ETHMilan26 - The Rise of Prediction Markets

Transcript

Hello everyone. So today we talk about the prediction market as David mentioned is a very new nent and uh interesting topic in crypto and not only so uh I think I just want to go through a bit like a history explaining for who doesn't know like what are prediction markets what's the status today uh and then uh do some reflection of what going to come so where we are today um like poly market and kali which are the two major venues both uh like US-based are have closed the April month with 23 billion of volume uh across all the

markets. So like a definitely not rookie numbers and if you can look at just two years ago it was only 80 million combined. So we can appreciate that there is really an exponential growth on these new uh type of uh uh markets and contracts. So um what we're going to talk about a bit of history uh when was the inflection point that make prediction market become uh viral and uh famous what can you do in prediction markets or what can you trade um a bit of critics about predictions or what people are saying what problems uh they currently have and what next and then uh just a

brief section about uh what DUN does uh for prediction marketing market. So a brief history. So what are prediction markets? Is a market where you can trade contracts whose outcome uh depend on uh uh future events. So basically uh the price of a market is their forecast. Typically, there are contracts that settle at zero or $1. And if you buy today that the price is 60 cent, that means that the participants are discounting a 60% um probability that the event in matter will will happen. Um so the price it's uh real time you can trade that every

second and the price moves instantaneously with uh any news any leak or any event that uh that happens and it's a very interesting uh concept the fact that you have now this uh very thick by thick like information uh drivers uh that are very different for example for polls or like even I don't know uh fed rates reports that are like once a month. So it's very uh it's very interesting how this uh information is going to be adopted and in the real world.

A bit of history. So poly market didn't uh invent uh uh the fire. So this was is a concept that is not new and it was actually almost 40 years old when uh an academic uh uh product uh the IO electronic markets first uh uh implemented this contract as a trial with a $500 cap and uh he bet the polls in terms of uh uh being able to predict uh what is going to happen on an event but then uh it just remained a project didn't expand. And then the 2014 there was another platform which was the first one um regulated by the CFTC which is

the commodity and futures agency in the US uh where basically you can always again trade on news but always with a cap. Then blockchain came in. So there was a new product called Augur. Uh maybe some of you remember it. It was on Ethereum. is actually very solid infrastructure but the UX was not really ready at the time for mass adoption. So it was a bit uh uh problematic and it eventually uh just uh didn't work and then poly market arrived in 2020 deployed on polygon using uh um optimistic or O oracle so what augur implemented uh with a central limit

order book and uh again it didn't pick up immediately but uh they settled the foundation for now this meteoric rise So when actually this uh uh product became famous uh we have one event in particular which is uh the 2024 uh US presidential elections when uh Trump was uh uh running against Kamalis and Poly Market really became like a one of the main information uh sources like a who's going to be the president like I was quoted by all the major uh uh US uh media companies and uh you can see here

exactly how it works. So you have an open uh the market open at a certain date and then you can see a time series of uh what are the prices so the probabilities of a candidate over the other to win the election and you can see uh Trump was favored for a long time then Kamala is raised and then eventually as we know Trump won. This was a pivotal moment because not only it reached mass adoption but also like a demonstrated that can work at scale. There are more than 300,000 users betting on this market uh for a total of almost $4 billion in volume.

Always in 2024 another participant it's called named Kalshi uh obtained a regulation from the CFTC again in the US to basically become a commodity and future uh regulated market uh it it was founded as well like a couple of years prior this but uh then after the regulation it suddenly like a spiked evolved and expanded to more markets and uh eventually now the latest data in April 2026 we can see that it makes almost 60% of the total volume also overtaking a poly market.

