This week in prediction markets:

• Harris primed for post-debate polling "win" after Trump odds survive highly unprofitable trading strategy.
• Kalshi bettors say Harvard applications will fall as Gay leaves a dark cloud.
• No Canadian election in 2024.

The second U.S. Presidential debate, and the first with now Democratic nominee Kamala Harris, is scheduled for Sept. 10 on ABC.

With the total amount bet on the Polymarket election contract approaching $850 million in the days leading up to the debate, bettors are still giving former President Donald Trump an edge against Harris, with the odds at 52-46. Polymarket is a blockchain-based platform, programmed onto the Polygon network as a smart contract, and bettors ante up by depositing the stablecoin USDC, a type of dollar-linked token.

Then it gets a bit meta, with prediction markets about polls (something prediction markets are supposed to be an alternative to) and even prediction markets about other prediction markets.

Polymarket traders are betting the traditional pollsters will give the debate to Harris, with a 75% chance that the Ipsos/538 survey will find she "wins" it.

At the same time, bettors also have their money on this "win" not mattering much.

Harris, in their eyes, only has a 22% chance of retaking the lead in the main Polymarket election contract the day after the debate. (Yes, there are Polymarket contracts about other other Polymarket contracts.)

The debate is also expected to draw an audience of well over 70 million people, which may put it in range of challenging the first Clinton-Trump debate, which holds the record as the highest-rated at 84 million, followed by the second Carter-Reagan debate , which had an audience of 80.6 million.

Meanwhile, the market seems to have righted itself after a very expensive failed manipulation attempt over the weekend.

The manipulation, on-chain data suggests, involved a group of whale accounts, led by DJTHolder , which apparently tried to artificially juice Harris' odds in the U.S. Presidential Election Winner market to profit from a derivative market that resolved based on the leading candidate on Sept. 6.

This group burned through $9 million of USDC and simultaneously bought large amounts of Kamala YES shares and Trump NO shares, temporarily driving Harris shares' price by anywhere from 2 cents to 27 cents. (Each share pays out $1 if the prediction comes true and zilch if not, so Harris' current 47 cent price indicates a 47% chance of victory.) The price subsequently fell back to Earth.

Despite spending millions, they failed to keep Kamala’s odds above Trump’s, leading to over $60,000 of USDC-denominated losses in the derivative market and six-figure losses in the main election market.

Right now, the apparent ring leader, DJTHolder, is showing a total loss on his account of over $46,000 based on $2.9 million in positions.

A highly profitable trading strategy this was not. They probably should have tried this first in the Minecraft version of Polymarket.

School takes hard knocks

The legacy of former Harvard President Claudine Gay, who departed this year over plagiarism allegations and an increase in violent anti-Semitism on campus, still hangs dark over the institution she once presided over.

Bettors on Kalshi are predicting another drop in applications for Harvard in 2025, for the class of 2029, forecasting the number to come in below the 54,008 undergraduate applications received in 2024 for the class of 2028.

While other elite institutions like Yale and Columbia reported increases in their application numbers, Gay’s legacy remains a burden. To her harshest critics, her legacy embodies everything that is wrong with post-secondary academia.

Many, including the university's largest donors , would like to see changes at Harvard, including more ideological diversity, and the dismantling of racial preferences in its admissions policy. While it seems that Harvard has acknowledged this and has begun to make changes , the market is still betting this won’t be enough to sway things meaningfully.

With all this talk of prediction markets and Harvard presidents, remember that one of the journalists whose work on accelerated Gay’s departure lost money betting on it on Polymarket .

Read more: He Exposed Harvard President's Plagiarism, Then Lost Money Betting on the Story

Kalshi, which unlike Polymarket is regulated in the U.S. and settles trades in dollars, won a court fight last week over its plan to offer election contracts. The Commodity Futures Trading Commission has asked the judge to delay the listing another two weeks.

Read more: U.S. Election Betting Delay Would Be 'Devastating' to Kalshi, Firm Says

Canucks shook

Polymarket bettors are giving a 22% chance of a Canadian election happening in 2024, despite a recent political shake-up, which makes it a possibility on paper.

Justin Trudeau won the last Canadian election, which took place in September 2021, with a minority government.

In Canada’s Westminster system, inherited from the United Kingdom, a minority government can fall if it loses a key vote in the House of Commons, like on the budget or a confidence motion, because it doesn't have enough seats to pass laws without help from other parties.

To prevent the Conservative party, led by an increasingly popular Pierre Poilievre, from triggering a motion of non-confidence that would call an election, Trudeau’s Liberals entered into something called a Supply-and-Confidence Agreement with the leftist New Democrat Party (NDP), which guaranteed a supply of votes should the Liberals support various NDP motions in Parliament.

That agreement ended last week .

The NDP ended its agreement with the Liberals because of frustration over the slow progress on pharmacare, a program to provide affordable prescription medications for all Canadians, and their opposition to how the Liberals handled labor strikes, which went against the NDP’s pro-union stance.

They also wanted to distance themselves from the unpopular government before the next election, and Canada’s CBC television network reported that the decision was made to pull the plug in June as Poilievre’s numbers surged.

So does this mean that the Trudeau government can fall with the next vote in Parliament? Not necessarily.

The Bloc Quebecois, a party that sits in Canada’s federal parliament but has the sole mandate of getting the best deal for Quebec, says it is willing to make deals with Trudeau’s Liberals to get more support for their pet issues like increased immigration powers for Quebec and more funding for the province’s infrastructure and seniors.

There’s also the option to "prorogue" parliament , which involves asking Canada’s governor general , the King’s representative in the country, and head of state by legal formality, to put a temporary pause on things and reconvene the House at a later date.

While there’s plenty of precedent for doing this, Trudeau did it in 2020 to fend off a growing scandal , and former Conservative Prime Minister Stephen Harper did it twice, once in 2008 to avoid a non-confidence vote and in 2009 to delay investigations into an Afghanistan detainees controversy and strengthen his position during the aftermath of the Global Financial Crisis. The maneuver is immensely unpopular with voters.

It looks like Polymarket bettors think Quebec has all the cards, and will keep Parliament functioning and Trudeau in power. But right now, this is a thinly traded market, and the balance might shift when more liquidity enters the fray.