taleb as miseducator (part 3)

on antifragility and disruption

And so we come to the big one … “Antifragile”, which I would argue is the Taleb book which has been most badly misunderstood and misinterpreted, with the most destructive consequences.

The trouble is that the full title of the book is “Antifragile: Things that Gain from Disorder”.  This makes it easy to think that it’s a book like “Manufacturing Consent”, where the major point of the book is succinctly summarised by the two word phrase “manufacturing consent”.  Although Taleb coined the word to suggest that “robust” was in the middle of a spectrum rather than at the opposite end from “fragile”, the point of the book isn’t just to say that some things do better than others when the shit hits the fan.

In fact, he points out early on, and repeats it frequently, that the simple fact that if you’re strong enough to take advantage of others you can often exploit them more in a crisis isn’t the type of thing he’s talking about; he gives a faux-biblical commandment that “Thou shalt not have antifragility at the expense of the fragility of others”, and I might argue (like for example, in part 4) that a significant part of the motivation for writing “Skin In The Game” was to try and stop the brooms from marching.

Rather than this, Taleb’s “antifragility” is a similar concept to Ross Ashby’s “ultrastability” (discussed here).  It’s the ability to reorganise oneself to deal with a change in the environment – the possession of enough capacity and bandwidth to deal with shocks that couldn’t have been anticipated.  Or, for that matter, shocks which could have been anticipated but weren’t.

“Antifragile” came out in 2012, right around what can be retrospectively identified as the peak of my banking career.  As a result, because I was doing quite a lot of international marketing at the time, I ended up talking about it a lot to the kind of people who picked it up and read it.  It interested me at the time that reaction to the book came in two sort of waves.

The real heads and enthusiasts got what I regard as the key point immediately – it’s a book about option value.  The ability to make a decision or refrain from doing so is in many cases equivalent to a financial option, and it’s an absolute basic principle of finance theory that an option has positive value as long as it remains unexercised.  For this reason, you keep options open in finance until the last moment possible – either when it is about to expire, or when keeping it open would require you to make some other payment which is bigger than the value of the option. (I oversimplify). 

This is a bit abstract, but it’s the way that finance quants (like Taleb) think about these things – the more decisions that you could take, but haven’t yet, the better you will do in an environment in which things happen which you don’t know about yet.  In the early days, people started writing pop articles and consultancy reports talking about “preserving optionality”, and we were all happy.

Then the “disruptors” showed up.

It wasn’t just a misinterpretation of “Antifragile” they were armed with – it was also a particular interpretation of the Warren Buffett “moat” concept, and of course they were for the most part operating under the hallucinogenic influence of a large pile of cheap money.  There was also the overhanging influence of imperial-period Jack Welch, who encouraged managers to “destroy your business, dot com” as a way of galvanising them to understand the competitive impact of the first Internet bubble.

There were elements of this business strategy in Hal Varian’s textbook, but a) that was difficult and technical and b) he tended to emphasise that you needed to develop increasing returns in order to gain a cost advantage so that your “moat” actually existed and you weren’t vulnerable to another party with even more VC money doing the same thing to you.  And of course c) Varian’s book was written in the first tech bubble, so it referred to actual software companies which really did sometimes have this sort of economics, rather than taxi and grocery companies pretending that they were software.

All these threads came together as a kind of bastard antifragility.  If you price below the cost of production, then everyone else will go bust, and you will be a monopolist who then has a “moat”.   If you spread out and do enough chaotic things, then one of them will work and you can “pivot”.  Move fast and break things.

Of course it turned out that the one thing this business model couldn’t tolerate was materially positive interest rates; a classic example of the “not exactly a Black Swan, is it?” species of Black Swan.  The big misinterpretation here was one that’s a consistent theme of Taleb-miseducation that I haven’t emphasised much so far – libertarians assuming that he’s one of them.  In so far as there’s a theory of government in Taleb’s books, then it’s one of matching the appropriate level of organisation to the activity.  Since he usually thinks most states are too big for what they do, he often sounds like someone who’s in favour of decentralisation per se.

And because antifragility is basically a property of organisation and reorganisation, they assumed that he must mean maximal decentralisation and self-organisation, without any planning at all.  People assumed that not having a plan meant you could never get loss, and that throwing bombs was the equivalent of urban clearance.  They also thought that stock options meant you had skin in the game, but that’s for another post.