General ThoughtsJoel Greenblatt

“Magic Formula” and Other Quantitative Results from 2013–Should the Computers Takeover?

Let me say that the following few sentences only represent my humble opinion. There are many smart, talented, and successful practitioners that participate in the field that I am about to comment on…

(Editor side note: unlike many who take pride in saying things such as “I tell it like it is”, or “I don’t care what others think”… I happen to dislike offending others and I do care what others think, but evidently I’m in the minority here. I certainly have my opinions, I’m not a fan of “political correctness” or anything like that, and I enjoy debates, but I don’t care about drawing attention to things that I disagree with when the sole purpose is to generate an argument.)

Rule #1: Don’t Lose Money

Having attempted to soften the blow from a very passionate crowd (many of whom I’m friends with), let me say that I’ve never been a big fan of letting a computer screen or “magic formula” tell me what stocks I should own. I think of risk management as the most important area of investing, and in my opinion, I feel that I would be taking on far too much risk to let a computer tell me what businesses I should own. For some reason, the stock market has always lent itself to this type of behavior. Probably because of all of the data, liquidity, and easily ascertainable financials. Buying and selling things based solely on mechanical formulas doesn’t really occur in private equity or business in general. I’ve never come across a successful business mogul who every December 31st liquidates all of his businesses, and then on January 1st acquires a whole new list of businesses that happen to be available at some low multiple of earnings, or worse yet–because those businesses happen to have gotten more expensive to buy in recent months (i.e. momentum).

It’s comical to imagine such a scenario. Imagine Sam Zell liquidating his real estate holdings at year end, and then the next day arbitrarily buying equal stakes in the 30 buildings in Chicago that trade at the highest cap rates (lowest P/E, or to be precise–EV/EBIT), and repeating that process each year. Imagine how shareholders would react if John Malone–instead of buying a stake in Charter and going after Time Warner Cable–abandoned his circle of competence and announced that Liberty Media would now be focusing on buying and selling each year the 50 stocks with the lowest price to book ratios because that strategy has performed quite well in a backtest.

His shareholders would likely revolt–and not because he chose to focus on low P/B instead of low P/E, EV/EBIT, or P/S. They would be upset because it wouldn’t make much sense. They would collectively feel that it doesn’t make much sense to abandon experience, knowledge, understanding, value principles, and common sense decision making–ingredients that resulted in outstanding shareholder results (Liberty Media is Malone’s holding company whose former parent–TCI–produced multi-decade shareholder returns of around 30% per year with Malone at the helm).

So if this type of strategy is comical when it comes to business investing, why is it so commonly accepted and respected in stock market investing?

This is a question I’ve thought about a lot… again–I respect many of these so-called “quants”. Some of them, such as Joel Greenblatt–has indirectly taught me an incredible amount about investing and I will likely be permanently indebted to him. Greenblatt is interesting because he actually left a value investing strategy that was ultra successful (40% or so per year for 20 years) in favor of his “magic formula”. I was always amazed that he made the switch, and would love to know exactly why, but both of his systems seem to give him great success (although the results of his original strategy–by his own admission–will always be superior).

Why Not Follow the Quants? 

So most of these quants are extremely smart, and many have built thriving asset management firms with excellent results so far. And the tests are convincing… owning low P/E or low P/B works over time. Even momentum works… in most of the backtests it is basically as strong as value. So why not try to emulate the quants? Wouldn’t it be easier to just let a computer program tell you what stocks to buy at 5pm on December 31st, issue buy orders for January 2nd, then take the rest of the year off? Well… even the quants will tell you it’s not that easy. They work much harder than this. Jim Simons (one of the most successful hedge fund managers of all time and a genius mathematician turned quant hedge fund billionaire) once said that as soon as you develop a system, someone else is already working on the same system and so you always have to be one step ahead–thus his reasoning for hiring 100 or so of the smartest math minds he could find.

But what about “quant” combined with “value”? This doesn’t take 100 PhD’s… just some common sense principles, a computer screen, and a healthy dose of behavioral fortitude (patience) to stick with it. After all, we’re taking common sense principles and creating a structure that forces us to stick with it right? Greenblatt has often said that simple value investing strategies work because sometimes they don’t work. If you can just stick with it, you’ll beat the market, etc… Yes, I think that probably is true, but the problem is most people don’t stick with it.

I personally have spent a lot of time considering these types of strategies over the years. And I think I have the temperament to stick with it (who doesn’t think that, right?)… However, the issue that I could never get comfortable with is investing in a business (or basket of businesses) that I knew absolutely nothing about. The results look great when assembled in a table, but take a look at the most recent magic formula screen, there will likely be a number of businesses that you know nothing about and some you have never heard of. That is the biggest issue for me, and one that I could never rationalize.

