Launch HN: Kalshi (YC W19) – A regulated exchange for trading on events
Everyone faces different risks: inflation, extreme weather, mortgage rates, Supreme Court decisions, student debt, and more. Our mission is to allow people and businesses to mitigate risks that relate to them.
We got into finance early during our time at MIT. Tarek worked at Goldman and Citadel, Luana worked at Bridgewater and Citadel. We noticed something common across all these places: a lot of trading stemmed from an opinion on a future event. For example, a lot of activity at Goldman at the time was focused on providing institutions with exposure to, or a hedge against, Brexit. To do this, we’d sell them complicated financial bundles (a bunch of swaps, options, etc.) at a high price… but these bundles were proxies: basically, a bunch of risk curves fitted together to approximate the binary/event exposure the customers were looking for. What you couldn’t do was just trade directly on the event itself, even though that would have been simpler and cheaper and was what people actually wanted. The option just didn’t exist.
Once we identified the problem, we noticed it everywhere. The more we thought about it, the more the idea of an exchange for people to trade directly on events seemed obvious. That was the inspiration for Kalshi’s event contracts.
We give people the ability to trade based on their opinions about a specific yes-or-no question. For example, if you have student debt and are worried about relief not passing, you can purchase a contract and get a payout even if it doesn’t pass. If you’re worried about the economic fallout of another lockdown, you can place a trade to hedge against it. If you’ve developed a model on inflation, you can profit from that….and maybe even offset your rising costs.
Our markets are phrased as simple yes/no questions. For example, “Will the Fed raise interest rates next month?”. Users place orders on either the Yes or the No side and choose a price to purchase the contract between 1c and 99c. When there is a Yes order and a No order on both sides where the prices sum to $1, a trade is formed. When the event happens or doesn’t happen, based on criteria clearly defined in our contracts up front, we give back that $1 to whoever was right. The way our markets are structured means these trades are fully collateralized—there’s no margin or leverage on our platform.
The price that a given event contract trades at is actually the market’s assessment of the probability that the event will happen. If something happens that should increase the probability, people will start paying more for Yes contracts, raising the price. This market determination of the likelihood of an event has made Kalshi the most accurate place for predicting where inflation will be next month: https://kalshi.com/forecasts/inflation. People have been building forecasting dashboards on all sorts of things based on our data, including Bloomberg and news outlets.
Before Kalshi, markets that allowed you to trade on economically relevant events were illegal or unregulated. From day 1, we decided to take the harder path: build a fully regulated exchange with a deep commitment to compliance. We spent almost 3 years working with the CFTC to develop our exchange, contracts, market structure, surveillance, and compliance programs to provide a safe, secure and orderly market. This finally led to us getting federal approval to launch the first regulated exchange for this new asset class.
This path incurred the risk of launching later than anyone who opted to eschew regulatory approval. We chose the hard path because we want to develop this asset class in a way that is safe and responsible, which we see as the only way to build for the long term....
184 comments
[ 3.6 ms ] story [ 222 ms ] threadEdit: Found the answer, yes "Only US residents are currently accepted."
It would make it very transparent, if the exchange is honest. As a history of all bids, payouts and judgments would be recorded on a blockchain.
It probably also would do away with the KYC, so everybody could participate.
And nobody would have to give away their private data.
Thinking about it ... if the smart contract would let users post new bets as tokens and appoint judges, the whole thing would be eternal. Individual users and judges might come and go. But the platform would live on forever and the history of bets would be recorded forever.
See PredictIt, Polymarket, YC is doing a great job of fulfilling Thiel's vision of building monopolies. https://astralcodexten.substack.com/p/the-passage-of-polymar...
https://www.usbets.com/did-kalshi-kill-predictit/
https://threadreaderapp.com/thread/1478370047735341056.html
There were several such "decentralized" projects (Augur, etc.), but I don't know of anyone who uses them. I kind of stopped following the "permissionless" space after several projects turned out to be too scammy for my liking.
> It probably also would do away with the KYC, so everybody could participate. Nobody would have to give away their private data.
That's exactly why they're not allowed. If you can't easily get fiat in and out of such contracts, they can hardly matter.
