Sunday, January 11, 2015

A Brief Sketch of My Approach to Value Investing


I generally find it easier to identify securities that are likely to appreciate by 2x or 3x over 2 or 3 years than I do those that go up 50% within a year. I am not capable of doing especially well in either “event-driven, special situation” ideas or in “forever portfolio” ideas. 

So, for me, a good idea is a security:
  1. whose value will very likely not fall below the price one paid for it;
  2. that will likely appreciate in value at a pace substantially above one’s hurdle rate; 
  3. whose worth will very likely be recognized by the market within a reasonable  (two or three years) period because its merits will be plain to even the most casual observer;
  4. that does not conflict with a current holding; and
  5. that allows for regular evaluation of the investment’s premise
If the idea has all of these features -- safety, upside, robustness, and timeliness; more on these later -- I feel comfortable allocating a meaningful (10% to 25%) share of my funds to it.

How many good ideas one is able to find (or recognize) in any given year is of course a function of one’s level of effort, experience, “circle of competence”, account size, and so on. 

I look at maybe a hundred and fifty to two hundred ideas a year spending fifteen to twenty hours a week to do so. In about half of these ideas I can see how I could lose money. In approximately a third of them I can't see a reasonable upside. Of the remainder, most are, when push comes to shove, just a little too complicated for me to really get behind or just a little too abstract; I can see that they're probably safe and I can also see that they will probably perform well, but I'm not sure. Of the remainder -- we are in single digits now -- I eliminate those that conflict with a current holding and buy those that I think are the most time-certain. In this way I figure that I am able to find two or three good ideas in an average year.

So in my highly stylized view of how I’d like things things to look, I ought, at any one time, have six good ideas in my portfolio paying out more or less in alternate years. 

In the real world, of course, some ideas will work out better and sooner than one expects, and some worse or later, but the central tendency should approximate this outline.

If the security doesn’t look as promising as when one bought shares in it, one can sell it and move on. If its price falls in the interim, one has the choice of adding to it or of lowering one’s cost basis by trading around the volatility in the stock price. But, for me, the principal advantage of doing things in this way is that selectivity helps keep my standards reasonably high (and the error rate reasonably low).

[The 20% cash position is there for use during important market declines – it is important psychologically and, if invested in a reasonable manner during a market decline, should help keep one’s full cycle CAGR at an elevated level].

Cash gains are reinvested and therefore compounded. Mark-to-market losses should, in theory, not matter – if one invests in a discounted security it is, by definition, underestimated and anything could happen to its share price until one’s appointment with it two or three years hence. In practice it matters – psychologically. That’s why it is important to do the necessary work beforehand, to avoid blind coat-tailing, to not be seduced by potential upside alone, and to know the limits of one's technical and emotional competence.
This, then, is a kind of road map that I use in charting my investment strategy, such as it is, and every major decision that I make is subject to its constraints and possibilities.

If I lay out the positions that I've written about that have constituted at least 15% of my portfolio by cost – that is, my highest conviction positions  – they correspond to the outline as follows:

Note the color coding: I replaced CEGID with Northgate, Lamprell with Emeco, Lombard Risk with Future Bright -- in each case because I thought the newer security lower risk than the one being replaced. Therefore my investments in CEGID, Lamprell and Lombard Risk did not last long enough to either reap the rewards or suffer the indignity of losses at the appointed hour.

Spare a thought also for Consciencefood Holdings which was in spirit one of my highest conviction positions and would have been sized at 15% or 20% had I the opportunity to buy more shares. It was taken out by the owner-manager before I managed to do so.

This is how they would have played out :


If one is not serious about risk management this road map leads straight to the poorhouse. Investment returns are therefore meaningless without a consideration of the risks taken to achieve them.

There are, I think, four ways to manage risk: knowing what you own, applying a “margin of safety”, hedging as necessary, and diversification. The appropriate balance between them will vary from person to person and will depend also on the nature of the securities involved. 

I, for example, prefer a heavier emphasis on MOS than I do on broad diversification because I think that I can apply MOS correctly most (though certainly not all) of the time. Four to six more or less equal positions and twenty percent cash is my sweet spot. I hope to attain a level of competence that will allow me to one day hold 8 or 12 names on average without increasing my risk profile.


Safety first. Upside second. Timeliness third. 

I am not persuaded that “margin of safety” is simply an x% discount applied to either some metric derived from the financial statements. 

