What you need to know 🐳 Polymarket is the first on-chain predictive marketplace to achieve widespread success, with adoption extending beyond the traditional crypto sphere.
Polymarket's key challenge will be to maintain a strong user base following the end of the US presidential elections.
Presentation 🧬 Polymarket is a prediction markets platform launched in 2020 on Polygon, a layer 2 of Ethereum. These markets allow participants to bet on the outcomes of various future events, such as elections, sporting competitions or film releases.
Unlike traditional betting sites, participants bet with each other, eliminating the risk of a bookmaker defaulting. The odds are purely the result of supply and demand, with no arbitrary intervention by a central entity.
The underlying theory postulates that the price of a market incorporates all available information. The market evolves in real time, allowing users to exploit their expertise or privileged information. This mechanism would reveal new and unbiased information, as any attempt to manipulate it would be costly. These markets are therefore a valuable source of information for various analyses.
On-chain prediction markets offer several theoretical advantages:
Global accessibility with minimal friction, no registration or KYC Automated, continuous and verifiable real-time operation Integration with DeFi to optimise capital efficiency Automated resolution via oracle integration However, these markets face major challenges, including attracting sufficient liquidity and implementing a fair, transparent and rapid resolution method.
Polymarket largely dominates the sector in terms of usage and liquidity. Its success increased in early 2024, particularly around the US presidential elections. Numerous competing projects have emerged, either to compete directly with Polymarket or to specialise in specific niches.
Here is the evolution of liquidity present on prediction markets:
Currently, $363 million is deposited on all prediction markets, including $327 million on Polymarket. Financing 💰 Polymarket has raised a total of $74 million:
$4 million in its seed round in 2020, with participation including Polychain Capital, Parafi, Stani Kulechov and Balaji Srinivasan $25 million in a Series A round in January 2024, led by General Catalyst $40 million in a Series B round in May 2024, with participation from Founders Fund, DragonFly Capital and Vitalik Buterin Team and community 👾 Shayne Coplan, founder and CEO of Polymarket, is a 26-year-old entrepreneur who has been immersed in the crypto ecosystem for a decade. The idea for Polymarket germinated in his mind during the COVID-19 pandemic, taking inspiration from Augur, the pioneer of on-chain prediction markets.
In 2024, Polymarket bolstered its team with high-profile advisors. Among them, Nate Silver, founder of FiveThirtyEight and a highly regarded political analyst, brings crucial expertise in election forecasting for the 2024 presidential election. Nate Silver is joined by J. Christopher Giancarlo, former Chairman of the CFTC. Both play a strategic role in the development and reliability of Polymarket's predictive data.
Function ⛓️ Only the Polymarket team can create new predictive markets. This choice aims to avoid too much fragmentation of liquidity between markets and to only create markets that can be clearly resolved via a credible source of information within a defined timeframe.
As it has developed, Polymarket has expanded its market categories, becoming a generalist platform where you can bet on a wide variety of subjects.
In a prediction market, prices are not initially fixed. They are determined according to supply and demand. Participants, known as market makers, act like traders by placing limit orders to buy, for example, "YES" or "NO" shares at a set price.
Executing an order requires an equivalent bid on the opposite option. For example, if one trader offers to buy a "YES" share at $0.60, another must be prepared to buy a "NO" share at $0.40. The two offers balance out to reach a total of $1, allowing the order to be executed.
When the orders on the "YES" and "NO" options balance out at $1, they are converted into real shares: a "YES" share and a "NO" share are created and allocated to their respective buyers. This exchange price - in this case $0.60 for "YES" and $0.40 for "NO" - becomes the initial market price, reflecting the perceived probability of each outcome.
This mechanism continues with other traders, with prices fluctuating according to supply and demand, until they reflect a collective estimate of the probability of the outcomes in question.
A major challenge for prediction markets is to attract sufficient liquidity to offer traders the best conditions and reduce the spread between buy and sell prices in the order book. In particular, Polymarket offers daily rewards to users who place limit orders close to the average price. This incentive is particularly important for less popular markets.
In contrast to traditional betting, participants in prediction markets can exit their position before the market is resolved. They own shares whose price fluctuates in real time, allowing them to resell them - subject to sufficient demand - to make profits or limit their losses.
At market resolution, each winning share is converted to $1. A participant's gain is therefore the difference between their purchase price and $1, multiplied by the number of shares held.
It is notable that Polymarket does not currently charge any trading fees on its platform.
Resolving a market 🤔 Each market is resolved according to specific predefined rules. To resolve its markets, Polymarket uses UMA, an optimistic oracle. This means that, by default, it considers any information transmitted to be true, even though it may be contested.
Anybody can propose a market resolution. This involves depositing a guarantee, usually $750, which will be forfeited if the proposal proves to be incorrect. If the proposal is validated as correct, the proposer will receive a reward.
Once a result has been proposed, it enters a 2-hour challenge period during which anyone can raise an objection. In the event of a challenge, UMA's stakers have 48 hours to deliver their verdict and settle the debate.
Using an optimistic oracle greatly simplifies market resolution, as players are financially incentivised to propose correct assertions to UMA. As a result, the project reveals that only 2% of cases are disputed.
