It’s Not Easy


I’ll start this post with a quote from Howard Marks, who himself is quoting Charlie Munger. I never said I was original:

In 2011, as I was putting the finishing touches on my book The Most Important Thing, I was fortunate to have one of my occasional lunches with Charlie Munger. As it ended and I got up to go, he said something about investing that I keep going back to: “It’s not supposed to be easy. Anyone who finds it easy is stupid.” As usual, Charlie packed a great deal of wisdom into just a few words…

… what Charlie meant is this: Everyone wants to make money, and especially to find the sure thing or “silver bullet” that will allow them to do it without commensurate risk.

The talk of the town is the Globo situation. I have nothing to add on that front, so I won’t presume to try.

One thing I do like to watch, though, is investor behaviour in the wake of events. Many chalk losses up to bad luck, or attribute the blame to some external factor beyond their or the company’s control. Sometimes this is reasonable; often it is not. Others look to their process – how they are selecting securities – and then try to figure out what can be fixed to make sure they do not fall victim to the same mistake again.

This is a noble endeavour – a bit like the race driver who figures out, through repetition, that he needs to swing a little bit wider on the second corner to avoid clipping the dirt. There’s an appealing sense of progression in self improvement; a logic that, if you can just fix what you did wrong with every misstep, you’ll end up being a consistently profitable investor.

I think it is often a misguided one. A sample size of one is (almost) never large enough to make a decision on anything in life. This is not to say it is bad practice to try – a post mortem is often a good thing to do – but it is a mistake to put too much stock in said post mortem. Even the implication of the phrase ‘post mortem’ is false, seemingly implying that you should only think through your logic when it was ‘proven’ faulty. The fact is that the market, by its very nature, will make you wrong very often, and hopefully right slightly more often. Variance is large, and many of the winners will be winners for reasons you had not envisaged at the start. Whenever variance is large, you should treat every single data point with a large dose of scepticism and a sense of perspective. Stocks, being story driven, emotional and long-term in nature (if you’re a value investor with a decent time horizon) detract from the ability to keep that perspective.

On this vein, and as a curious observer of the discussions on Stockopedia, I noted a once-more recurring theme on many of the comments on the Globo saga. Stockopedia operate an algorithmic ranking system for stocks, whereby they rate stocks on three axes – value, momentum and quality. It trawls through the numbers and evaluates financial statements based on their similarities to a number of investing criteria which have outperformed in the past.

Unsurpisingly, Globo ranked highly on their algorithms. The headline numbers, falsified as they have now been proven, were fantastic.  This leads people to ask the same questions they asked when Plus500 (another stock that ranked superbly) blew up:



These sorts of comments miss two fundamental truths.

Firstly – it is easy to fight the last war, and everyone trying not to look stupid does it. You can see this in broader markets when you consider the great efforts expended by regulatory organisations to tighten up banking systems post the financial crisis and ensure sufficient capital is available through future stress scenarios. The US have even tried to make a list of ‘systematically important financial institutions’ in the hope that identifying and imposing stricter rules will prevent another financial crisis. There will be another deep recession. It will probably not be caused by what people think it will be caused by, or what people are preparing for.

Secondly – the whole point of markets is that it is almost impossible to meaningfully profit from anything that is obvious. It is obvious, it has been priced in. You could eliminate the vast majority of dud companies with a strict set of criteria. Would it be profitable to do so? The answer to that question isn’t at all clear.

People, for perfectly understandable reasons, want one number to be able to capture everything about a company and sum up whether it is likely to be a profitable investment. When said number has a false positive, they want to fix the number to account for the error. This is a futile quest. You will end up fighting every last war, and progressively excluding more and more signal with your noise.

There really is only one answer, and it’s not really a very surprising one. You can screen, and you can use aids for finding companies to look at – ways of narrowing the ocean to a small pool in which you can fish – but you cannot outsource doing due diligence on an actual company. If you do decide to outsource said due diligence to an algorithmic screener – and for many investors (if only because most investors meaningfully underperform) this might be a good idea – what you absolutely cannot do is start to pick apart bad or good situations and try to second guess output based on that.

If not tinkering with the output of a screen sounds too light-touch to you – too hands off and unreliable – there is only one answer… actually engage in all the due diligence yourself.

… and if the Charlie Munger quote at the start was a bit bland for you, I’ll leave with the immortal words of Ronnie Coleman: “Everybody wants to be a bodybuilder. But don’t nobody want to lift no heavy-ass weights”.

10 Replies to “It’s Not Easy”

  1. John at UK Value Investor

    Excellent points indeed. I am a big fan of post-mortems and continuous improvement, and you’re right that it’s important to not “learn” the wrong thing from an anecdotal experience. As you say a sample of one is usually to small, although often it’s all we have, and we have to work with what we have.

    Also, love the fact that you’ve somehow managed to fit a Ronnie Coleman quote into a post on investing. I never thought I’d see the two on the same page together! Excellent. “Ain’t nothing but a peanut.” etc.