Kalshi: not a novel idea, but it's good to see another implementation.
Several solid blockchain projects (also permissioned/KYC, but based on blockchain, with distributed oracles) are launching similar markets in 2023. I think they will have an advantage because all participating financial institutions will be able to create own markets without reinventing the wheel or building their own in terms of underlying infrastructure (APIs, smart contracts, etc.). Some already have simple apps (decentralized price feeds, etc.) in beta now - they just need to add the ability to bet on outcomes. I don't want to "advertise" their names here but you can look them up on the Internet.
You could even have trusted middle-men that collect a Fee iff ambiguity is discovered and an appeal is wagered against the escrow.
I'm sure it exists and got rugpulled already.
We, like traditional financial exchanges, define the contracts upfront with clear rules on how they will be settled, then follow the letter of the law very strictly when settling a market.
In general, we've found market certainty to be more important that accuracy: ie. having pre-defined rules that everyone can agree on and that are pre-set is more important than those rules being the "correct rules" (not that having correct rules isn't important).
By contrast, I signed up for Kalshi and bought 3 contracts in the last 30 minutes.
Additionally, smart contracts don't actually solve the primary source of (unaccepted) risk: the oracle.
It only took me a minute to use Polymarket, but I already had the prerequisites above.
I'm surprised if there wasn't already a pool of smart contracts out there basically already emulating the space.
The obvious first principled questions come to mind: definition ambiguity in the event and thus result, the Oracle problem confounds this by essentially either giving leeway to the market deciders or by using a concrete agreed upon source of truth, such as data published to .Gov domains and other government/market API's.
The need for human capital to moderate/adjudicate claims is apparent, but would vary per domain. No need for a human to decide on public CPI data (assuming no wrongdoing or mispublishing on the .gov part), and this could be further minimized by making clearer positions/events that eventually are glued to a dedicated Oracle API.
Definitely an area that could be implemented quickly but haphazardly in an unregulated/cryptobro fashion via tokens and smart contracts, but cool to see its been developed above the table.
Cool economic calendar too, I have been looking for exactly this. https://kalshi.com/economic-calendar
There are definitely areas where automating market resolutions/creating smart contracts can be useful....the most important thing is having a very well-defined contract upfront, so that outcomes are clear, the validators are authoritative, and unexpected edge cases don't come up. This is something we spend a lot of time thinking about when listing markets.
Note that Kalshi lobbied to shut down its competition (PredictIt), so overall benefits are not clear.
One thing I'd suggest: add pictures to the events. Makes it much easier for me to navigate and identify markets. I.e. a Biden picture next to a Biden event.
They’ll have a monopoly for a bit then. Would ‘t be surprised if Robinhood dips their toes in.
Any plans to support that space more deeply?
Who will win the 2024 presidency
GOP
Dem
Or the hundreds of scenarios that could have been bet on during the midterms.
Is the point of this platform to serve as a hedge against other asset classes or is this supposed to be thought of as a completely separate asset class and not as a hedge against other investments?
I see the gambling aspect of it of just betting on outcomes, and that is interesting to me; however, I am curious how your team thinks about it.
In your example, there's two things you could do depending on your usecase: 1) You think inflation will go up and capitalize on it. Today, you might think that a good way to do that is to short SPY. That's good, not great - because it's a proxy: inflation could still go up, and you SPY could go up as well (correlation is not 1:1 for a number of reasons).
The best way to express that view is by buying inflation event contracts: more direct, no basis risk, cleaner.
This use-case was super common when I was at Goldman and Citadel, which is where we got the idea.
2) You hold SPY but worry about the exposure of your holdings to inflation: you can use event contracts to hedge that exposure very precisely... think of it as a precise, meticulous surgery on your portfolio to eliminate (or even take) risks that are very difficult to eliminate with traditional instruments.
One other question, how did Citadel and Goldman do this before your platform existed? Were they big enough to just call up some bank and create a specialized product for them?
For example en the Moon landing example, I guess NASA will make press release, but it looks impossible to automate reading it and deciding. (What if the starship colides with the Moon? What if they can't return? What if it explodes 5 seconds after landing? What if it's a joint NASA+ESA project?)