Instead, I try to imagine the worst case scenario that can reasonably come about for the business’ (or assets’) prospects over the relevant investment horizon. I then overlay the least attractive capital allocation decisions consistent with management’s prior behavior, and value that

That’s the margin of safety value at a particular point in time: provisional, inexact, contestable though not controversial, and sufficient for the time-frame that I'm concerned with -- two or three years.

Again, for context, here are my contemporaneous attempts at MOS assessments of my high conviction ideas. Two prefatory points first:

First, I am far more strident when talking to myself than I would ever be in a blog post and the tone of what follows reflects that I'm afraid. I also think now, in hindsight, that I was on at least a little less sure ground in sizing up the safety of some of these names than I imagined at the time.

Second, the premises vary according to the nature of the business but I am in each case looking for something that can be absolutely counted on to permit a net present value of cash payouts greater than the price paid for the security.  

An asset-light business that could cover its operating expenses from maintenance service contracts alone,  engaged in a line of work that is both non-discretionary and not subject to near-term technological obsolescence. Sixteen years of dividends are covered by cash already in the bank. Sustainable free cash flow is 105% of enterprise value and this at a low point for infrastructure spending in France.

Although dominant in France, it is also exposed to emerging markets where revenues are growing and margins are much fatter. Toll roads are everywhere a growth business. There is a verifiable backlog. There is customer concentration but the relationship between it and its customers seems as much partnership as client-supplier. 

So revenue probably won't fall by much. If it does fall, operating profit won't fall by much more than that. The only effect of a share price decline would be to increase the efficacy of the buyback program and therefore make the equity more valuable still. All things considered I find it hard to imagine how the net present value of the dividends received by GEA shareholders will not exceed 45/share. 

An unusually high and stable gross margin combined with accelerating sales growth through the 2007-9 period is the most obvious sign that economic conditions would have to get a great deal more desperate than they have recently been for this business to suffer even minor indignities along its long term growth trajectory. Hospitals would have to stop being built even in emerging markets. Replacement of installed flooring in preexisting hospitals would have to be deferred. 

A rock solid dividend policy means that all discretionary cash flow not invested for growth has been and would be paid as dividends. No growth therefore would mean a 9% inflation-protected dividend yield for a very long time, longer certainly than the time to recover the value paid for the shares. 

90% market share in interisland traffic means pricing power. Oil goes up, fares go up by more. Oil goes down, fares go down by less. Meaningful competition in this segment would require 100 additional daily flights and would therefore require additional airport capacity to be built. But the main airport has already just been extended by the state -- and partly paid for, and occupied by, Hawaiian. 

This interisland business generates $220 million in EBITDAR.  It's not particularly cyclical but give it a haircut anyway and call it $200m. 

Alaska Air Group is larger and better capitalized than Hawaiian, is structured in a similar way, and is also its direct competitor. It trades at 5x EBITDAR. If Hawaiian's West Coast and International businesses were to fail, it would make compelling economic sense for ALK to purchase Hawaiian. 

If ALK were to acquire Hawaiian at a full turn less than its own EBITDAR multiple, close down the West Coast and International operations, and liquidate the associated aircraft, the effective purchase price of the interisland franchise would be $700m or 4x FCF. Doing so would also lead to reduced capacity -- and therefore higher fares -- in the West Coast-to-Hawaii routes, allowing ALK two ways to benefit. 

So Hawaiian's "acquisition put", or low-end "private market value" is ~$17. 

.. to be continued in next post. I'll also talk about upside and timing, tie the three (safety, upside, timing) together, and relate them back to the road map idea. In the post after that, I'll give my current holdings the same treatment as I did the names that have played out and see how they stack up. I'll then say something about how and where I look for ideas. 
I don't want anyone to get the impression that my approach is appropriate for most people or even for most active investors. It is not. My last post in this "series" will be about that.


  1. I enjoyed this write up on HOS

  2. Red - Really enjoy these investing style write ups in addition to the specific investment analysis on a company. Looking forward to the next write up on where you think Texhong, Future Bright etc. stand. Also appreciate that you put it all out there and accept that there will be a few idiots out that go beyond trying to show an opposite hypothesis and stray into borderline insults. Here's hoping that 2015 brings you some good returns.

  3. Nice post, really enjoy reading some of these philosophy/strategy/methodology posts in addition to straight write-ups..looking fwd to more!