  2. Jane

    Excellent article. The Stockopedia ranking and screening systems are a very good starting point for researching investments and Paul Scott’s commentaries are then very useful for flagging up areas of concern that sometimes seem to slip through the algorithms but then there is a need, imo, to complement and complete these by extensive due diligence into the “business reality” of a company, This third element is perhaps the most challenging for most investors as it takes us into commercial and technical knowledge areas perhaps outside of our experience and because it requires a great deal of time and application, as well as personal discipline. This process also throws up areas where we just need to realise and accept that we might just not know enough or understand a business well enough to want to risk investing in it. As you say, “it isn’t easy” but investing personal time and effort to getting it right (or to avoiding getting it wrong!) before actually investing your money must give an investor an edge over others who can’t be bothered to do it.

    • Lewis

      I’m not sure it must give an investor an edge – lots of well-informed investors who put considerable time and effort into researching and understanding prove to just have systematic biases which seem to prevent them from ever achieving any real outperformance.

      I also think that there might even be some easy wins to be had from tilting away from the market-cap weighted indices that people call ‘the market’ (i.e. – “you can’t beat the market”). It’s a reasonable sounding phrase touting market efficiency, but I consider the very structure of these indices to be bad enough in the first place that just by applying a few formulaic filters you can probably beat it.

      That said, we’re obviously on the same page. If you start stock picking, you should take the time to try and properly understand exactly what you are buying. The growing prevalence of products which state that you can pick the best stocks just by running a filter on ‘highest return on capital for last 10 years’, ‘best share price performance’ and ‘lowest earnings multiple’ – or whatever some smart provider comes up with next – creates even more of an opportunity for those who want to spend the time actually looking into a company’s prospects. It’s not a slam dunk and one will make mistakes, but I think there’s certainly value you can add, particularly in small caps. The very fact it is challenging, requires broad and deep knowledge and necessarily is zero-sum (namely you must be better informed than other market participants) implies as much.

  3. Frank

    All these complications ! It is a zero sum game and the average will always underperform by the margin of their costs. You may overperform today, but you will underperform tomorrow. Why not be satisfied with market returns, invest monthly in an index tracker, minimise your costs and get rich slowly? The S&P500 has been going up over 9% annually for the last 50 years. “Enough for the day, ’tis the evil there off”.

    • Lewis

      More power to you.

      I think ‘the market index’ is constructed in such a way which makes a small amount of overperformance not enormously difficult to achieve with some effort. A ‘small amount’ compounded over a long period of time will lead to dramatic differences.

      Of course, I also hope I am able to outperform by a meaningful margin and not just a small amount. It’s been 4 years, so I’m not ready to make that claim yet.

  4. Paul

    Interesting article. I have been having quite a bit of debate with a friend recently about the relative merits of:
    a) statistical ‘factor’ based analysis and investing using algorithms (accompanied by a mechanical trading approach) in the way Stockopedia does (or UK value investor who commented above)
    b) using qualitative analysis and judgment (which could be by an ‘intelligent’ investor who has done a lot of research).

    With b) I think some people can perform very well. However, these people are few in number. One problem is that with b) you expose yourself to too much information that is potentially relevant (to what is actually a binary decision) and it is easy to (either consciously or subconsciously) focus on wrong or misleading information e.g. my impression is that many investors will base decisions on the basis of someone else’s view (for professional investors often the company management) and it is very easy to be persuaded by somebody else’s confidence. With b) we are also exposed to numerous behavioural biases that are detrimental to our returns e.g. a strong reluctance to buy after a price rise; a desire to sell after a price rise; a reluctance to crystallise a loss; a tendency to overtrade etc. etc. It is very easy to underestimate the strength of these factors and the significance of their impact on returns. With b) you are betting on the fact that you have mastered these biases more than others, you know what information is relevant and are more clever (or hard-working) than the professionals in the market. I think it’s actually very difficult for most people to be confident on whether any of these points are true…

    There is of course also some scope to do better or worse than the market average doing the mechanical factor type approach a) but my impression is that for most investors there would be a strong likelihood that you would do better than the market average if you focussed on broadly the right factors and stuck to it (e.g. there seems to be quite a bit of research showing that even the most simple momentum strategies outperform in the long-run). The advantage is because by being mechanical you avoid the pitfalls of b) above. I think many people have an inherent scepticism that something like a) could work in the long run – because if it did then surely all the professional investors would do it and arbitrage away the profits. This is a very sensible question to ask but my view is that it vastly overestimates the efficiency of markets: why assume that most other people will have the discipline to behave in the most profitable way when you cannot (or choose not to) do so yourself?

    Of course you can also combine the two in some way, which could potentially be even better for some people. However, following the reasoning above, for most I would imagine the more discretion you give yourself the worse you will do if you are an average investor.

    • Lewis


      A very thoughtful comment, thanks.

