The Rise of Decentralized Prediction Markets
What they are, How they work and Why they are the Future
Mutual contracts, agreements, transactions and the recorded database of them are some of the defining structures of our economic, legal, political and financial systems. They facilitate transactions and govern interactions among nations, organizations, communities, and individuals. They set organizational boundaries and establish and verify identities and chronicle events. Unfortunately, these critical tools of administration and human society have not kept up with society’s digital transformation.
Blockchain technology attempts to solve these problems, and it is becoming increasingly popular these days. In a nutshell, blockchain technology relies on a chain of blocks that contain data or information. It is an open, decentralized and distributed ledger on a peer-to-peer network that can record transactions between two parties efficiently and in a verifiable way. Interestingly, this distributed ledger can also be programmed in a specific way to facilitate transactions automatically. Since everybody on the network has a copy of the distributed ledger with them, no central or single entity holds the authority of the overall network, nor they can modify or corrupt the database for their own vested interests.
This feature of blockchain technology enables decentralization and helps to maintain the transparency and security of the overall network. The technology is very dynamic and is transforming multiple industries as new use cases for decentralized networks, products, and services continue to emerge. Some applications are secure sharing of medical data, decentralized voting mechanisms, supply chains & logistics monitoring, cross border payments, better weather forecasting, improved auditing and so on. The prediction market is one such sector that is rapidly adopting blockchain technology as new decentralized platforms are emerging to challenge the existing centralized prediction market platforms. Let’s dive right in and see how it works!
What is a Prediction Market?
A market is a place where two parties or two individuals can gather to facilitate the exchange of goods and services. This can be physical goods such as vegetables or consumer goods, services as in Uber or financial assets in the stock markets. Essentially, it is a group of people (and robots?) buying or selling things.
Similarly, a prediction market is a marketplace where people can buy and sell predictions of future events. These are exchange-traded speculative platforms where traders can place their bets on the outcome of future events. In a typical stock market, you buy/sell shares of a company but in a prediction market, you get shares of an event outcome. This could be any future event such as whether it will rain after 2 weeks, whether India’s economy will go up next year, whether Lionel Messi will win Ballon d’Or in 2021 or whether Donald Trump will win the 2024 presidential elections?
Prediction markets are also known as information markets, predictive markets, decision markets, idea futures, event derivatives or virtual markets. Some people refer to it as “legalized betting” as they provide an opportunity to make predictions about wide-ranging events and topics of interest, but in reality, it is much more than that!
The main purpose of prediction markets is crowdsourcing information from the general public to obtain and aggregate beliefs about the outcome of an event. Subsequently, traders then bet on the binary beliefs that they predict to be the final outcome. If the bet of a trader turns out to be correct, then he/she/they earn money from those who bet incorrectly but if their predicted bet is wrong, then they lose the wagered amount. The ultimate payoffs are determined by the proportion of traders that bet on a specific outcome.
How do the Prediction Markets work
As discussed earlier, in a typical prediction market, you essentially bet against the market how an outcome will turn out for a future event. Let us now understand how the prediction market works in detail with the example of US presidential elections, a very popular prediction market in the present era.
To begin with, there are two types of shares in a prediction market: “YES”, or long shares and “NO”, or short shares.
Every YES share pays out a dollar if the event in question occurs in the future, and the worth of your share becomes zero if the event does not occur.
Every NO share pays out a dollar if the event in question DOES NOT occur in the future, and the worth of your share becomes zero if the event occurs.
But which factors determine the price of these shares? Is there any central or single entity that dictates these prices? Well, these shares are valued at a price determined by buyers’ willingness to pay and sellers’ willingness to accept a certain amount. This amount is directly proportional to the probability of the event outcome in question. Just like how stock prices represent the aggregated investors’ prediction about a company’s future earnings, the prediction shares are priced according to supply and demand and represent the aggregated probabilistic predictions of an event.
If the aggregated belief of the overall market consisting of numerous investors is that the probability of a certain event happening is 70%, then the YES share will be priced at $0.70 and the NO share will be priced at $0.30. Subsequently, if the event occurs, every YES share (worth $0.70) you purchased will be worth $1 and NO shares will be worthless.
Still confused? Let’s go through an example for a better understanding.
Event A: Will Donald Trump win the 2024 presidential elections?
Event B: Will Joe Biden win the 2024 presidential elections?
Suppose we create a prediction market for the 2024 presidential elections for two events as described above. After setting up a market, traders can now invest a certain amount, say $1000 and receive 1 Event A token and 1 Event B token. If the aggregated expectation of the overall market is that Trump has a 60% probability of winning the presidential election then the share of Event A will be priced at $0.60 and on the other hand, Biden’s prediction share will be priced at $0.40 given that the probability of Joe winning the election is 40%.