So now what what can you trade on this prediction market? So short answer would be everything but uh we can dig a bit more into it and we can see like I always using Dune data like we can observe what categories of markets have mostly traded uh sports in the US particularly is the main uh uh market but there is a decent volume also on crypto of course being poly market cryptonative and politics around like a 10 15 billion uh on those markets. But uh you can see that uh 10 1050 10 15 million uh total volume and you can see sports is the major uh uh is the major

category but uh this image it doesn't reflect like the real truth of what uh like the two how the two markets differentiate each other. So KCI is mostly on sports and where has most of the volume works but poly market instead is more uh distributed. There is a lot of crypto and politics as well and uh we're going to dig a bit deeper on see what particularly um this category look like. So sports um I mean uh interesting betting on sports but I think the interesting thing here is that uh by introducing sports prediction market didn't become only a eventdriven market

so once on every four year you bet on the presidential uh campaign and uh uh all the rest of the time you just uh don't use that but uh sports happen every day every week and by implementing these markets Kashi was able to get a very consistent volume uh of bets like every day of the week. So you can see here the NFL is one of the most uh traded sports. And also I think something interesting for us here in Italy is that Poly Market just reached an agreement to become the official and exclusive prediction market for Syria which is interesting as well as poly market cannot be used in Italy but uh that's another that's another thing and

uh of course you can take you can trade crypto uh a trend that we found uh particularly interesting is that uh is very trendy now to bet on these very shortterm markets. So like uh will the price of Bitcoin for example be up or down in the next five minutes? That's uh uh one of the markets that has the most volume on poly market. And then there is another number of categories of course like a politics, macroe events like a fed rate decision uh with the geop political condition also many markets related to wars and uh consequences of such uh but also like a

weather AI culture like a vision Oscar Grammys uh stocks earnings. So literally there are markets for everything. One thing to uh point out is that uh these venues are uh um most of the markets are curated by the venue itself. So the venue decide what markets to open. On poly market you can actually also create your own markets. But the reality of the fact is that uh uh typically markets created by users are very thin liquidity and not very traded. So 99.9% of the numbers I'm showing it's all stuff that is curated by the venues themselves.

So um of course it's a new product and it doesn't come with only like a benefits and uh uh nice numbers and uh funny stuff to use but also uh a few critics. So one of the most uh insisting one is uh insider trading. So now you have a public market where you can earn money on something and if you have privileged information uh you can trade on those and on poly market for example you can potentially remain anonymous so that's a very uh interesting conflict of interest and just to mention a couple of cases like a um a US soldier was um uh was arrested

in the US because was trading the same week that the president Maduro was captured. Uh trade that that week would have happened the the arrest and they made almost $400,000. Then they went back to recognize him and he was arrested. But there are many other cases and uh in particular lately like with the Iran war there were a couple of uh uh addresses always connecting to the same traders where um made a lot of money on a very high percentage like a win rate. So you can assume that people has privilege and information.

Um there are also other type of um uh problems. So so like a settlement dispute. So when a market concludes like a who says if the event actually happened or not uh there were a couple of cases where a market resolved in a certain way and there was not actually that clear that there was the actual resolution and uh it was definitely controversial because you can see that the numbers here are not like a few euros is a lot of millions uh at stake. And last but not least uh again like a A lot of this trading is said by certain studies to be uh wash trading. So you

can always be careful to see um like a report number. You always have to dig deeper a bit of like a what happens. Um yeah, so this is a bit like a an overview and I think now I want to just show a bit like a what's happening like currently these days in the in the space and uh what uh could happen. I think this year is going to be a very interesting year for this type of market. So we can try to see what bit what are the events or uh the catalyst of this uh of this industry.

So um for some of you might know what is hyperlquid. So is a perpetual uh trading venue on blockchain. Uh very popular. They just launched prediction markets as well. So definitely a bit another big offshore exchange that is uh going into the into the prediction market race. Uh there are other markets of course that I haven't mentioned so far but like a limitless predict are all like a onchain native markets that have a very high grow and very decent volume but not only like even uh uh traditional finance company are entering the space like interactive brokers uh possibly the most

famous broker in the world just launched a product called for forecast trader where they're going to focus mostly on like a macro macro and economical uh announcements and uh is just been bootstrapped today like in May 2026 and also Poly Market is expanding partnering with NASDAQ and other traditional finance institutions to for example offer private uh stocks uh trading on the on the on their on their platform.

what's need to change, what needs to be improved, what are the risks a bit as we mentioned like uh the integrity risk like uh insider trading is not just a a separate event is a thing that is now coming up very often in this uh uh when we talk about prediction market. So that's definitely something that uh to for the prediction market to establish as a information driver has to change and has to uh be addressed possibly with K KYC and anti-money laundering measures but uh also I think there have been studies specifically in the US that uh observe how prediction markets can be used to influence uh foreign policy and national policy.