    On James Halstead, curious how you got the NPV of dividends of 300m? And what calculation you used to view it as a "9% inflation-protected dividend yield"?

  4. Thanks for reading, Barry.

    I'm not sure I understand the 1st question exactly so correct me if I'm off.

    2010 FCF = 23, dividend policy of 2/3 implies 2010 dividend of 15.

    I assumed 3% LT growth in line with what most reasonable people would say was the LT growth rate of the world economy in real terms.

    You can therefore back out my implied cost of capital at 8%: [15/(8%-3%)] = 300


    Inflation protected means that I thought that JHD has pricing power.

    It's a conclusion that I arrived at after looking into what it does exactly, who it sells to, what its customers are likely to prioritize in the quality--price spectrum that constitutes "value" etc.

    Big picture: This is 2009/10, the worst environment seen in 30 years for commercial / institutional flooring, which was down 25% to 30%. The public sector institutional flooring sub-market was in even worse shape, as you can imagine.

    Against that background, JHD grew by 11% and (plus/minus sterling devaluation and raw material inflation) kept its gross margin intact.

    That was startling to me at the time and made it worth a look.

    Imagine that it is late 2015 and a midcap spare parts supplier to deepwater O&G not only grew revenue but accelerated its revenue growth. Imagine also that it gets back 50c every year on every $1 that it invests in growth. And imagine that it has a dividend policy trailing back a generation wherein 2/3 to 75% of FCF is dividended out.

    That's the starting point and the conclusion. In the middle is the hours of work that it takes trying to make sure that it is neither a probable dud (like, say, Dialight) or resting on uncertain ground(like, say, Xaar).

    I think now that I wouldn't buy JHD again at that valuation. Truth is that I wanted to own something that had the whiff of quality about it after owning lots of mispriced garbage in 2008-2009 and I didn't hold out for similar quality at a better price.

  5. Hi Red,

    What about Pan Orien Energy Corp.?

    They have no debt and fantastic onshore wells in Indonesia and Thailand starting to produce this year and in the comings.

    It´s difficult for me to analyce it well but maybe a great oportunity very close catalyst.

    Thanks in advance

  6. Did you look up what FB is worth without Macau casinos?

  7. Red,

    I looked it up. I saw that the revenues of FB are more influenced by the visitors inflow to Macau than the reveneu of the gambling.

    In the first half 2014 the total restaurant sq.ft operated were 142.000 of wich: Hotel Lisboa 7.000 (has a casino),Hotel Venecia 21.500 (has a casino),City of dreams 8.000 (It´s a casino), Sands Macau 4.500 (hotel-casino), Sands Cotai central 6.800 (hotel-casino), Rio Hotel 15.500 (hotel-casino).

    So more or less 30% FB restaurants are in hotel-casino. Maybe the strong drop of the revenues in gaming it´s going to affect to FB revenues but I think is very dificult to calculate.

    How do you see it?

    And Real Madrid knock out by Atletico... :(

  8. But I 'm confidente because the occpancy rate in hotels grow a little bit respect 2013. So maybe the drop in gaming revenus it isn't important.

  9. Red FB looks really interesting now. Except I don't like the whole Hengqin project. It feels like a state planned project to me that could fail horribly. 1.5-2 hours away from Macau. And that is what will eat up most , if not all of their cash.

    You have any insights on that? If that has a high probability to succeed, then obviously the stock is a complete bargain. Otherwhise it seems just cheap.

  10. Well Galaxy is spending US $1.6bn there, Shun Tak has been investing heavily etc. It's a "build it and they will come" approach for sure but then that has been the history of Macau and its environs. And Vegas too, I suppose.

    FB's value doesn't rest on Hengqin. I think what we'll see is that the company will sell 1/2 of its development right to a bigger cash-rich player with property development experience.
    (It's what I would do, anyway, and the CEO seems like a sensible fellow).

    The sale of that stake, if it happens, may mean a cash receipt equal to the company's current cap.

    Other than that, there are the Hong Kong and Zuhai eateries, the airport and school/University cafeteria, the franchised restaurant concepts, the moon cakes,the investment property, the Japanese food wholesale operations and so on.

    I suspect those are worth something, too.

    Expansion of industrial catering to casino workers would be a plus but it's just a free option.

    Who knows where it trades first, though?

  11. Any idea of the reason for such a drop in FB?

  12. David,

    Maybe this , but who knows?