However, these shares/tokens can further be sold freely to other investors. If you are not convinced with the markets’ probability distribution, then you can buy the desired (subjectively undervalued) share and sell the overvalued share, which will impact the prices. Eventually, as more people buy or sell more shares as per their own subjective analysis, the prices of these shares will fluctuate depending on the combined information held by the market. Ultimately, if Event A i.e., Donald Trump wins the 2024 presidential elections, then the tokens of Event A will be worth $1 and Biden’s tokens will be worth zero.
The Importance of Prediction Markets
Prediction markets are not a new phenomenon. They have been existing for centuries. As this paper from the Massachusetts Institute of Technology describes:
Political prediction markets date back to the sixteenth century, when betting on the next pope was considered common practice and banned by Pop Gregory XIV in 1591. Gambling odds were printed daily in newspapers such as the New York Times in the early twentieth century. They only declined in popularity due to the advent of scientific pooling before interest was again re-kindled in this area.
By now, I am certain that you perceive prediction markets as just another form of gambling or some exaggeratedly glorified gimmick that investors throw their money at, but in fact, they are considered to be quite valuable to make accurate predictions of future events. You would ask, “How are they even different from gambling if both are essentially zero-sum activities”? Well, in the prediction markets, the payoff depends on the accurate prediction of an outcome of an event while in gambling, luck plays a major role in the final outcome.
Back to our topic, why these markets are necessary? What is its significance? The main objective of prediction markets is the aggregation of beliefs over an unknown future outcome. A market is a place consisting of various players having a wide variety of knowledge, thoughts, and opinions. Prediction markets make it possible and easier to advise policy decisions by giving more accurate estimates of the “aggregate consequences” of those decisions. The economic incentives provided by the prediction markets motivate the participants to do extensive research, conduct an analytical study and taking some efforts before investing their money.
Individual players are allowed to take advantage of proprietary or secret information about the future event and turn it into a profit without revealing the source or content of the information. The market punishes and discourages those players by imposing costs if they fail to conduct proper research before coming to the most accurate conclusion.
Prediction markets follow the concept of “Wisdom of the Crowd.” Wisdom of the Crowd refers to an idea that a group of people hold a more substantial, important and wider range of information on the market as a collective than individuals would as singular or central entities. It allows extracting the necessary information from different corners of the society to arrive at the nearest forecasts about a future event. In some instances, the markets not only extract existing knowledge in the decentralized parts of the society, but motivate the generation of new knowledge. For example, prediction markets on the weather might motivate meteorologists to develop improved forecasting models.
Prediction markets are not only limited to areas of general interest such as political predictions, weather forecasts etc. but are increasingly being adopted by corporate companies such as Google, Ford, and Microsoft. In a research paper titled “Corporate Prediction Markets: Evidence from Google, Ford and Firm X” the researchers concluded that forecasts from predictions markets outperform other forecasting tools available to the management companies. They also highlight that prediction markets get better and more efficient as time passes by. An example of how prediction markets can be useful in companies is portrayed below:
A 2008 paper titled “The Promise of Prediction Markets” authored by several Nobel Laureates and Google’s Chief Economist made the case that prediction markets can let governments and businesses make better forecasts and policy decisions. As the paper describes:
The range of applications is virtually limitless - from helping businesses make better investment decisions to helping governments make better fiscal and monetary policy decisions.
In his book “The Wisdom of Crowds”, James Surowiecki recognizes the power of prediction markets as a tool to capture the collective insight of the people. As he explains:
We generally have less information than we'd like. We have limited foresight into the future. Most of us lack the ability - and the desire - to make sophisticated cost-benefit calculations. Instead of insisting on finding the best possible decision, we will often accept one that seems good enough. And we often let emotion affect our judgment. Yet despite all these limitations, when our imperfect judgments are aggregated in the right way, our collective intelligence is often excellent.
In his 1984 paper, Richard Roll studied the relationship between the orange futures market and the weather. Orange trees cannot withstand freezing temperatures for over a few hours, and the paper found that the prices of orange futures at the close of market at 2.45 pm, predicted errors in the weather forecasts of the minimum temperature later that evening. This serves as an example of how the general public, in this case, the aggregation of orange buyers and suppliers, may end up revealing information about the weather that even the weather experts tend to miss.
Please note that these research studies are based on centralized prediction markets that are presently limited to analysts, experts and rich investors. As we proceed, we will see the limitations of the centralized prediction markets and make a case for decentralized markets.
