new strategy will produce mediocre results in real life. I’ve said before that after studying investors with excellent long term returns, I’ve noticed two main ideas: one is “Original Greenblatt” type concentration on cheap/good businesses or special situations, and the other is a Graham/Schloss approach that you follow that owns a basket of truly cheap stocks. The “in-between” is what most investors follow… aka buying too many average businesses at 10 times earnings, which leads to average results over time.

      Anyhow, thanks for the thought provoking comment.

      1. I am a little confused by the end of your response to Nate where you say Greenblatt’s “new” strategy will produce “mediocre” results in real life and yet in your post you state:

        “it’s worth noting and respecting the results of the Magic Formula over time:

        •23.8% annualized returns from 1988-2009 (according to the book)
        Last year, I started tracking two “mock” magic formula portfolios (both have 30 stocks that I take from Greenblatt’s Magic Formula on 1/1/13—one is composed of stocks with market caps over $250 million, and the other has market caps over $2 billion) and both have done incredibly well…”
        So who wouldn’t want to use this formula to invest? … I think over time the results will be difficult to beat…”

        I assume the “new” strategy you are referring to in your response to Nate is the “Magic Formula” ? Please clarify.

        1. It’s a good question Dave, and I probably made that confusing with my comments. Basically, I think one could systematically use Greenblatt’s method of picking 30 stocks from his “formula” and likely do quite well over time (most likely beat the market). I don’t think this system will produce 23% annual returns as it has in the backtest… at least they haven’t in the separate accounts so far. Also, Greenblatt’s mutual funds that use his formula own hundreds of stocks, which will make it difficult to beat the market by much more than a few percentage points.

          1. And yes, the “new” strategy is the magic formula. The old strategy is his concentrated approach to investing, where he typically owned 5-8 stocks.

      2. John,
        Thank you for the thoughtful information that you produce. Your website is like an archive of all the articles, books, etc. that I have on my computer, on my bookshelf and in my file cabinets. I have tried to keep it organized, where it is readily available for reference, so is marvelous in this regard, as well in all regards. I thoroughly enjoy all of your post and your investment methods mirror mine in almost every way…probably because your reading list also mirrors mine..

        You mentioned you’ve noticed two main ideas from studying investors with excellent long term…I would add a third, excluding Graham/Schloss, and that is really enlarging the size of the bet when the odds are heavily in your favor.

        1. Thanks for the nice words Randall, and yes, great point on position sizing when the probability is heavily in your favor.

  3. Hi John,

    Thanks for another great post! I am learning a lot from your blog. So, please keep up the great work for all of us.

    I had a question regarding the ROIC screener to check companies that have been producing 50% ROIC over a period of several years. Is there perhaps a site that offers such screeners for free? I know it is cheap to ask, but I just started and I prefer to keep my costs under control.

    Thanks a lot!

    1. Hi Joris, thanks for the comment. I like to use Morningstar for my screener. I don’t screen specifically for Greenblatt’s method of calculating ROIC, but I use simple ROA. From there, you can manually look at the businesses and make adjustments as needed. Basically, the idea is to look for businesses generating high returns, and ROA will generally achieve that objective. You can also use ROE, but that is much more subject to the leverage a business uses, so I would use ROA and then examine each situation. I referenced a screen I put together that looked at companies that produced 20% ROA. That is a good place to start. Most free screeners (Morningstar has a free version) have this. I’ve also used Gurufocus, which has a nice screener as well. I like Morningstar because you can search the history going back 5 to 10 years (10 years is a paid service, 5 I think is free). Also, I think Gurufocus actually has a Greenblatt ROIC screener, but I prefer to simply screen for ROA and make my own adjustments.

      Any screener using ROA over 20% will produce a list of businesses that certainly have generated above average returns on their capital. But with any screener, you need to be careful because there are always manual adjustments and analysis you’ll want to make…

      Thanks for reading!

      1. Gurufocus does have an ROC (Greenblatt) screen, but it requires a premium membership to use. They do list the Greenblatt ROC under the 10 Year financials for free, though.

      2. Hi John! Thanks a lot for your reply! I understand and actually agree most of your concepts, our team claim to do more or less the same thing. So now is more down to apply and real weight you put on each factors in the final decision, that really make a lot of difference though everyone claim to do the same, in my eyes. And too often it is subject to judgement. I’m really glad that I found your blog. There are lots of questions I may have a chance to gain more insight through this blog.
        Now starting with another small technique question here. In the part 1, you write Greenblatt’s formula to calculate return on Capital (I have read his book too), EBIT as numerator, net WC+ fixed assets as denominator. I thought you imply this is a most appropriate one. Now here, you says you use ROA, so it is net income divided by total assets. I’m wondering why it’s ROA, which seems really not refined parameter? No adjustment tax rate and capital structure. Especially total assets, it is such a “rough” accounting number. It is far from reflect real capital needs to invest as you repeatly mentioned in the part 1.

        And the reason I’m going after all this details, it is because it does make a difference. For example, assuming here is a soft drink company, goodwill around 300m, adding some other franchise rights, bottling agreement, total intangibles goes up to 400m, PPE roughly 250m, net working capital negative, so we count it 0, it earns EBIT 125m, if we use Greenblatt formula, it is ideally 50% return! However, see its balance sheet, it has 100m cash, 400m current assets (inventories and receivables), 70m financial assets, Then ROA only 10%, far from your 20% screen filter.
        So I want to know what’s your opinion? What do you really do in your process?

        1. Hi Nina,

          Yes, you are correct. The numbers will certainly be different. But I wouldn’t get too caught up in the exact calculations. Each business is different, and needs to be analyzed on its own.