I think that over time, most of these systems will end up doing quite well, but when it comes to my money and that of my investors, I prefer to stick to owning things I know something about. I will rely on the same value principles that Graham, Schloss, Buffett, Greenblatt espouse–the same basic principles the value quants use–but I’ll add a layer of common sense and understanding to the equation. This is far from a guarantee that my strategy will do better… in fact, many of the quant systems will likely prove a very tough test. And it’s amazing how well a simple system such as Greenblatt’s magic formula has done. But in the end, that’s all we’re really doing, right? We’re just trying to buy good businesses (high ROIC) at cheap prices (low P/E). So it shouldn’t be surprising that it works.

I guess what I’m doing is simply using the principles and just ensuring that I understand what businesses I own.

But I do believe it is less risky–and more common sensical–to understand something–albeit not everything–about the businesses you own. Adding common sense and logical decisions with tried and true value investing principles seems the best combination to me. 

But again, this is just my opinion.

Enough Theory–Let’s See Some Quant Results

Without further ado, and with the understanding that showing these outstanding results might seem contradictory, I’d like to show the results of some very simple value investing strategies from 2013. For fun, I keep track of a few simple value portfolios each year. These are just hypothetical portfolios, and I keep them more or less because I love data. This is the first year I actually kept them on a mock portfolio on Finviz. These are not portfolios that I manage–these are only hypothetical results that are gross of costs such as commissions, fees, slippage, taxes, and other transaction costs that one would incur in a real money portfolio. But nevertheless, I was astounded by how well some of these simple strategies did in 2013. Obviously it was a banner year for the market–best since the mid-90’s. But take a look at how some of Greenblatt’s ROIC (“magic formula”) strategies did:

Hypothetical Portfolios

 (disclosure: these are hypothetical results-not real money portfolios)

Again, these are just hypothetical passive portfolios that I track for fun. The magic formula had one of the best years in the history of the backtest going back to 1988 (although there were a few years that surprisingly surpassed this one). The index portfolios are the results of simply taking the S&P 500 and taking the cheapest decile (or the cheapest third of the Dow). The highlighted portfolio takes the cheapest S&P decile (50 stocks within the S&P 500 with the lowest P/E ratios) and then takes the 10 stocks from that group with the highest return on capital.

So it was a good year for quality and value as measured by low P/E and high ROIC. My Greenblatt hypothetical portfolios simply took the top 30 stocks in the screen at the beginning of the year. Here is a list of the 10 stocks within the S&P low P/E decile with the highest returns on capital (as measured by taking current assets plus net fixed assets less current liabilities) along with their 2013 total return including dividends:

Highest 10 ROIC Stocks 2013

Wow, what a list to throw a dart at in 2013… returns on each stock ranged from +15% to +127%. Unlikely to be matched anytime soon.

These are tough results to beat, and I think that if I kept track of these for the next 25 years, we might not see a year as good as this one for most of these portfolios. In the end, I enjoy checking these once a year–more just to see how “value” does vs. the market, and I think these portfolios would actually represent better than average results over time compared to the S&P. My objective is to do better than the average and better than these portfolios over time, but I refuse to waver from my focus on risk management, and that requires me to do some thinking. Some feel that because of their very nature, it might be less risky to follow these types of automated strategies. To each their own, and for those that follow these strategies, congrats–you certainly had a great year!

33 thoughts on ““Magic Formula” and Other Quantitative Results from 2013–Should the Computers Takeover?

    1. That’s a good point. I’ll take a look at your piece. Greenblatt wrote his third book that had some nice discussions of that topic, and also “value” weighted indices as well. Interesting…

  1. funny you bring this up, I was just thinking about this last night. the more I try to rationalize, why I shouldn’t simply follow a quant formula the more the facts say I am working to hard…

    let me ask you this, what do you consider getting to know a company? it seems to me, ( of course I could be wrong) that you find a ” small pond” to fish in and pick the best ones. what do you find that makes one company more attractive than the next?

    I am leaning towards following a quant strategy, with an occasional special situation ( or two) thrown in. of course keeping an eye out for a compounder that is going throw a rough patch

    1. Yeah I think that quant strategies that are based on simple, value principles such as Greenblatt’s, or even just simple low P/E or P/B will likely do quite well, and I’d certainly bet that they’ll beat the S&P if they are implemented in a disciplined manner. I just personally feel more comfortable knowing something about the companies I own. That’s not to say you have to be an expert in each field (someone will always know more than you about the business)… I’m just talking about understanding how the business operates, the business model, how they make money. If I can’t quickly describe their operations to someone who doesn’t know anything about stocks or business, then I’m usually not interested.