Limitations of Centralized Prediction Markets
The contemporary prediction markets are very centralized that limits the utility of markets and their ability to extract information from people to make accurate forecasts. A centralized prediction market creates a prediction task, determines the participants, rules of engagement and is highly regulated. Let’s address some of the issues with this framework:
Centralization: A basic illustration of a prediction market is presented below. There are three main stakeholders, specifically a centralized platform, publisher and users. A publisher publishes a prediction task on a centralized platform. Subsequently, users trade their money for their perceived predictions and eventually, the platform compares these predictions with the final outcome and decides whether the participant’s prediction is right or wrong.
It is quite evident that a single or centralized entity has a significant influence on overall activity in the prediction market. Centralization is inherently bad because it cannot avoid the building of trust among the stakeholders such as trust between platform and publisher, trust between users and publisher, trust between publisher and users, etc. Due to this “trust” problem, the centralized platform may have some vested interest to adjust many parameters to maximize their profits, after all, it’s a for-profit platform. In the worst case, the centralized platform may manipulate the final outcome as well which would make you lose your money despite predicting the final outcome. Ultimately, participants have to “trust” the authenticity of the platform in order to ensure privacy, security, fairness and safety of their funds. In a decentralized prediction market, participants need not trust any central entity for the transfer and safety of their funds as smart contracts will automatically execute the action.
They are Expensive: Investors in centralized prediction markets bear various costs for participation as the operators charge high fees. The justification for high fees originates from the fact that these centralized prediction markets are for-profit exchanges recovering their regulatory and administrative costs from high fees. Participants are also required to pay very high trading fees of 4-10%, deposit and withdrawal fees while the market takes a cut from their profits. This is an obstacle for reliable participants as the entry barriers and high fees would discourage them from participating in prediction markets. This can lead to less reliable predictions as the useful information required to make accurate predictions would not be encouraged. In most cases, there are withdrawal limits as your funds cannot be withdrawn instantly.
Closed: The strength of any prediction market is directly proportional to the number of participants contributing to “crowdsourcing” of the information needed for an accurate forecast. However, traditional centralized prediction markets, like most financial markets are limited by stringent regulations, capital controls and regulations. The privileges of prediction markets and their significance are only enjoyed by the developed countries that tend to have advanced financial markets, sufficient liquidity and institutional framework. It is near to impossible for a person from a developing country, say Bangladesh or India to participate in prediction markets of developed countries.
Regulations: Given the nature of prediction markets, federal regulators usually have an unfriendly approach towards prediction markets. Market supervisors and regulators act as gatekeepers that restrict who can participate and which events participants can speculate on. Additionally, you cannot simply create your own prediction market for a specific event unless you run a regulatory and licensed platform. Nobel laureates who co-authored the paper “The Promise of Prediction Markets” observes:
Unfortunately, however, current federal and state laws limiting gambling create significant barriers to the establishment of vibrant, liquid prediction markets in the United States. We believe that regulators should lower these barriers by creating a legal safe harbor for specified types of small-stakes markets, stimulating innovation in both their design and their use……..(regulatory) approach could suppress innovation and reduce opportunities to aggregate information and improve decisions.
Therefore, it is very important to transform the existing traditional platforms into decentralized ones. It will help to displace the control of prediction markets from centralized entities and give it back to a lot of people in the community. As I have explained above, the major disadvantages of traditional prediction markets arise from their inherent feature of centralization but blockchain technology can solve many of these problems and ensure the safety of funds and fairness of the whole process.
The Need for Decentralization
In a nutshell, decentralization refers to the transfer of control and decision-making from a centralized entity (individual, organization, or group thereof) to a distributed network. It strives to reduce the level of trust that participants must place in one another, and discourage their ability to exercise authority or control over one another. The benefits of decentralization range from providing a trustless environment to optimizing the distribution of resources. Integrating blockchain technology in the prediction markets not only eliminates many issues but also provides other benefits that encourage participation and create an environment for democratizing the prediction markets to the masses. Let’s explore some advantages of DPMs.
They are open: Similar to cryptocurrencies, decentralized prediction markets (DPM) are open, peer-to-peer, public, borderless networks and have no transnational boundaries. Participants from any part of the world can buy ether or other cryptocurrencies powering prediction markets to bet on event outcomes and also create their own prediction markets without any regulatory and licensing requirements. As previously discussed, traditional markets are closed and geographically restricted. For example, it is very hard for an Asian or African investor to get access to American prediction markets but with a DPM, anyone from anywhere can join and participate as they wish. In the coming decades, the openness feature of DPM is going to play a major role in the democratization of prediction markets.