          Regarding ROA, I was really just suggesting it if you have to use a screener and you’re interested in finding high return businesses. The objective is to find businesses that produce high returns on capital, and in most cases, depending on leverage and working capital, these businesses will produce high returns on its assets and equity. So ROA can be a quick way to identify higher quality businesses if you have to use a screener.

          I don’t personally look for businesses this way very often though. But I think it would certainly identify some high quality firms if you look at all the businesses available that have made 20% ROA recently, or better yet, over a period of time. (As you say, you’ll have to make adjustments here and there, depending on the specific situation).

          The basic principle is what’s important here. I do think Greenblatt’s method is very good as a broad based measure to use across capital structures and timeframes, but I wouldn’t be too rigid with it. Greenblatt himself makes various adjustments from business to business in his own portfolio and calculations.

          Think big picture, and think like a business owner. As an owner, you’d ideally want a business where you can reinvest a lot of earnings back into the business, and those invested earnings yield a high return on investment. Different businesses would need to be analyzed in different ways, but that is the concept that you’re trying to get to the bottom of.

          Thanks for the question Nina!

  4. Does Greenblatt prefer high ROIC companies with significant reinvestment opportunities versus companies in fully saturated situations? I have been grappling with whether or not it is more attractive to own a business that has a fully developed footprint and is taking excess cash to buy in shares versus spending capital to further build out its footprint. All else being equal, shareholders prefer to invest in a growing platform, as economies of scale should cement competitive advantages and EBIT margin expansion over time. That said, growth is pricey in public markets, often costing extra multiple turns. I’d be curious if Greenblatt has a view on buying excellent businesses trading at reasonable multiples with management teams using FCF to buy in shares versus similarly high ROIC businesses in expansion mode, without near-term buyback capacity and trading with higher multiples.

    John – Thanks for the perspective. Well done.

    1. Good thoughts Danny. My guess is Greenblatt would prefer–valuation aside–the business with 30%+ ROIC that has ample reinvestment opportunities and a long runway. That type of reinvestment opportunity will likely create more value than buybacks, and certainly more than dividends (unless the owner who receives dividends can match that type of return). Buffett mentioned that in the 1992 letter as well, and I’m assuming Greenblatt would echo Buffett’s view that he articulated in 1992:

      “Leaving the question of price aside, the best business to own is one that over an extended period can employ large amounts of incremental capital at very high rates of return. The worst business to own is one that must, or will, do the opposite – that is, consistently employ ever-greater amounts of capital at very low rates of return. Unfortunately, the first type of business is very hard to find: Most high-return businesses need relatively little capital. Shareholders of such a business usually will benefit if it pays out most of its earnings in dividends or makes significant stock repurchases.”

      So my thoughts are Greenblatt would echo Buffett’s thoughts that the best thing is to find a compounder that produces very high returns that can reinvest at similarly high rates, but since those are so rare, the next best thing is a compounder that–while it produces more capital than it can reinvest–still compounds value through buying back shares… which indirectly provides that capital with access to the same economics. I think Munger once said that he didn’t want to franchise See’s Candies because it was already “drowning in capital”. Sees might be an example of the latter where the business produces incredibly high returns, but can’t reinvest all of the capital it produces.

      Still not a bad investment option, but if Sees could reinvest all of the capital as opposed to just some of it at those high returns it has produced, the business certainly would be much larger and much more valuable.

      It’s an interesting question.

  5. Hi John! In 2nd paragraph, you mentioned, Greenblatt said he looks to fill his portfolio with business that have historically produced 50% returns on capital. Here historically, usually, in the textbook or our practice, it’s 5 years. I’m just wondering how many years / how long backwards do you look at, 5 or 10? (I understand, it might be case by case, some companies might restructured, or new management… etc. just generally speaking, how long would you be interested to look back?
    Meanwhile, a related question, now when you assess historical financial performantce, how many years do you look at? If again, as usual we use 5 years backward period, it mostly begins from 2008/2009 which is the bottom for most companies, then trend most often looks good.
    Thanks a lot if you could reply!

    1. Hi Nina, I always try to go back as far as I can to get an idea of how the business performed in various cycles. I try to look back at least 10-15 years if possible. And it’s always good to look at how companies performed during recessionary periods or in times of general economic headwinds/high unemployment, etc… I like to find businesses that are very durable and are able to sustain profitability even in the tough times. 2008-2009 is a great crucible!

      1. Thanks a lot! Indeed.
        Somewhere, yours blog or someone else…, I read and totally agree that, there is difference between typical investor and CEO mindset. Buffett said? “I’m a better investor because I’m a business man; and I ‘m a better businessman because I’m an investor”. It’s better to meet both mindset.

  6. Hi John,

    Thanks for the great post. I enjoyed reading your articles. Your posts are well written and thought provoking. I find that the part 1 of this topic is not longer available. Can you email you the valid link or the PDF of your post on the part 1?

    Thanks………Chee Wai

  7. Do use ROIC or ROC? ROIC seems more useful to me but I have issues with a lot of companies that have negative working capital. I personally use ROE but am a huge Greenwald fan and would like to start screening by ROC.

    PS. Great blog!

    1. Thanks Jamie. ROC and ROIC are the same thing–return on capital (or at least I use them interchangeably… there seems to be a few different ways people like to define that). But all of those metrics (ROA, ROE, ROIC) are basically there to tell you how effective a company is at wringing out profits from the capital it has at its disposal. You need to adjust for leverage and other factors, and each business is different, but any of those metrics are useful to consider.

  8. Hi John,

    Part 1 of this post seems to have disappeared. All the links are dead. Would love to follow the full write-up.


Leave a Reply to Nina Cancel reply

Your email address will not be published. Required fields are marked *