      But this is not to say that I think quant strategies won’t work… As I say, I actually think they will work well over time, but I just choose to take the principles of those strategies and add a few “manual filters”. We’ll see in 10 years how that works out… 🙂

      With investing, I think the first thing any investor has to think about is risk control, and the second thing is to feel comfortable with their own strategy. They have to understand it and feel comfortable implementing it, otherwise they won’t be able to stick with it. And there are many different ways to make it work over time…

      1. Also, as far as getting to know companies, I just like to read the 10-K’s and learn about how they make money. I don’t do much projecting or predicting, and I try to keep it simple. I like to look at how they use capital, how much capital investment do they need over time, how much value have they created over time, what kind of products or services do they sell… are they simple things? I like simple business models. I like the parking garages in prime locations in the center of cities. I like toll bridges with high traffic counts. And of course I like them at bargain prices…

  2. hi, excellent article! i like focusing on quants such as low pb. some of them can be great winners in the future! it is a contrarian process and it goes against human nature!

    1. Hi Lei. Yes I think you are correct and I think you’ll do quite well over time. That’s the Schloss way of investing, and although he read the 10-K’s and liked to buy simple things (as I do), I think you could actually quantify a low P/B strategy and make it work. Again… for me, I like to look at those cheap stocks and try to cherry pick the ones I understand. But as you say, buying a basket of net-nets or low P/B over time is like the insurance underwriting business model and I believe that over time you’ll do quite well with it.

      1. hi. john. i agree with you! i think the essence of value investing is to buy low and sell high. low pb or low pe can tell me those stocks price may be low. only what i do is to assure they can turn around. i always think it is hard for most investors to do because i can often see some opportunities.

  3. do you publish critical intelligent responses to your articles or just the ones that you like? good going with the censorship … keep up the intellectual honesty huber.

  4. Don’t see why you had to prevent publication of a comment that highlighted a weakness in high roic strategies that you seem to hold so dear … try and engage in intelligent discussion to arrive at the best investment strategies rather than focusing solely on broadcasting your own views.

    1. Hi Critics. Thanks for reading the post. I am not sure what you are implying though. I have never restricted a comment (other than spam) here at Base Hit Investing. The only comments that wouldn’t make it through are ones with vulgarity or other inappropriate language (and I’ve never had one of those-so I’ve never had to restrict a comment). I’ve always encouraged comments (feel free to be as critical as you’d like).

      If you tried to post something that didn’t make it through, it wasn’t because I screened it. Maybe try re-posting it.

      I would be happy to engage in a thoughtful discussion about ROIC or anything else related to the post.

      Thanks for reading.

  5. Hi John,

    I had a long post typed up but it got rambly and I was distracted by other work so instead I’ve decided on more of a “hit and run” reply. Here it goes:

    You said,
    But in the end, that’s all we’re really doing, right? We’re just trying to buy good businesses (high ROIC) at cheap prices (low P/E). So it shouldn’t be surprising that it works.

    I want to take the sentiment of this post to it’s logical conclusion (in my mind). The sentiment was, “Quant investing isn’t businesslike, you don’t even know these businesses and any actual businessman who bought and sold as a quant would, would scare the living daylights out of his investors.”

    The question I would ask myself here is why the conditions would exist in the first place that one could make a lot of money over long periods of time by simply passively investing in operating businesses they had no say or influence over. I am not asking, “Why does this strategy make money?” (A: “Arbitrage”), but rather, “Why is such an opportunity even available?”

    It is a weird state of affairs when one can engage in “business-like” investing and generate economic profit without actually making any of the individual contributions to the business that are historically considered the domain of the businessman (ie, managing inventories, developing processes, marketing to customers, choosing strategic vision… “going to work”, etc.)

    We call the few “business-like” investors who get so uppity “activists.” But an activist in any other area was simply a businessman who controlled what he owned.

    I’ll leave it at that, I have much more to say but you’d have to catch me over a whiskey or something for it to be coherent and worth hearing.

