6 Comments

*Nonetheless, if I were a betting man, I’d say there’s value to be had in betting ‘Kamala yes’ right now given the prediction markets’ divergence from election models’ odds.*

For individuals in the US at least the tax code eliminates most or all of the value from the divergence. Winnings are taxed at ordinary rates, while losses are not deductible. For a top-bracket CA resident, buying a 50-50 bet for 40 cents is barely breakeven.

Professionals can deduct gambling losses. But I suspect most professionals are looking to make lots of small bets, each with a little edge; not a single big bet on a coin-toss once every four years.

Expand full comment

Great points all worth considering! Although there are probably numerous ways to circumvent this tax liability by transferring crypto around to different jurisdictions and methods of that nature. Not particularly well versed on all of this, though.

Expand full comment

how do you explain that the betting markets and polls way underestimated trump in both 2016 and 2020 and what was done to correct for this bias? Perhaps if you suggest a gold standard method rather than cautionary notes about every method currently out there, we can assess confidence in betting market odds vs. polls vs. Nate silver, etc.

Expand full comment

Great question. As I understand it, the '16 & '20 polling errors regarding Trump were probably driver (to some extent) by "shy Trump voters,” and (to an even larger extent) by nonresponse bias: Democrats are more likely to respond to political surveys. I've heard that pollsters are very aware of these past biases and have been changing their methods in response. Nate Silver posits that "If polling firms were still applying the same techniques they did in 2016 and 2020, we’d probably be seeing a Harris lead in the Electoral College right now. Instead we have a toss-up, more or less." ... Covid obviously had an impact in 2020 as well, in that there were a disproportionate number of voters voting early and mailing in and things of that nature. My gut says that the polls will be less erroneous this time around than they were in '16 and '20, but I'm no expert.

As for the betting market odds, it's hard to make the argument that they aren't being distorted. If not because of Fredi whale with seemingly infinite liquidity, distortion could also be a product of bettor demographics: gambling skews heavily male, and makes skew heavily toward Trump. Again, there's more to it and probably we won't get the full picture until after the fact.

Zooming out, I think we should lend credence to the fact that all indicators (polls, prediction markets, "vibes," etc.) are trending in Trump's direction. Sure, there's a divergence between the betting markets and polls, but they are moving in the same general direction.

Hope this helps, and thank you so much for commenting. Cheers!

Expand full comment

However it is now entirely possible there are shy Harris voters who are either former Indy or Rep who will not say such due to fear of reprisal from family or co-workers.

And of course… the betting markets have a job - Separate people from their money. And MAGA fall for every scam there is. Worth noting Peter Thiel has a controlling share in Polymarket.

Expand full comment

All good points. Really hard to tell which direction the polling error will fall until after the fact. I feel a significant factor is still the fact that democrats tend to be more inclined to answer polls than republicans. (This is independent of notions around shyness of Trump/Harris voters.)

Expand full comment