On Bullshit

Ten years ago I was the first French publisher to be interested by a book released in 2005 by Princeton University Press. Even if I mainly publish academic books (that means: for hundreds of readers, not thousands), I thought that this book would have been highly successful in France. The book is named On Bullshit, written by American philosopher Harry Frankfurt. I tried to buy the French rights of the book but unfortunately it was so successful in the US (“the book appeared for twenty-seven weeks in the New York Times Best-Seller list”) that I couldn’t pay the advance requested (thousands of dollars). The book was then published by a big French publisher under the title De l’art de dire des conneries, and tens of thousands copies were purchased by the French readers – too bad, for once I missed the opportunity to make a lot of money. I loved this both serious and ironic study on the art of “bullshit” (translated baratin or conneries in French); Frankfurt distinguished between bullshit and lies and concluded that “bullshit is a greater enemy of the truth than lies are”.

Last week I was on a train, coming back from Paris to Brussels. For once I bought the main French newspaper, Le Monde (technically, Le Monde is not the main newspaper but it’s considered as the main serious newspaper) and I came accros an article by Paul Jorion, a very well-known economist in France (and I suppose he’s known in the US too as some of his books were translated in English). The article is titled Everything just got better if the markets are ignorant! Here is the note:

Programmer Alberto Brandolini once said that “the amount of energy necessary to refute bullshit is an order of magnitude bigger than to produce it”, so I won’t spend a lot of time to comment some weird theories. For instance, Jorion writes that if the participants/algorithms know a lot of information, this situation leads to crashes, or mini-crashes (because of the mimicry); conversely, if the participants are blind and don’t know nothing, then everything if perfect (that’s weird because if the participants don’t know anything, there is no price anymore, and no market anymore; it reminds me the fun story about an algorithm named Sniper, invented by Todd Kaplan and involved in a contest in Santa Fe, back in the 1990s; the story is told by Rust, Palmer and Miller in “Behavior of Trading Automata in a Computerized Double Auction Market”, and that’s why I choose the “character” Sniper as the narrator of my book). In the last paragraph, Jorion suggest to slightly tax “transaction cancellations” generated by algorithms, so “as if by magic all the mini-crashes will disappear!”. Yeah, just do it – it’s easier said than done.

The main problem here is about “transaction cancellations”. This is very strange by someone who is supposed to know the markets well (Jorion worked in the financial market world in Wall Street some years ago) to confuse “orders” and “transactions”. An order is a quotation, an intention to trade at a certain price; a transaction is deal between two counterparts who agree on a certain price. Jorion states that “98% of the transactions are cancelled by the algos”, and here is the connerie. Only the exchanges can theoretically cancel transactions (a lof of them were canceled after the Flash Crash in 2010; most of the Knight Capital trades were not canceled in 2012 when the firm had serious issues with their smart order router and crashed down, but when Goldman Sachs had a trading software glitch in 2013, all the exchanges canceled most of the trades, “limiting losses that could have cost Goldman $500 million” – I always thought that God must have called the exchanges and said “Hey, the bank works for me, so you have no choice: cancel the trades!”). Obviously, algos don’t cancel trades; they only cancel orders (Matt Levine did an interesting piece recently on that, “Why Do High-Frequency Traders Cancel So Many Orders?”). That’s a shame to see that an old student of the great French anthropologist Claude Lévi-Strauss writes absolute bullshit.

My point here is that Le Monde is viewed as the most serious newspaper in France (“le quotidien de référence”), but publishing this kind of baratin is not good at all – for the readers and for the market participants, because most (educated) people are educated by reading newspapers, and if newspapers publish bullshit, people’s judgement on markets are falsified. Both as a book publisher and a teacher, I think good education is vital, and I am struck by the the lack of seriousness of some people who are given carte blanche to write this kind of connerie. It’s the pits – sorry for the pun, as I have a lot of respect for the old human pits.

PS

1/ Some newspapers also published bullshit recently about the Virtu case in progress in France (the French regulator AMF prosecuted Virtu – more specifically Madison Tyler, who merged with Virtu in 2011 – for spoofing activities in 2009 and 2010), I’ll post a short story about that soon.

2/ A new academic paper was just released yesterday about “Limit Order Placement by High-Frequency Traders”. Even if some will moan (because the scientists used the now well-know criticized Nasdaq dataset from 2011), this article is far more interesting than the baratin published by Le Monde.