So imagine uh I want to trade like a presidential US campaign like you can have a potentially foreign countries that spend money to uh change the optics of who can be the candidate rather than the other and we know that uh how the human mind is if you see something that uh is been said as having high probability of uh of happening then uh it's uh it might it's going to happen like maybe people don't do are not going to vote vote or they change their mind uh in that process and of course like from an infrastructure point of view as well like uh the uh dispute resolutions I mentioned but in general um there's

going to be a lot of work to do in how actually the mechanism of this prediction market uh work. So um where Dune fits in. So I just can give you a brief introduction of uh what Dune is for who doesn't know. So we are a blockchain data platform. Uh we have data for 100 and plus blockchains. We don't have of course only data for prediction markets. We have data for stable coin perpetuals dexes lending protocols like everything that you can uh analyze on blockchain is on dune. Uh we have a very established company been active since 2018. So um as the introduction say probably the best known

company uh in the space for data um and uh just working particularly on prediction markets like uh where was a lot of difficulties to to build these data sets because uh is a very different uh to put together data that is structurally very different. So poly market is purely blockchain data. Kali is not settled on blockchain. So it's fully just offchain. Uh they have very different categorization, very differently uh structure the market and uh with Dune we built a data set that uh normalizes and put together all these differences and make it very easy for

you to compare venues, compare like a markets and where the trades happen. So we do a lot of the work so you don't have to uh well just a bit of uh structure of how this uh data looks like. So we get data from sources that can be blockchain or APIs. We build um particularly three interesting data sets for each venue which are markets having uh reference data and details about all the market that are traded. uh trade table. So you have uh specifically all the trades that happen in that market as a time series.

So you can go back since the inception of the markets and see everything. You can back test strategies. You can see uh how this data could be able to predict actually future events. And then we have a candlestick prices. So as well you can see everything like a open, high, low, close, volume uh for that. And then we put together basically we union these data sets and uh we build the final uh prediction market uh data set that is very easy to consume. If you want to do reports, if you want to write news about it, if you want to uh build a prediction market data set and observe the a prediction market venue and observe the competition as uh what you need. Uh

yeah. So this is uh um this is a bit as I mentioned the the data set we've built and u thank you everyone uh I don't know what time is it so if anyone has questions yes please >> one second I accessing data and autonomously building strategies and operate in full autonomy. I know it's just a dystopian future but maybe it's not too far away.

>> It's not too far. I think I think yeah he builds his own algorithm in checks your data on your platform or other blockchain data platforms and takes decisions and and operate by itself. Uh what do you think is the scenario? what is likely to happen maybe now or in a few years. >> Yeah. Uh good question. I think that definitely is already solved the fact of accessing data. So Dune as a MCP CLI any like a agent can freely access or easily access data on Dune. They also now solved a bit like the part of uh if they these agents wants to actually connect to platform have accounts spend money.

Now there are a lot of protocols like uh X42 by Coinbase or where these agents can have their own wallets. They can have their own money rails. In terms of strategies um I mean they can potentially do it already. I think that uh it's early to say if these strategies are uh valuable like uh generally when there is not much history on uh uh maybe trade agents might have already started trading a few months ago but then doesn't mean that uh they can consistent provide alpha uh on these markets and uh uh be traded but I think that uh there are already all the tools to do it. It's just a just a matter of uh um directing

them maybe a bit uh in the direction and uh see what uh what a longer history can tell us. >> Another question. Thank you everyone.