No machine could have predicted this VLCC spike

Mr Prospector is not a fan of algorithms predicting the markets.

It is fair to say that just when you think you understand shipping something always happens to make you realise you don’t. Not even close in my case. Take for example a close friend of mine who drinks with me at the Airdrie Working Men’s Club. He was a tanker trader for many, many years. He then took some time off, before coming back with a bang at a start-up data company that was going to redefine the way shipping worked. Moreover, it was going to redefine how it was traded, using algorithms and whatnot to show that with enough data, and most poignantly with enough money, the beast could be tamed. In other words, we’ve all been doing it all wrong, but if you hire enough people who went to Harvard, they will tell us all how to do it correctly.

My pal spent a lot of time telling me the genius of these Harvard colleagues (he did a B-Tech in Drama at Airdrie Institute For The Educationally Bereft by the way, an institution that has provided a huge number of shipping alumni over the years). He ran a ‘black box’ trading model, that we dubbed Metal Mickey. Sadly, Metal Mickey didn’t really work very well. Another tweak, another Harvard expert hired, it was turned off and on again, yet still it didn’t seem to predict the future any better than I did after a dozen vodkas at AWMC (future was always kebab followed by hangover).

As the years went by he seemed to be less and less enthused by the latest missives from the brains factory. It got to the point where Metal Mickey was just a flashing box in the corner, on which people stored dirty coffee cups. That’s not the way tens of millions of buckaroos should end up, unless signed by Man Utd of course. Finally about three months ago, and (ironically) predictably, his tenure in charge of Metal Mickey was ended. At around the same time, the tenure of most of the Harvard geezers was also ended by their ambitious sugar daddy funder. You didn’t need an algo to work out that that was coming. In fact, as far as either of us know, the whole damn thing has gone into the bin now, along with dustbins of cash, quite a few dreams and some gold-plated CVs.

Fast forward to this week, and I bumped into my mate in a ropey boozer in Airdrie. All he could say was ‘I’m sitting at home thinking I wasted five years waiting for some sign that we could predict the future. Three months after we give up and the biggest boom ever in the tanker market happens. And all I keep reading is people saying., “Nobody could have predicted this!’’.

Had the doors stayed open for another three months then all of that work would have paid off handsomely. He’s now sat on the sidelines watching all of those binary ‘idiots’ making out like bandits. Being right for the wrong reasons, better to be lucky than good, etc, etc. You’ve heard it all before.

Is there a moral to this tale? For once I think there is. No analyst, no report, no algo, no Harvard tech expert, no sooth sayer, no caster of runes had a clue that this was going to happen to the tanker market. I don’t care what anybody says, no oil production figures, AIS pictures of tankers loaded or in ballast, no refinery data, none of the things that we are being told and sold predicted the biggest boom in VLCC history. Easy to throw your hands up and say ‘the market is the market’, but in reality the endless study of supply and demand of a single commodity showed nothing useful. Price lovers also didn’t see it. So let’s think about that for a moment. If you had perfect forward vision and were gifted 100% accurate data in advance of events in terms of production and consumption of oil, along with 100% accurate fleet data to the deadweight tonne, you still could easily have been bearish in the face of a $300k/day bonanza. Don’t lie to yourself, you easily could have.

“You cannot predict political events,” I hear you say. But if they impact markets in such a powerful way, overriding all other factors that we are sold as being so important, then shouldn’t we be trying? Forget the oil/fleet data. It doesn’t work when it can be tipped over and trampled by one geopolitical event. Do you want to come to a barbecue at my house tomorrow? The weather forecast is sunny all day, provided there isn’t any rain/snow/fog/locusts/badgers, which are all realistically possible. What should I wear then? Maybe the hundreds of millions spent on research should just be spent hiring political analysts? After all, that’s where the action is. And if that doesn’t work, then very possibly, and I do not say this with joy or conviction, you would be better off saving the entire industry research budget and paying it back to shareholders or, better still, spending it at the blackjack table. At least you get free drinks there!

Not heard of Mr Prospector before? Here are some recent tweets to give you an idea of the thoughts and opinions of the most scurillous voice in shipping.


  1. Meanwhile, all thank you cards and letters may be sent to:

    Donald John Trump AKA #45
    1600 Pennsylvania Ave, NW
    Washington, DC 20500

  2. And send a thank you card to Wilbur Ross, at least recently a big owner of stock in DSSI tanker company. He is making a killing selling of some of his stock now. No doubt he has to thank Trump for it, but who gave Trump the idea???

  3. Good one. As they say in New York around the bar at Bobby Van’s (I guess sort of like the Airdrie where you hang out), “who wudda’ thunk it?”

  4. Whole article appears to be as the aggressive – defensive approach of a person that has been rejected.
    If we want this world to move forward we have to abolish self-centric psychological deficiencies and open up to the future (which by the way is already the past for most other industries than shipping).

    Come on! Look ahead! Forget about coal…

Back to top button