Generally, event contracts are legal contracts technically. We have a team that spent a ton of time developing templates to incorporate all types of events (hard problem). Here's an example definition: https://kalshi-public-docs.s3.amazonaws.com/regulatory/ruleb...
This is hand-waving a bit, and not only doesn't address the OP's question (does a fatal collision count,) but is a good example of the ambiguities that could plague any "simple" contract.
I would think for ultimate compliance that Kalshi would be the first order-judges, but if the parties disagreed, these ultimately get decided upon by an actual Judge giving a good-faith determination of the question with a reasonable persons' definition of the terms.
Dang let me know if I should take them down/if they're super annoying.
This is branded as if it's some new type of stock exchange when in practice it's just another betting platform, no? Other than fancy UI, how is this any different in practice than oddschecker.com or betway.com (I'm not familiar with those sites, just a quick search)?
>inb4 all stock exchanges are betting platforms
Note in the post - the buyers who head to goldman to get exposure through a basket of positions are not "traditional" asset managers.
Put another way - the relevant asset managers were not restricted.
In gaming, you are trading on risks that are artificial, eg. betting on a dice roll or on a roulette spin. Those risks do not need to exist, and they risk solely for the purpose of betting.
In the derivatives markets, the risks you're trading on are not artificial, they're already existant. The purpose of markets is to transfer those risks from people that bear them and need to offload them (hedgers) to people that have appetite for them (speculators).
This debate has been at the core of these markets since the dawn of derivatives markets. See this fascinating Court Case on grain futures: https://en.wikipedia.org/wiki/Chicago_Board_of_Trade_v._Chri....
At the time, there was a lot of debate on whether grain futures were gambling. The final decision by justing Wendell was not because (paraphrasing) "sure a lot of people will speculate, but saying that these markets exist purely for speculation is off - there's a subset of participants that need to hedge risk, and that makes these markets crucial" (note: intent to deliver was the early form of hedging).
The original response:
"[T]he plaintiffs chamber of commerce is, in the first place, a great market, where … is transacted a large part of the grain and provision business of the world. Of course, in a modern market, contracts are not confined to sales for immediate delivery. People will endeavor to forecast the future, and to make agreements according to their prophecy. Speculation of this kind by competent men is the self-adjustment of society to the probable .... It is true that the success of the strong induces imitation by the weak, and that incompetent persons bring themselves to ruin by undertaking to speculate in their turn. But legislatures and courts generally have recognized that the natural evolutions of a complex society are to be touched only with a very cautious hand. .... It seems to us an extraordinary and unlikely proposition that the dealings which give its character to the great market for future sales in this country are to be regarded as mere wagers or as ‘pretended’ buying or selling, without any intention of receiving and paying for the property bought, or of delivering the property sold, within the meaning of the Illinois act"
But in a commodity market there are real participants whose primary business model is production, being able to hedge against future price swings provides real value to these businesses.
(despite the snark, I'm also curious how the situation is different)
The entire local economy surrounding sports venues. For a small bar or restaurant near a stadium, the difference between a good season and a bad season can be huge. This is magnified to the nth degree in places like Green Bay, WI where renting out one's home for a single at home post-season game can easily net a family several month's worth of mortgage payments.
Considering how Capitalist governments behave in financial crashes, you mean:
"offload them to people who have appetite for slanted bets, or in case of significant losses, to the public who has no appetite for it."
Was there ever any formal waiver or guide lines?
As a type of investment though, there's nothing wrong with it - if risk-mitigation is what you need.
If the binary nature was a problem they could easily structure these to be non-binary, such as bet on a game and don't return 1/0 but the actual game score 21/15.
I guess the number of markets we post next year will be a function of how crazy 2023 will be. In some ways, we capitalize on volatility.
Antifragile...
We're a derivative exchange regulated by the CFTC. We're no more gambling than grain futures are!
There's a number of unemployment-related markets that allow you to create insurance for yourself against unemployment.
And these are just things I've thought up of in a couple of minutes, not even considering what actual financial markets might want to use them for.
