It is 11pm on March 21st 2019 as I write this. Press releases from EUCO indicate that an unconditional extension is granted to the UK, rendering my bet on departing the EU on March 29th lost. It may be instructive therefore to examine the bet: its aim, placement, and implications.
I bet from a variety of motives. There is the desire to know the world and measure myself against it, to commit to a prediction, simple greed, and the purity of the wager itself: in gambling there is just you, the bookie, and no excuses. There is also a motive that has been out of fashion for some years: the insurance bet.
The 70s was the time of high-tax Labour governments, and businessmen used to bet on Labour in the hope that their winnings would salve the wound. Similarly for no-deal Brexit: projections of £1=$1.15 or lower and £1=€1 concerned me greatly and I wished to offset it, to compensate for the malevolence of others in these foolish times. So I decided to bet on a no-deal Brexit.
These days bookmakers offer a wide range of options on a given event, and a list of them is in previous posts. But there were two constraints that made the selection difficult. I stick to high-street bookmakers so online options were not viable, and all the options matured on the March 29th so I could not insure against a no-deal post-29th. After some consideration I selected Betfred’s 9/4 on exit by March 29th, got £500, and resolved to place the bet.
I work away from home during the week, and this year I am in a Midlands town: a large town or small city, dependent on mood. As you walk away from the train station the architecture changes, becoming rougher: bay and forecourted, but peeling plaster, dirty from car exhausts, an edge place between suburb and town. The shop door shines white under dirty yellow street lighting and I go in around 8:30pm: screens to my right, FOBTs behind me, counter to the back, bloke tidying up, nobody else in: there are no monsters here.
I show him the screenshot on my phone: political betting is so rare I have to screenshot the website so they know where to find the odds. After some confusion (he thought I was protesting) bloke nods, finds the odds and the slip is filled out. He takes the notes, I keep the slip, I walk out and back to my weekday room.
I thought that an extension was politically impossible, but after the bet was placed somebody on PB (@AndyJS ?) posted a link to a meeting of Conservative party association chairs approving one. The impossible was becoming possible before my eyes.
A later post by @RichardNabavi linking to a Guardian article reinforced it: if an extension was requested then the UK would grant one, and the ERG would bleed away the time to a no-deal. This eventuality was not covered by the bet, and the news tonight confirms it. The bet is lost.
If no-one is corrected then no-one learns their lessons. So what do we learn? The error was twofold: the bet I chose did not match the risk to be covered, and I inaccurately estimated the probability of an extension. To estimate probabilities “Superforecasters” (Gardner and Tetlock, ISBN 13: 9781847947154) and “The Signal and the Noise” (Silver, ISBN 13: 9780141975658) both recommend the selection of a prior probability then incrementally modify it over time as news comes in.
This is good technique but is time-consuming and I have little free time. But it seems I will have to endure. The second part is more serious: the mismatch between bet and risk. My self-imposed restriction against online betting meant that I was extremely restricted in my choice of bet and I had to select one that did not fully cover the risk. Opening up an online account or even a telephone account would enable me to select a bet more tailored to the risk, but that is difficult as I would have to run it past Compliance and they do not respond quickly. But I shall consider it.
The extension of Brexit day leaves the risk of no-deal uncovered: I am exposed, in the jargon. The ERG will still agitate for no-deal and nobody to date has defeated them, so I still need to cover myself post-29th. I will look over the weekend for a new bet but it might be simpler to just buy dollars. At the time of writing £1=$1.31 according to XE.com, so it would prove a good hedge against a putative fall in sterling. As a matter of policy I am transparent in my betting and I note my bets here as they happen: it enables good feedback and better decisions. So I will let you know what happens.