    1. Prax, long posts are always welcome. Thanks for the comment, and feel free to follow up with the remainder of your comments. Or if need be, next time you’re in North Carolina, let me know and I’ll buy you that drink. 🙂

      Yeah, I’d agree that in some ways it is unique that the opportunity persistently exists to passively create superior results over time. As others such as Greenblatt has said, I really think value investing will always have a permanent edge given the market structure that we participate in. In other words… “value investing works because sometimes it doesn’t work”. Human nature causes the opportunity to be ever present (over time). This is in part because of institutional behavior (what have you done for me lately?), an emphasis on short term time frames and near term outlooks, and just good old fashion fear and greed. All of these things tend to create “forced selling” or irrational selling of stocks for reasons that are independent of their intrinsic value. These reasons for selling will likely persist forever, as they are embedded into our collective psychology. Some might say they’ve actually increased with the increase in institutional and individual participation. Of course, sometimes the “spreads” narrow and other times they widen. But over time, there seems to be a consistent discount (price to value) that exists with this section of the market.

      I think that a value practitioner will do quite well over time by simply sticking to their guns and playing the insurance underwriting game with a basket of cheap stocks (some work, some don’t, but on balance, the whole thing works with the right underwriting guidelines). But as I say, my preference is to do some additional underwriting within the subset of ideas that are already part of that value universe. It’s basically the same concept, just different tactics.

      My short answer to your main question is that the opportunity exists simply because of human nature.

    2. John,

      I’ll probably have to take you up on that offer when I am in the neighborhood as I am not sure I have gotten my point across (or maybe I don’t have a good point, that’s possible, too).

      I understand the way human nature and mass psychology drive opportunities in the market. What I am getting at is what conditions create those opportunities themselves.

      People don’t make these kinds of mistakes, at least not routinely, in the private market. Just about everyone pays/receives full price for their ways. Why is there an opportunity to be a “business-like” investor who nevertheless does none of the things the typical businessman must do to eke out a profit actually operating a company?

      I’ll leave two hints as to my answer: taxes, and inflation (ie, changes in interest rates driven by changes in the supply of money and credit)

      We’ll discuss more in North Carolina one day!

      1. Sounds great. One quick thought… Buffett once commented on the topic that you are talking about. He said that he couldn’t have made 26% annual returns in the private market. It’s too efficient. Maybe the auction market has something to do with it. I personally think that the liquidity of the stock market actually creates more (not less) inefficiency. It’s certainly true that some illiquid small cap stocks can get very mispriced by the market, but so can a company as large as Coke. And I think that the liquidity of the market creates a lot of the inefficiencies I mentioned in my previous comment (i.e. institutional buying and selling for reasons unrelated to value… window dressing, forced selling, quarterly rebalancing, etc…) there are all kinds of reasons for buying and selling in the stock market that don’t exist in the private markets (or to a far lessor degree at least).


  6. Thanks for answering all my questions and for this blog.

    As I think more and more about diversification vs concentration. I’d feel more comfortable being an underwriter of some sort. Collect a bunch of investments and it would work out in average. The only problem is like you said that you have to concentrate on your best idea. I read a document saying that you get the statistical benefits of diversifying by owning 6-8 stocks. So a 10 stock portfolio is fine with me.

    As you mentioned you either need to concentrate or own a basket of crap. What are your thoughts on owning a basket of crap and similarly concentrating? I’m conflicted with the choice of concentration vs diversifying. I want to buy the cheap banks that’s making money, or a basket of cheap for profit education companies. Then I also want to concentrate on my very best ideas.

    Once again thanks for this blog and your insights.

    1. If I may answer, it really comes down to personal temperament.

      you can be fairly concentrated while still spreading your bets out to account for risk.

      also are you willing to own secondary stocks?? Many arnt willing to own a basket of net nets. ( or the like)

    2. My thoughts are always evolving on this topic of portfolio management, but basically, I think that most portfolios own far too many stocks. The gross benefit to adding say a 40th stock to a portfolio (assuming all roughly 2.5% weighting) is less than the hindrance that the new position adds (in other words, as you add more positions, they each become less and less beneficial in terms of diversification, and become more and more significant in terms of trying to create significant performance). Plus, it’s a lot more maintenance to keep up with 40-50 positions.

      Of course, there are basket approaches to buying cheap stocks where you’re making a statistical bet, and that reduces a lot of the work as well. So again, both strategies can work, but I think in terms of risk adjusted returns, if your goal is creating big returns over time, it’s going to be extremely difficult with a diversified (conventionally defined) portfolio.

      But this isn’t advice per se… you need to be very careful with a 10 stock portfolio. But certainly I’m more comfortable owning a small basket of undervalued stocks that I understand, and have both good quality and cheap prices.