[1] https://kalshi.com/events/FRMMAX-22DEC29/markets/FRMMAX-22DE...
Also even if you can't hedge the exposure in its totality, it is still worth hedging a fraction of it. "Under-hedging" is a common term in commodities markets - people often want to cover a portion of their exposure and leave the rest in the hands of mother nature.
Bureau of Labor Statistics could announce a new equation for the CPI next month. Which means the inflation will be calculated to be much lower than everyone expected.
If you're betting on Bureau of Labor Statistics press releases, you just lost your bet just because a human decided to word things differently.
If you're investing in a complicated financial bundle, like actual contracts to buy sell oil, wheat, stocks. Then you could still be winning. Because a single politician can change your win to a loss when you're betting on words (like Inflation as calculated by a person in the US government), but a single human can't lose you money when you're investing in actual equities
This is more like a roulette table than an actual investment
1) Hand-wavy exchange of money. Most of the derivatives market, including some of the most traditional instruments like energy Futures, have massive amount of cash-settled activity, rather than physically-settled -- ie. nothing physical is actually exchange hands, and traders are purely exchange financial risk.
That doesn't make the transaction any less important and valuable: at the end of the day, participants are hedge financial exposure and cash-settlement is a great and more efficient way to get that hedge in.
2) Single-point of failure OR arbitrary change to the underlying. There's an extreme amount of scrutiny and safeguards around how CPI is calculated and how it is changed. CPI impacts trillions that are traded in traditional assets like interest rate swaps, inflation swaps, mortgages, etc. Also, CPI is a large factor in the Fed's decision to raise interest rates at every meeting. It cannot be changed by a single person, on a whim. And this tends to be true for all the data sources underlying our markets.
Arguably, a stock has much more key man risk (or "arbitrary whims risk") than something like CPI: eg. Elon ripping a bong and tweeting something's impact on Tesla stock.
3) More like roulette than actual investment. The lack of physical underlying doesn't make this any less of a financial instruments. Most liquid markets today do not actually exchange the underlying, eg. interest rate swaps, index futures, etc.
What differentiates a financial product from gambling is the presence of an economic purpose. A roulette spin does not need to happen - it happens solely to create an artificial risk for people to bet on. In the case of event contracts, things like an election or a CPI print already expose the market to risk... that risk already exists and our markets allow the transfer of said risk from people that have and can't bear it to people that have the appetite to bear it (that's actually the whole point of the commodity futures and derivatives market).
Totally untrue. PredictIt was legal and regulated and well-loved for many years, operating with the permission of the CFTC under a no-action letter. Then Kalshi hired former CFTC commissioner Brian Quintenz and soon, PredictIt was suddenly deemed by the CFTC to have committed still-unenumerated violations of the no-action letter.
This is a rotten way to do business and shame on YC for being involved in it.
https://karlstack.substack.com/p/a-textbook-case-of-regulato...
* because https://hn.algolia.com/?dateRange=all&page=0&prefix=false&qu...
(edit: for anyone wondering, I didn't flag the GP comment)
Regulation in this space is really tough and thorough - we've been battling through it for years now (also explains why this post is coming so long after the end of our batch!).
Like this most people will have to assume you really just paid off / bribed Brian Quintenz.
https://astralcodexten.substack.com/p/mantic-monday-81522
> If someone from Kalshi wants to swear, in so many words, “We promise we put no effort into convincing CFTC to quash PredictIt”, I will believe them (although I would still suspect the CFTC was following a thought process like “now that Kalshi exists it would be embarrassing to let less-regulated markets exist alongside it”). Until then, I think cui bono remains the right question.
Here's your chance, 'tmansour: if you didn't do anything wrong, why not say so yourself?
I have no horse in this race or involvement with either party at all; I'm just interested because this feels like quite the betrayal of the usual ethos we'd expect from YC affiliation, and your responses on the topic so far are evasive enough to be a red flag in and of themselves.
Respectfully, I think you should have stopped there, or perhaps earlier. There is quite a body of evidence already provided in this thread suggesting that Kalshi is an anticompetitive regulatory-capture play.