      I don’t disagree with anything you’ve said. I suspect the majority of investors would do better following a simple approach; just trying to eliminate behavioural biases and signing up to a rules-based methodology would get rid of the most egregious ‘value destruction’ individuals exhibit, like selling down massively in 2009 and not getting back in until the market ‘feels good’ again. This very blog came about from me trying to figure out if I fall into the camp that can outperform through bottom-up analysis, and with (as I mentioned above) only 4 years of data, I’m still far from being able to even guess an answer to that question, though I am now employed in the sector!

      The only real pushback I’d give you is on your third paragraph. There is plenty of evidence for a mechanical factor approach, but I am uncomfortable with any ‘factor’ that I can’t clearly perceive as being grounded in behavioural psychology. I can sort of stomach that belief with the value or momentum effects, but I do worry that these factors are becoming so well known that easy profits are being arbitraged away, as you mention. My slight suspicion of factor-based investing generally is that the lens with which we view these back tests and ‘experiments’ to see what does and does not outperform is borrowed from the scientific world. In the scientific world we conduct experiments, holding conditions constant, and measure the results. The problem is that we take that approach and try to adopt it in a world that is inherently reflexive and ever-changing in the market. Things are not constant here – as soon as information becomes available it starts being priced in; if the information ever did contain genuine alpha and was not just a statistical quirk, that is. So you could show me any amount of research showing that a simple factor outperformed over however long a time-period you like, and I would still be deeply wary of extrapolating the results. The very existence and observation of the effect may prevent its recurrence unless, as I said first, the effect is grounded in some stable constant of investor psychology that is not liable to change over time (pessimism and negative shocks persist longer than optimism, herd mentality, difficulty in buying stocks which have gone up lots etc.). Even then you’d need to get comfortable that the market cannot ‘override’ its biases – which may seem a naive thing to say, but plenty of ETF providers are launching factor-tilted products which may significantly lower the hurdle for people to surprass their inefficiencies. Much easier to buy the ‘Vanguard Value ETF’ than actually pick from the dregs of dire stocks.

      Do I think this will ‘fix’ these biases? No, not really. It does give me pause in extrapolating factor performance, though.

      That said, to complete the circle, even that esoteric argument is irrelevant for the vast majority of investors, who would do much better following that approach that doing what they do currently. I don’t say that on a high-horse, I simply look at the statistical massive underperformance of actively managed retail portfolios.

      And to further complete the circle, the shrewd observer will note that I still pick cheap-looking stocks with quality characteristics. I do a lot of screening for ideas on this basis. So arguably I use factor-like ideas to create a sub-set with which to do further due diligence. Where does that place me on my own scepticism..?!

      • Paul

        Hi Lewis

        Sorry it took me a while to reply to this – I forgot I had left a comment here.

        Yes I agree with you very much on the difficulty in having statistically reliable evidence on factors that one could extrapolate into the future. This is the fundamental problem – the level to which any investment approach is statistically ‘proven’ to generate outperformance in the future is limited as there is so much noise / randomness and things change over time. So an unproven belief in why a particular approach should be profitable is required if you want to outperform the market.

        Partly because beliefs about what approaches work are very difficult to test satisfactorily (and also because approaches are very diverse and the flow of information between market participants is limited and very noisy) I think it is likely that the tendency of certain factors (like momentum effects or outperformance of ‘cheap’ or ‘quality’ shares) to get arbitraged away over time is actually pretty low. Of course it’s pretty much impossible to test the actual answer to this question.

        But, important to note as you say, that your approach, which uses an element of screening and some judgment to identify stocks you think are cheap or high quality is not really any different in this respect. For this type of approach to make sense you have to either have an idea about why the market has specifically valued a particular stock badly or (generally more relevant) why in general ‘cheap’ ‘quality’ shares according to some definition would be likely to perform better on average. If you believe the latter then you should recognise that you are implicitly placing some faith in the logic behind a factor type approach.

        I think the key difference in practice is more to do with the trade-off between the flexibility to account for additional qualitative information and the discipline that a mechanical approach brings. Or more accurately (as everyone is different): ‘to what extent should I mechanise my decisions as an investor?’ The right answer for most investors may be ‘a lot more than currently’

        • Lewis


          Very cogent thoughts, thanks.

          I agree with everything you say; particularly that you will require an unproven belief. Looping back, that’s actually a neater explanation of some of what drove me to write the piece in the first place – the thought that you cannot satisfactorily quantify what a ‘good’ investment will be based on looking at past winners. You need conviction in something, based on some more fundamental understanding (behavioural or otherwise) of market dynamics.

          You last paragraph is key – the ‘progression’, if I can use that word, in my own investing is trying to understand factors which will influence performance which cannot be screened for. I think that will be a more enduring source of competitive advantage. It is also the holy grail which leads the vast majority of investors away from a framework which would serve them much better, namely mechanisation.

Leave a Reply

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