      1. @Christmas: Position sizing is related to expected return and risk. The common formula to use is the Kelly criterion, since it gives optimal geometric growth. You can read about it on Wikipedia. But the summary is you can have fewer stocks if you have lower chance of losing money on any given bet, or higher expected value. These slides are pretty good:

        I believe there was a previous post on this blog about that also.

    1. Greenblatt ran a concentrated portfolio that had 80% of his assets in 5-8 stocks. He averaged 50% returns (before incentive fee) in his hedge fund from 1985-1994 when he closed to outside capital. He reportedly continued to do extremely well for another 10 years or so before he changed his strategy to a more diversified statistical approach based on his “magic formula”.

  7. Interesting discussion.It has never occurred to me to run my portfolio in “cruise control” using these methods. I have a clear understanding of why I manage my own portfolio…..1) To make money 2) To learn how world works 3) To have fun

    Following these automated formulas neither I will become wiser or have any fun….all I will have is a promise to become rich over time………so what if I have more or less money than following a formula when I retire…….its not the destination but the journey that counts.

    Maybe if making money is your sole purpose to investing in businesses then this is worth a debate.

  8. Thanks for your answer Jonathan. Yeah, I’m willing to own secondary stocks. Knowing thatowning a basket of net nets on average will give me superior returns is fine with me.

    I was considering concentrating on secondary stocks, although I fear that risk may be too large.

    1. as long as you invest with a margin of safety, and buy in bulk I see nothing wrong with that method. I personally have seen the good net net investing can have. my advice, trust the system, and focus on the process. you can amplify your returns by trying to focus on the “best” of the bunch. like shooting fish in a small pond

      goof luck

  9. I had some long posts I tried to post showing statistics and the rest but for some reason they are not going up, maybe they are to long…..

    I will test this one to see if it goes up or not.

  10. It’s also worth noting that it isn’t as simple as -> formula in -> results out.

    Greenblatt spent $20m on a team of analysts and programmers to massage the data for his Magic Formula database. They re-formatted all of the financial statements for companies in the US, reclassified certain types of assets and added/removed items from the income statement. There was quite a lot of human input to determine and derive the formula, as well as a lot of effort to normalize the statements.

    There is a reason Greenblatt’s results can’t be replicated, no one has access to his exact formula or his data other than him. We don’t know what adjustments he made.

    A quant investor is beholden to their data, if the data is bad the results are bad. They are also beholden to their formula, which is merely a consistent process.

    In the book The Value Investors Irving Kahn has a great quote regarding quant investing. He says that people have been telling him that computers are better investors since the 1970s, yet no one has programmed a computer to consistently make money since then. He asks if it has been so simple and easy why isn’t someone sitting and printing themselves money from a computer?


    1. Yeah it’s a good point Nate. I’ve thought that it’s interesting to note that Greenblatt has a much larger team of analysts now that he’s a “quant” than he did when he started his fund in 1985 (at that time was just him and one partner I believe).

  11. Hi John!

    I am a frequent reader of your website and really admire the simplicity and logic of your writings.

    As for me, I don’t feel my circle of competence goes beyond quantitative value investing That’s ok with me because I feel comfortable not ‘understanding’ the businesses I am investing in. A bit like a insurance company owner who feels comfortable not knowing the driving habits of each of his clients. The reason for that is that I trust the fact that my calculations tend to put the odds on my side on average, over many commitments.

    But to profit safely from any business where profit margins relies on statistics there are two things that are absolutely needed :

    1. Diversification though many commitments: maximum % of assets in one commitment at any time and a minimum number of new commitments made each year.

    2. Understanding why the odds are on your side on average: Conviction that our calculations put us at an advantage is key for long term success. Without solid conviction, chances are we’ll get discouraged by badluck or underperformance and sell our business at exactly the wrong time. The casino owner does not think about selling is business because someone won the mega jackpot yesterday. Bad luck happens and he knows it. He does not get depressed and blame himself by saying “I should not have let this smartass in my casino” or “I should have selled my casino two days ago”. I think that quantitative investors should act pretty much the same.
    Focus on logic, statistics, patience and diversification while putting emotions on the side.

    1. Nice comment Stephane. I’ve always thought that a Ben Graham style quantitative approach will work very well. I prefer to manually pick the stocks, but I think a pure quant system will work over time. My only addition to that is that I manually “underwrite” my insurance portfolio. But a pure numbers game will likely do well over time if it’s based on the value metrics that we’ve discussed. Diversification in this case is not necessarily to reduce risk (although that is one reason), but it is to increase the odds that the desired outcome is achieved.

      Ben Graham talks about this in the Intelligent Investor.

      Thanks for the comment…

Leave a Reply

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