Just because we moderate HN less when YC or a YC-funded startup is part of a story does not mean I won't respond like a normal human being.
* Yes, the revolving door is bribery-by-any-other-name. Anger at this type of legalized overt corruption of our institutions is literally ripping the country apart.
* be small-scale and not-for-profit
* be operated for academic and research purposes only
It doesn't really say what "small-scale" means, but obviously, if PredictIt keeps being successful and growing, eventually they will no longer be small scale. PredictIt was always in a situation where they had to make a different deal with regulators, eventually, as long as they continued to grow.
If you want to blame the CFTC for shutting down PredictIt, Intrade, all of those other prediction markets, then yeah, I agree! I wish they had just been much more permissive long ago. We should blame the CFTC for being too strict. But it doesn't seem fair to blame Kalshi just because they might be the ones to finally convince the regulators to allow one of these.
I hope Kalshi is the group to figure this out, make prediction markets popular, and that prediction markets can finally get the attention from wider society that they deserve.
Brian Quintenz is a Republican financial manager who was nominated to be a commissioner of the CFTC by Trump in 2017.[1]
During his time at the CFTC, as I recall, there was heavy bipartisan action, which he was praised for.
Well, there's a slightly more cynical take that he was indeed completely crooked.
My memory is foggy, and I don't have time to grab sources just yet as I have a meeting coming up so I'll update this shortly, but he was then turfed out of the CFTC under Biden and promptly hired by Kalshi.
Given his record, it's less of a jump to see what Kalshi is doing as shady, but more of a smoking gun... PredicIt, which had been operating under a no action letter (NAL) from the CFTC since 2014 just happened to have that letter withdrawn a few months after Kalshi raises $30m at a Series A to take their platform live.[2]
From my point of view, this is classic American lobbying mafia style stuff. Prediction markets have been prevented from flourishing for decades, and now that the time has arrived, the crooked CFTC et al have cleaned the house so their chosen startup that they're deeply in bed with can thrive.
The whole thing, quite frankly, stinks.
[1] https://en.wikipedia.org/wiki/Brian_Quintenz
[2] https://www.pymnts.com/news/international/2022/millions-comm...
Unfortunately, this seems to be endemic of American politics and government organizations.
In Kalshi's case, Brian Quintenz aside, let's talk about Jeff Bandman, who spent close to 3 years at the CFTC.[1] Well, after doing so, he became a regulatory strategy advisor for Kalshi.
Here I quote a Bloomberg article[2] on Kalshi from earlier this year:
"Eventually they tracked down a former CFTC official, Jeff Bandman, who assured them the landscape was changing; he agreed to help them navigate the agency and its characters."
So yeah, I'll concede that there are a few un-generous assumptions that need to be made to fully paint Kalshi as a bad actor in this situation, but given the various pieces of context available, I don't feel it's an overstep to call behavior like this out.
Ultimately Kalshi is set to be a unicorn startup, and I wish the founders success in their endeavors—they appear to be exceptionally bright and hard working individuals—but what appears to have transpired for them to have their shot doesn't sit right with me.
[1] https://www.linkedin.com/in/jeffbandman/
[2] https://www.bloomberg.com/news/features/2022-05-26/kalshi-s-...
Archive link as the original has a paywall:
https://archive.ph/20220527010753/https://www.bloomberg.com/...
If PI were operating on the basis of a no action letter they were unregulated pretty much by definition. A no action letter is basically a regulator saying that it won't take action against a market participant for doing things that would otherwise merit regulatory action.
(I'm not affiliated with any of these parties, just going by what I read in the comments and the various publications from CFTC.)
And predictit will have its day in court [1] where both sides will actually need to present evidence, unlike on an internet forum
[1] https://www.globenewswire.com/en/news-release/2022/10/06/253...
Because it was one of the longest running and largest prediction markets to ever exist.
> trade-size limited
Kalshi is trade-size limited, too.
Which is definitely a significant difference, but still seems too small for cases like tmansour’s motivating example (“providing institutions with exposure to, or a hedge against, Brexit”). Though I guess it depends; if you bet $25,000 on a 99:1 long shot, then your position could be worth $2.5 million if it pays off.