👉 For more details on oracles, see our industry analysis
However, this dispute resolution process can be complex and sometimes relies on the subjective interpretation of UMA's stakers.
For example, in the market concerning potential links between Barron Trump and the DJT memecoin, UMA had initially ruled that Barron Trump was not involved. Nevertheless, Polymarket re-examined this conclusion when new contradictory information emerged and ultimately overturned UMA's decision, refunding users who had bet "yes".
The Polymarket team can therefore intervene as a last resort. The challenge is to decentralise the power of resolution and to clarify the methods of judgement as much as possible, to prevent anyone from harming participants in markets involving several hundred million dollars.
A Polymarket token? 🪂 Polymarket doesn't have a token at the moment, and it's possible it will stay that way.
A possible token could nevertheless play a role in resolving conflicts or selecting which markets to launch.
Participation in the platform's various markets could be an eligibility criterion in the event of a future airdrop.
Ecosystem and integrations 🤝 More and more applications are integrating Polymarket's markets.
dYdX, for example, enables leveraged trading of the prediction market concerning the winner of the US presidential election. Ultimately, the project aims to integrate all Polymarket markets.
Ostium offers the possibility of automating the opening of short or long positions on various assets, based on the evolution of the probabilities of specific prediction markets.
However, the low liquidity of most prediction markets makes them vulnerable to manipulation. This risk is currently holding back their integration into derivatives.
Other prediction markets ⚔️ Azuro
Azuro adopts a model where data providers are responsible for providing the information needed to settle bets and help attract liquidity at the start of a market by taking positions with odds they initially set.
The protocol uses a hybrid model that combines odds issued off-chain with onchain betting streams. Each bet thus modifies the initial odds provided by the data provider, to reflect a balance between the provider's opinion and that of the market.
Azuro relies on an automated market maker (AMM) model to simplify the provision of liquidity across its markets. The idea is that its pools generate returns by providing liquidity based on the quotes provided by the data providers: the liquidity providers' returns are therefore largely dependent on the data providers' ability to offer quotes that are advantageous to them.
For the time being, the data providers are not yet elected by the DAO, but appointed by the team. Azuro's operation is therefore still largely centralised.
Drift
Drift is a decentralised trading platform initially specialising in futures products. The project recently announced the creation of its own BET prediction market, which aims to offer greater capital efficiency to its users.
It is possible to trade on BET using around thirty tokens as collateral and not just USDC or USDT. Users earn collateral interest that can be borrowed by other platform users.
Market resolution is handled by a security board elected by governance, again creating a fairly centralised resolution system.
MetaDAO
Contrary to most prediction markets, MetaDAO does not seek to compete with Polymarket, but seeks to specialise within "decision markets" i.e. prediction markets that are supposed to measure the repercussions of certain decisions.
MetaDAO thus enables various DAOs to create markets to evaluate the evolution of the price of their token depending on whether or not certain governance decisions are accepted: the idea is to let the market choose whether or not a particular decision is positive for an organisation.
👉 For more details: see our interview with Proph3t, the founder of MetaDAO
Regulation ⚖️ Crypto prediction markets face significant regulatory challenges, particularly around oversight, manipulation and policy implications. In the US, the Commodity Futures Trading Commission (CFTC) initially banned political prediction contracts. However, a recent ruling allows certain platforms such as Kalshi (non-crypto) to offer such contracts, although the regulations remain unclear. Decentralised markets, such as Polymarket, are vulnerable to manipulation by influential investors and could also face restrictions.
In 2021, the CFTC sued Polymarket for offering unauthorised binary options. The following year, the platform paid a $1.4 million fine and restricted US users to simply viewing its markets.
In Europe, the lack of a common framework leads to varying regulation from country to country. In France, prediction markets are considered gambling and regulated by the National Gaming Authority (ANJ). In the UK, they may be treated as financial products under the supervision of the FCA, although the status of decentralised platforms remains uncertain.
The MiCA regulation, scheduled for 2024-2025, could impose transparency standards on crypto platforms in Europe. However, it does not specifically address event contracts, casting doubt on the legality of prediction markets in the EU.
👉 For more details on regulation, check out our report on prediction markets
The Big Whale's view 🐳 The prediction markets sector has become really attractive, with many apps launching to try and cash in on Polymarket's popularity.
However, as with any market, liquidity is the key to success. It represents a virtuous circle for those who manage to attract it and creates a high barrier to entry for competitors. This dynamic really sets Polymarket apart - no application seems to be able to compete with it in any way at the moment.
New innovations, particularly in terms of improving capital efficiency, may attract a few users. Nevertheless, this does not seem to be enough to generate a real shift in liquidity between applications.
The other major challenge for prediction markets is to manage to find a transparent and decentralised resolution system. The aim is to prevent individuals from being able to take control and potentially manipulate the markets to divert liquidity.
It is highly likely that much of the liquidity will disappear with the end of the US presidential elections. However, Polymarket appears to have diversified sufficiently to retain a solid user base.
Other experiments, such as decision markets like MetaDAO, suggest new use cases for prediction markets. It remains to be seen how the various regulations will view these platforms.