They are inexpensive: In a blockchain-based prediction market, the fees imposed for participation are considerably low because there are no middlemen. The only charges are the network fees that are needed to maintain the overall network security. In general, these fees are negligible.
Eliminates Counterparty Risks: DPMs do not have any intermediaries and thus eliminates the counterparty risks involved with middlemen. They are also more resistant to censorship and corruption as they cannot be arbitrarily shut down by regulators.
Fund Safety: A participant’s fund is secured in a completely decentralized process. Every fund is protected with a multi-signature transaction. If a participant wants to move funds from one wallet to another, then he must have all signatures of the needed masternodes for this transaction. Before signing a transaction, each masternode will check whether the conditions required for a transaction are met. If the decision is “no”, then masternodes will not sign it. If a malicious node wants to steal funds from a participant’s wallet, then he needs to own many masternodes on the network which would be too expensive for him to operate - thus a decentralized network of nodes ensure the safety of funds.
Popular Decentralized Prediction Market Platforms
Multiple decentralized prediction markets have developed over the past few years. Augur and Omen were early developers in this space, with Augur launching as far back as 2014. Many of the other projects disappeared eventually due to the following factors:
Low Liquidity
High Gas Fees on Ethereum
Technical Glitches
However, since the last year, Decentralized Prediction Markets have started attracting the crypto community. With ethereum scaling solutions such as Matic became operational, upcoming EIP-1559 upgrade of Ethereum that could possibly reduce the network fee and so on, there is now a certain possibility that this sector will continue to grow and develop. The last two years have been full of unpredictable events such as the 2020 presidential elections, COVID-19 vaccines, the longevity of government-imposed lockdowns, uncertainty in the financial markets and other sets of binary outcome events that have actively prompted interest in DPMs. Today, we have dozens of decentralized prediction markets that are well developed, more efficient, cheap, and secure platforms for trading. The most popular ones are:
1. Augur
Augur is one of the first decentralized, trustless, open blockchain-based prediction markets. It is built on Ethereum blockchain and uses smart contracts and oracles to allow for the prediction of real-world events without the involvement of any central or third party.
As Augur’s whitepaper highlights, Augur follows a four-stage progression:
Market Creation: The platform allows any participant irrespective of race, caste, gender or sexuality to create a market for any upcoming event with uncertain binary outcomes. The market creator sets the event end time and chooses a designated reporter to report the outcome of the event. The assigned reporter does not unilaterally decide the outcome of the markets and participants have always an opportunity to dispute their claims.
Trading: After creating a prediction market, participants are free to predict future outcomes by trading shares of those future outcomes.
Reporting: Once the event is complete, decentralized oracles (we will shortly discuss what these are) help to determine the true outcome of the event in order for the market to finalize and begin settlement. The process is facilitated by some random profit-motivated reporters who simply report the actual, real-world outcome of the event. The reporters whose final judgement is consistent with the real-world results are financially rewarded, while those whose reports are not consistent with objective reality are financially penalized.
Source: Augur’s Whitepaper Market Settlement: A participant can close their position in one of two ways: by selling the shares they hold to another trader in exchange for currency, or by settling their shares with the market after the event. Smart contracts automatically execute the transaction and the participant receives the money if his prediction was correct.
Apart from Augur, some popular blockchain-based prediction markets are:
The Importance of Blockchain Oracles
As we have understood by now, a decentralized prediction market relies upon a set of smart contracts that settle who will get paid how much if an event occurs. You might also wonder if there is no central entity then who determines which outcome actually occurred in the real world? In a typical centralized market, this is ensured by a single operator. The centralized operator has the final say, but in the decentralized market, there is no centralized supervisor. So how do DPMs resolve this final judgement problem? This is where the role of oracles kicks in.
Oracles are third party or data feed services that provide necessary data that smart contracts may need to execute under specific conditions. They work as a bridge between real-world events and blockchain prediction markets by sending data from the outside world, such as temperature data or the number of votes a political candidate has received to a blockchain such as Ethereum. A smart contract built on top of that blockchain can then use the data and determines whether to dispense money and to whom. Centralized applications rely upon centralized data feed entities for their data requirements. For example, Foreca provides weather data for Microsoft Windows and MSN.
