Quant Chad

Polymarket's fees are now 36% higher than on Kalshi.

2 day ago Polymarket US replaced its 3.475% standard taker fee with a tiers based system.

You now need >$250k taker volume over the last 30 days achieve the same 3.475% rate.

Under that threshold, you will now pay a lot more.

Here’s what you now pay in for a buy at 50c based on tier:

- Under $10K -> 4.75%
- $10K to under $50K -> 4.25%
- $50K to under $250K -> 3.75%
- >$250K -> 3.475%

Kalshi’s standard rate on the same trade is 3.50%.

To think that Polymarket used to have 0 fees less than a year ago, and that Kalshi was considered the expensive platform.

FALSE. Prediction markets dont "converge" towards model prices.

it's the other way around.

I analysed Nate Silver's vs Kalshi's forecasts on 137 close races in the 2 month leading to the election.

at NO point in time did Nate's FLIPR model give higher average odds to the eventual winners, meaning Kalshi was always more precise.

If prediction markets really "converged" towards FLIPR odds then you would expect to see FLIPR lead in precision at each update, with Kalshi playing catchup.

It's pretty obvious why when you know how election models work.

All FLIPR essentially does is aggregate public information :
- weighted polls
- district partisanship and incumbency
- public expert race ratings.

These are all publicly available and used by @Kalshi traders before they have a chance to make their way into a Silver Bulletin forecast.

Sorry but you cannot get alpha by buying a $5 Substack subscription.

Kalshi is actually way more accurate than Election models.

I pulled data from 488 US election races from 2024 to now.

I compared Kalshi’s pre-election probabilities against :
- Silver Bulletin
- FiveThirtyEight
- JHK

Here are the average Brier scores:

🟢 Kalshi: 0.0282
🔵 Models: 0.0320

That translate to roughly 12% less average forecasting error for @Kalshi.

Some model outperformed Kalshi on some races (blue dots), and even on some categories (ex: JHK on 2026 senate), but overall Kalshi was way more accurate.

Kalshi's data also happen to be public and not locked behind paywalls.

I'm sure the comparison would be even more damming if Nate Silver's 2026 forecasts were publicly acessible.

@mansourtarek_

Does the Blue Wave mean the end of Prediction Markets ?

TLDR; It could be really bad for Polymarket but prediction markets will be ok.

here is why.

First we have to remeber that are in a bubble, to the average normie, Prediction Markets are issue #99 after AI, gas prices, War etc.

Yes Democrat politicians took some shots at prediction markets before :

- "STOP Corrupt Bets Acts" in March aiming to ban election and sports.
- 17 Democratic senators requesting restrictions against the CFTC.
- Michigan dems banning Polymarket US from sports.

But (1) Trump still has a veto, (2) except for sports, these got little traction.

So there is a risk of Prediction Markets losing access to sports, but not much else.

I would be more worried for @Polymarket in particular, because of the Trump connections

- Donald Trump Jr. is a major investor trough "1789 capital"
- Peter Thiel too through Funders Fund.
- Donal Trump Jr. sits on the advisory board

But as @shayne_coplan explains below, Polymarket accurately predicting the Dem victory could actually play in their favor 👇

AirBnB just started offering weather Prediction Markets to it's users.

Users can hedge bad weather at their destination by betting on rain levels.

this is a trillion dollar usecase.

here is how it works:

- users choose to pay a small "insurance fee" (stake) when booking
- payout is determined automatically based on the odds of rain
- If it rains, payout is automatically credited to their account

No claim, No verification, 100% automated.

Essentially, this is a prediction market bet disguised as insurance.

This is called a "Parametric insurance".

And it's what will take prediction markets from the $600B gambling industry to the $8T insurance industry.

Right now AirBnB is settling the bet through Chubb, a legacy insurance company, but they explicitly do NOT disclose their oracle anywhere.

I wouldn't be surprised if they were already using @Kalshi behind the scenes.

It would be incredibly easy for them to use Kalshi to :
- receive accurate weather odds
- handle market resolution
- even settle payouts

All the insurer would do is charge a premium over the odds and collect the difference.

If this usecase takes on, @Kalshi & @Polymarket could become the backbone of a trillion dollars industry.