(PredictIt also limits each market to 5000 participants, while as far as I can tell Kalshi has no such limit.)
Here's more about our regulatory status: https://kalshi.com/learn/how-is-kalshi-regulated?
We're fully regulated by the federal government as a derivatives exchange, and that's a big difference in an of itself (it took us years to figure out the right model to offer derivatives dynamically, with events as underlying). Regulation allows us to plug into the financial ecosystem, and offer the asset class to hedge funds, market makers, brokers, etc.
For all intent and purposes, we're a financial exchange that offers derivatives on a broad range of things that have been offered before.
We are fully regulated by the CFTC. Our ethos has been do regulation from day 1 and we spent 3 years getting regulated before we launched a single product. Regulators are already caught up and are working with us constructively to expand this marketplace and asset class.
https://en.wikipedia.org/wiki/Intrade
We're doing a fully regulated, full scale version of Intrade. Regulation allows us to take this to the next level: getting market makers onboard, integrating with brokers, etc.
Edit: the crypto part isn’t true
Though we've built our exchange from the ground up, we're focusing on innovating on the markets side, rather than the financial plumbing side.
We took the approach of addressing the elephant in the room first. We spent 3 years to crack the regulation case and we got federal approval to make sure we have the potential to get this market to its full potential.
So, in a way, this is a good and tricky questions that applies much more broadly to non-securities markets (e.g. commodity futures). (IANAL)
1. we spent years working with the regulators on defining and building our surveillance systems. They basically ingest data from the exchange and run stats/some ML to flag suspicious trading patterns to an investigation team in our compliance department (similar to when NYSE flags a Goldman trader for insider trading) -- our systems have gotten really sophisticated and you'd be surprised by the amount of commonality there is in cases of fraud/insider trading/manipulation/collusion and so on...
2. we tie people's trading activity to KYC we run at signup. We also pass people through Politically Exposed Persons (PEP) lists, that flag anyone that works in gov and their relatives etc.
3. when we investigate cases of inappropriate behaviors, the consequences can range from fines to criminal prosecution by the CFTC (similar to stock trading)
4. obv all the above are not a single hammer solution, they're heuristics, but people generally don't commit a federal crime to trade with $100, they tend to trade with much more meaningful sums, which fortunately and intuitively is much easier for our oversight programs to flag.
Overall, this question presents a number of fascinating challenges/questions, but it's not a more difficult problem than flagging insider trading/market manipulation in traditional markets like stocks and commodities.
Even as a relatively sophisticated investor (compared to the median) I wouldn't really know how to hedge correctly on these markets.
For example if inflation is high, how many units do I buy. Am I fully hedged, or can I lose the bet AND lose in real life?
I think users need to understand these things to use it. Some of that is education, and some will be tools.
For example, enter your salary, expenses, budget etc, and it recommends a bet, and then shows you the outcomes if you win or lose.
It should be clear that for example YOUR cost of living might go up more than inflation because you buy different things from the inflation basket.
This is a bit like insurance, in the sense you may pay for health insurance, or travel insurance then find your thing isn't covered.
I would also be interested in the tax implications. For example you do a bet to hedge your foreign holdings of say GBP. GBP goes down so you win from the bet and lose in real life. But you still pay tax on the bet (or do you ... ?).
I think this is a cool idea.
I think the more useful idea for the everyday person is something akin to a "1kg of tomatos a week prepaid subscription for 5 years" aggregated over their typical shopping.
That all said trying to tame the future is a hard beast. And expensive to insure yourself against all the things. Especially likely things, like shit getting more expensive!
Just wonderful!
While it's true to some extent, insurance is indeed crucial: one of the counter-parties is meaningfully reducing the risk of catastrophe, or extremely detrimental outcome at the very least.
This can be extended to all sorts of events, including Covid, the economy, politics (Brexit had devastating consequences), hurricanes, etc.
"Biden approval rating" https://kalshi.com/events/538APPROVE-22NOV23/markets/538APPR...
How is that comparable to an insurance and not just simple betting?