  13. Red why buy calls for HA on such short notice? Seems the market can do a lot of things over such a short time? Why not just buy the stock?

  14. Hey Red,

    Any chance we will see a Paradise Entertainment writeup? You certainly seem to have some conviction in the idea. I was sorry to see AIQ replaced.

    I also wanted to say that this is by far the best investing blog I have ever read, and your research has generated very considerable sums in my PA.

    That said, a streak of bad luck can make any process look invalid in the short term, and you've certainly had your share of that in recent times.

    I can't wait for that luck to mean revert, and sincerely hope you continue to keep this blog alive and humming with brilliant investment ideas.

    Thank you for your contributions to the investing community.

  15. Very kind of you and thanks for reading.

    I'm okay with the draw down, though. Emeco has been unlucky, I think -- I had been relying on oil and copper to offset iron and coal so that disposals could occur in at least a somewhat orderly fashion. But there you have it -- they don't always work out.

    As for the rest, I'm comfortable with them.

    Too many names, though, and keeping track of them all is starting to approximate proper work. I keep promising myself that I'll get down to 6 x 15% and let them ride.

    Paradise: Sure, writing it up now and will post within the next couple of days.

    Thanks for reading

  16. Red, Emeco can buy back their bonds early right? They don't have to wait untill maturation in 2019? Otherwhise it seems like an extremely shitty deal to me. If utilization hangs around 55%, that is ebitda of about 75m$? That is only just enough to pay interest and maintenance capex.

    And after that they would need to raise over 300m$ from selling equipment.

    Looking forward to read your thoughts on Paradise :) .

  17. Hey red, might find this interesting:

    These guys just upped their position in Emeco. Their commentary is interesting.

  18. Thanks for the link. EHL's diversification is of course why I preferred it to other mining services names in OZ in the first place.

    But this formulation

    "The Canadian operations have 12 major customers, all which are profitable at oil prices of even $50/barrel"

    is misleading and, if intended to encourage other money managers to look at EHL, counterproductive.

  19. yeah saying cash flow profitable would be more accurate. They did put their money where their mouth is though.

  20. No - I don't mean that. I mean that talking about the customers being profitable at $50 is not the same thing as talking about whether their mining operations are profitable at $50.

    Those customers have a wide range of operations -- SAGD, pipes, refining, etc -- that are profitable at $50 and below but have nothing to do with EHL.

  21. right, thanks for clearing it up :)

  22. Hey Red,

    Have you ever taken a look at RCAP?

    Complex capital structure, Schorsch overhang, but seems very inexpensive on a forward basis-about 5x cash earnings including synergy projections, which appear reasonable at first glance.

    The rate sensitivity is also attractive and asymmetric, as cash accounts are producing almost nothing currently.

    Insiders have been buying, and future acquisition potential including tax shields are free options.

    Selling monthly puts generates pretty unreal IRR's also.

    Hope perhaps it interests you.

  23. Sure, I've looked at it but thanks in any case for mentioning it.

    How are you thinking about Schorsch, the veracity of the financials, and the risk of indictments?

  24. I see the issue at ARCP as more of a window-dressing event than outright misrepresentation of financial results as seen in Chinese frauds etc. So, I expect the economic impact, if there is some misrepresentation, to likely be insignificant relative to the current stock price and the earnings it implies.

    The managers involved in RCAP have good long term records at ING, LPL etc, know Schorsch, and have been buying stock. Corvex involvement in ARCP is also possibly worth noting given their depth of due diligence and track record.

    If you have any thoughts on the issues you noted, I would love to hear them. Both my forensic accounting and general investment knowledge are quite limited, as I have only begun investing seriously this past year.

    Looking at Hogg Robinson, the valuation seems similar to RCAP, but without the fraud overhang. Do you mind sharing your 1000 ft view on it? That is if you are not intending to write it up formally.

    I also noted your sale of Future Bright-having recently entered a significant position I would love to know if your decision was driven by opportunity cost, a lack of cash, or some impairment to the equity value?

    Thanks for your time.

  25. re: Hogg

    I guess the 1000ft view would be that:

    (1) SpendVision's worth something like 60 million so, setting aside the pension liability for a moment, I've paid just over 3x FCF for the travel management business.

    (2) It seems to me that they've guided a 28 million cut in annual operating expenses so that cash flows will go up if unless revenue goes down by a lot.

    I don't think it likely that revenue can fall all that far from here.