In the worst scenario, trusting a centralized data source might lead to serious issues. For instance, the owner of an oracle data feed could manipulate the data in order to sway smart contracts in some party’s favour. Malicious participants could hack the data feed to sway the final outcome in their favour. Subsequently, if the smart contract relies upon such manipulated data then it might lead to a market failure. As a result, researchers and developers are exploring ways to develop resilient decentralized oracles. Today we have decentralized oracle services such as Chainlink that provide tamper-proof data, verifiable randomness, external APIs, and much more to smart contracts. Therefore, in a DPM, the duty of decentralized nodes of the oracle is to just validate the state of the world and upload it periodically to the blockchain.
More Applications of DPMs
Global Financial Markets
In the next 10-15 years, as the adoption of Decentralized Prediction Markets will increase in the developing world, it would democratize the participation of people with access to mobile internet in prediction markets. One possible advantage of the mass adoption is that it would allow users from developing countries to have some exposure to stocks, commodity, forex or any other financial asset.
For example, a daughter of a hardworking fisherman from a developing country with a mobile phone, internet access and little savings but with limited access to banking or have trivial knowledge of how to invest in the U.S. stock market or other financial assets could possibly add Facebook, Apple or Tesla in her portfolio and monetarily benefit by predicting the direction of the stock’s future price movement.
A talented teenager from the Balangir district of Orissa state could gain exposure to his favourite stocks in the United States with the help of mobile application or web, without having to register for the brokerage or agent service. In a traditional financial market, he can only buy a full share of a company but with the decentralized prediction markets, he could purchase a small fraction of a share, a feature that is not available on popular traditional trading apps.
Governance
As we have discussed earlier, a paper titled The Promise of Prediction Markets, written by a group of economists that includes four Nobel Laureates have argued that prediction markets can enhance social welfare by helping businesses and governments make more accurate forecasts and better corporate governance and national policy decisions.
Futarchy is a governance model proposed by economist Robin Hanson in which elected representatives would formally define and manage an after-the-fact measurement of national welfare, while market speculators would say which policies they expect to raise national welfare. Under futarchy, prediction markets will be highly used to determine which public policies will have the most positive outcomes. While Robin Hanson’s model of futarchy is based upon centralized prediction markets, the crypto community have started their research and serious experiments with futarchy systems built on decentralized prediction markets. The ongoing advancements are promising and it would be really interesting how the developers and crypto community plans to convince the general public and government about its potential importance to ensure transparency and accountability in governance.
You would argue that futarchy would limit the participation of those with little to no income. But if governments, institutions and organizations collectively acknowledge the importance of futarchy for transparent governance then they could simply subsidize participation costs and pay them the required amount to participate in prediction markets in order to ensure the central idea of “Wisdom of the Crowds” is sustained.
I plan to cover futarchy, decentralized governance and Decentralized Autonomous Organisation (DAO) in a separate article next month but if you are interested, then you can read Ethereum co-founder Vitalik Buterin’s article and the research paper titled “Blockchain Technology and Decentralized Governance: Is the State Still necessary.”
Conclusion
I am happy that you made it to the end of this long essay. I understand it must be perplexing to comprehend all the underlying new technicalities, concepts, graphs, statistics and the consistent repetition of the word “decentralized” might have made it more irritating. But we are in the very early days of exploring the scope of decentralized prediction markets into our daily lives and their potential utility of crowdsourcing information for accurate forecasting and other applications.
The prediction market is itself a powerful idea. Open, transparent, peer-to-peer and decentralized prediction markets manifest into even more powerful one. The market is very nascent and further research is required to explore the extent to which these markets can offer more benefits to broader segments of the society. DPM is only one aspect of Decentralized Finance (DeFi) as it has a lot more to contribute, such as lending, borrowing, derivatives, decentralized exchanges, etc.
We are witnessing the birth of revolutionary technology and its range of applications can play a crucial role in transforming our institutions, societies, workplaces, public administration, governance and approach towards financial inclusiveness. As global adoption continues to grow, I am now quite confident that cryptocurrencies, smart contracts, DeFi and the underlying blockchain technology are about to change everything the way the internet had changed everything in the late '90s.
If you find my research and analysis quite helpful in your understanding of economics and financial technology, then please consider sharing my article with your friends. Since I do not charge for my articles at all, sharing my blog is the only way you can help me! And if you are new here, please subscribe to my newsletter for awesome articles like this!
References and Further Reading:
Why & How Decentralized Prediction Markets Will Change Just About Everything. | by ConsenSys
Decentralised Oracles: a comprehensive overview | by Julien Thevenard
An Introduction to Futarchy | Vitalik Buterin
A Decentralized Prediction Market Platform Based on Blockchain and Masternode Technologies
What is Decentralization? by Amazon Web Services
Corporate Prediction Markets: Evidence from Google, Ford, and Firm X
The Wisdom Of Crowds by James Surowiecki