Learning from investing mistakes

One of the great things about investing is that you can learn a lot from other people’s mistakes without spending any money. It’s inevitable that you are going to make mistakes yourself, but you don’t get points for reinventing the wheel. So when someone decides to do a post-mortem on a failed investment I’m all ears. Especially since you usually see investors talking about their latest idea’s: not their failed idea’s. A good blog post that I recently linked to on Twitter was this one from Glenn Chan about junior miners, doing your own homework and how some things are just too difficult.

Sometimes you also read a post-mortem, and you wonder if the author really learned anything. Today I came across this blog post from Vitaliy Katsenelson titled “What I learned from the J.C. Penny Fiasco”. In the piece he argues that the JCP investment wasn’t a bad investment: it just didn’t work out. That’s of course possible: if you bet when the odds are in your favor you will win in the long-run, but it doesn’t mean that every bet will pay off.

But if you want to review your investment thesis you can’t just assume that your initial assessment of the probabilities was correct. Vitaliy thought that the probability of a turnaround at JCP was ~70 percent (with a lot of upside) and the probability of failure around 30 percent (with ~40% downside). If that would have been true an investment in JCP would have been a fantastic bet.

I doubt that this was the case, because there is simply not a lot of logic that supports the 70/30 probability distribution. How often are declining companies able to turn around? How often do retailers turn around? How often do companies turn around by trying a radically different strategy? How much experience did Ron Johnson have in turning around a struggling retailer? I don’t know the exact answers to all those questions, but I’m pretty sure that there would be absolutely nothing in the data to support a high probability of success. When you grab some bullshit probabilities from thin air it might be a good idea to start there in your post-mortem…


No position in JCP, so far just following the story for entertainment value

6 thoughts on “Learning from investing mistakes

    1. Alpha Vulture Post author

      It’s always a bit lame to prove your point using quotes from famous people, because you can always find some quote, but this one seems to fit pretty well: “I`ve said many times that when a management with a reputation for brilliance tackles a business with a reputation for bad economics, it is the reputation of the business that remains intact.”

  1. Thomas Braziel

    All good points. I would have to say that your post bring up my biggest disagreement with a number of investors (value guys included) which is the use and abuse of probabilistic assumptions to justify investment cases (or to argue for a situations asymmetric risk/return profile). I’m not saying the framework is incorrect, but it very difficult not get the rose colored glasses on without knowing it. At that point, you’ve essentially left the numbers without knowing it. Also, it is very difficult to learn and become a better investor from these situations, because even if it doesn’t work out, you might convince yourself that it was still a good bet (and you might be right).. BUT you’ll never know.

    1. Alpha Vulture Post author

      I guess that is true for every single investment: you never really know if you got lucky or not. For example: a thesis might play out exactly like you predicted, but it might just be because a very real and very big risk that you didn’t see didn’t realize itself. Getting high quality feedback from decisions made in an environment where random events play a big role is just very difficult.

      But if you make a couple of (perceived) 70/30 bets that don’t work out you quickly get a sample size that is big enough to know that you are either extremely unlucky, or just wrong. Trying to estimate probabilities of various scenario’s isn’t a bad idea I think: it’s better than nothing. But you have to base the probabilities on something solid, and you have to be cognizant of how there is not only uncertainty about the outcome, but also uncertainty about the probability distribution itself.

  2. Matt Brice

    Emperor has no clothes. You hit the nail on the head about these probabilities. I always laugh when people write about these probabilities in such precise terms. Interestingly enough, as you have subtly noted, those that use such “scientific and precise” language are the ones who are usually writing a lot of the content out there.

    Luckily, I hope other investors (presumably readers of Institutional Investor) continue to use their Kelly criterion.

  3. red.

    Well, I think it’s worth wondering if it was anything more than a copycat trade. I certainly thought so when I read the original piece last year and am more inclined to believe it now. Ackman’s hit rate on activist positions approximated 7/10, and he exited when Ackman did even though the opening stanza of the long thesis was premised on time arbitrage.

    The Tesco piece seems quite similar, as I say.


Leave a Reply

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