    (a) The function that TMCs fulfill will always be necessary and that function is more efficiently/effectively outsourced than performed in-house by Hogg's clients;

    (b) net contract wins seems to bear this out; (c)they are only now really focusing on what one would have thought was their natural client base -- global multinationals with out of the ordinary travel management needs, like O&G companies etc.

    So, if we split the difference -- 300 in fwd run rate travel revenue and 252 in travel opex -- we'd get to 2x FCF for the consolidated travel business.

    The unconslidated JVs, NCIs etc have value also. Of the 85m in cumulative retained profits accruing to HRG virtually all of it has been reinvested for growth.

    With respect to the pension liability, a 2% rise in corporate bond yields more or less eliminates the net liability and also the 7 million in associated pension financing charges on the income statement.

    That's the fundamental picture as I see it -- more or less: inexpensive, sturdy cash flows with an interest rate hedge attached.

    I invested in it b/c I think the resolution in the valuation discrepancy might be quicker than average. I'd like to see 80p in relatively short order.

    Future Bright, on the other hand, can wait a few months. I'm okay with a ~20% position for now. If I can see the catalysts coming -- Macau sentiment, sale of part interest in Hengqin -- I'll bulk up again. So yes, it's just a cash management issue.

  26. Hi Red,

    I'm loving all your recent posts, so thanks so much for taking the time to write them! I was wondering whether you had any plans to continue this series about method (particularly re where you find ideas); it's especially useful for us younger and relatively more inexperienced investors out there (in a teach a man to fish sort of way). I of course understand if you're too busy/prefer the specific writeups so you can get feedback on your ideas, but just thought I'd ask.

  27. I will, Toby. I'm trying to keep a balance between writing up ideas and writing about method. I suspect it'll take me three or four weeks to finish the remaining chapters of the method series.

    There's a also a series on business models
    that I need to finish.

    Thanks for the prod & for having been patient.

  28. Thanks Red! I hadn't actually noticed the business model series, but it looks fascinating too.

  29. Hey red I looked at Texhong and Future bright, and what I noticed is that they were so cheap on an price earnings basis. They kept growing earnings and returning capital, but between 2005 and 2010 it seems as if the market did not really care. Seems like those Chinese only look at dividend yield?

    It seems to have traded no higher then 7-8x earnings, and often at 3-5x earnings, despite rapidly growing.

  30. Asian markets look at NPV of dividend stream, yes.

    Plus, (1) if it pays out less than 20% of earnings it will trade at a heavy discount to NPV of dividends (2) if it pays out > 50% it will trade at a premium to NPV of dividends.

    I understand it; it's an entirely reasonable means of accounting for the uncertainty and risk inherent to equities.

  31. Interesting discussion on premium vs. discounted NPV of dividends in Asia...

    how do you view a stock like Pico Far East (752 HK)? Trades at 6-7x EV/NOPAT, pays out 50-60% as dividends, and growing at 9-10% per year. Seems like no premium?

    Also, really enjoy the posts on investment process / strategy in addition to individual write-ups. Keep it up!

  32. Right, one would look at PICO and 70% payout on trailing $0.15 in earnings = $0.10 dividend and 5% payout = $2/share.

    No traling 5 year growth, really, so you'd expect a 6% or 7% yield other things being equal.

    So there's some premium there and that's due entirely to the quality of Pico's mgmt and it's outlook as it transitions into an Asia-wide business.

    Good deal if you could pick it up at $1.20 to $1.40 per share.

  33. Red how do you take into account overall China risks?

    What happens in the case of a revolution with companies like keck seng and texhong? So not some property bubble, but full blown political unrest.

  34. Allthough some of the arguments in that argument are rather weak, I have to admit.

  35. Hey red,

    Any plans to continue this series? I don't think I've seen anything come out on it since this post but might've missed something. Would be really interested in hearing more about your process, especially the idea sourcing phase.



    1. Will do, Peter. Sorry it's taken me so long to figure out what to say and how to say it. The blog will be 4 years old in July so I'll write it up at that time and in that context.

    2. Glad to hear it! Looking forward to reading.

    3. Also would like to heard how you managed to keep up with everything. Most of us has a full time job and other responsibilities. It's mindblowing that you research and followup so much for stocks in your watch list.

  36. Hi Red, any follow-up on the series? Would it be possible to write a little bit about the difference between your method vs warren's method, in terms of valuation, moat, etc? I learned a lot from your blog. Very much appreciated.