First Published on @IOSG Medium on Jan 20, 2020

In an article published in 1945, Friedrich Hayek pointed out a centrally planned economy would never beat the efficiency of a decentralized economy. Knowledge is unevenly dispersed among different members of society and a single agent only holds a small fraction of the total knowledge. As a result, decisions are best made by those with local knowledge rather than by a central authority.

What if individuals have their own opinions? What if people may get rewarded when their opinions are correct? Will the market gather more wisdom? Will it help us make a better decision?

Blockchain technology enables it and we see that even before the hype of DeFi, projects in the field of prediction markets such as Veil and Augur have attracted huge attention.

In this article, we will analyze pain points of the current prediction platforms, discuss the attempts that ambitious teams have been working on, and explore the future investment opportunities.

TL;DR

1. Crypto prediction market has managed to attract large numbers of betting products even though currently, it is only a small market with 0.08% total value locked in the DeFi field.
2. While the Scalability Trilemma exists in the broad blockchain industry, there is also a dilemma between liquidity and compliance in the prediction market. To achieve high liquidity, one most common way is to introduce more entertained scenarios. This leads to application scenarios like betting or gambling, which are unfortunately strictly regulated in different jurisdictions. Liquidity remains still as an obstacle.
3. Our thoughts: · The prediction market is still at its early stage and has the potential to capture the mass market for greater usage. We believe financially-attractive decentralized protocol with a friendly user interface will be designed shortly. · Automated tool providers are in urging need. As most players are focusing on improving protocols, UI/UX and value-added tools have been heavily ignored. · Products dealing with insurance or hedging might be a killer use case in the prediction market. The market that insurance giants are hesitating to cover due to compliance and jurisdiction, is a ~250 billion US dollar market (total market cap of cryptos as at 19 Jan 2020).

Major pain points

All prediction platforms share a few problems in common: Small user base in the crypto world→low volume (even the biggest player Augur)→low liquidity→bad user experience→even smaller user base. Liquidity is hard to improve due to long resolution time, high learning cost, unattractive profit, and the trade-off between compliance and liquidity. These pain points impede the large-scale usage of the prediction market. Currently, it has less than 0.1% total value locked in the DeFi field (source: Defipulse).

First, compared to centralized platforms, which have almost instant resolution time (including derivatives exchanges and online betting), decentralized ones rely on crowd decision systems to find an honest result or a result that at least majority agree, and hence it is time-consuming. However, this pain point seems to have been gradually solved. For example, Augur’s V2.0 will cut resolution time from at least 7 days to around 24 hours.

Second, the learning cost is super high. Non-crypto users, who understand the technical theory behind the prediction market and how to bet on those platforms, prefer fiat currency. Let’s picture how complex and expensive for them: users need to learn how to convert fiat money into multiple cryptocurrencies to participate, which cost money (transaction fee), complex (download wallet, log in exchange, buy crypto, etc.) and not fun (due to some terribly-designed applications). This would definitely prevent new users from entering crypto applications. For crypto users, the learning cost might still be high as users need to learn how to create a valid market, how to use the tokens and all different mechanisms behind the oracles. The market creation is highly non-standardized and impedes its large scale. Fortunately, we do see some promising projects notice it and are trying to improve this feature such as Veil (failed sadly) and Guesser. They build on Augur and positioned themselves to provide better user interface and services.

Third, the financial incentive is not attractive considering the cost. We can see a low return model through the Augur’s example. Market creators need to stake ETH (or Dai in V2.0) to create a market and stake REP to designate a reporter, which will be returned with no bonus/interest if the market creator did everything correctly. The only return for them are fees on settlement volume, which is the creator fee in ETH, typically in the range of 0–5% determined by the market creator. On the other hand, a market creator has a lot of costs to bear: the opportunity cost of staking ETH and REP, a mandatory initial liquidity cost, and the learning cost we mentioned above. Overall, we think oracles with better-designed incentive mechanisms could help the prediction market attract more attention.

Finally, there is a dilemma between compliance and liquidity. Veil (owned by a Cayman Island company) which built on Augur and 0x provided instant settlement for traders before resolution by acting as the counterparty. This feature improved liquidity and user experience while drew the attention to authorities and eventually make itself sunset. Just as the Scalability Trilemma existing in the blockchain industry, in the prediction market, there is a dilemma between liquidity and compliance. The less decentralized the platform is, the higher risk it may face. Centralized platforms that allow real money play is strictly banned. In the crypto world, the regulation risk is getting high with centralized attempts such as off-chain resolution or act as counterparty to provide liquidity. On the other hand, projects try to provide more entertained scenarios to attract users for the ultimate goal of gaining liquidity. However, the current traction pool gathers most users who have been betting and gambling, which is strictly regulated in different jurisdictions. Even projects like Gnosis received a license from Gibraltar, is still prevented from touching a few related scenarios.

Project briefing

Let’s dive in:

1. Prediction Market Oracle/Protocol: Augur, Gnosis, Stox, Flux, and Bodhi. A lot is happening in this sector.

· Augur is the first decentralized prediction market built on Ethereum that allows users to create and trade outcomes of events in any category. It conducts a fully decentralized oracle from market creation to resolution and as a result, it inevitably leads to long resolution time. It is no surprise that its V2.0 will have to cut resolution time from at least 7 days to around 24 hours to improve user experience. The Augur team will integrate 0x to allow partially off-chain trading so that people can create, modify, and cancel orders faster and cost less. Maker will also be integrated to allow Dai as collateral token and trading token in an event.

· Gnosis will launch its prediction platform in 2020 soon. The project has big ambitions to establish a prediction ecosystem for greater use. It has multiple products pipeline such as a trading DEX-DutchX, an ERC-20 wallet-Gnosis SAFE and its multisig wallet. The major difference of its oracle compared with Augur is its centralized resolution- a default centralized Oracle for judging prediction results.

· Stox is a product of a CFD trading brokerage. The platform uses its own token STX as collateral token and wagering token. It uses a centralized oracle for resolution with a decentralized dispute mechanism. Unlike Augur and Gnosis, it mainly attracts centralized market creators such as invest.com to operate markets. However, having a consortium of centralized providers means that Stox is not technically censorship-resistant as those providers are easily shut down by authorities.

· Flux is an early-stage project and it will soon launch its Alpha version. It is built on NEAR protocol for scalable transactions and has data driven tiered oracle for an instant resolution which can be between 30 mins to 24 hours as the team claimed. Furthermore, their solution for solving liquidity problems besides having chosen a scalable blockchain is to have a standardized market making API for use by traditional and blockchain based centralized market makers.

· Bodhi aims to provide a prediction market platform for Chinese users. In contrast to Augur, Bodhi uses third-party Oracle to automatically judge the prediction results. Different from Gnosis’ centralized Oracle solution, Bodhi enables BOT holders to take over the voting process and make the final decision of a prediction event in the event of Oracle failure. Similar to Stox, it also employs a single token model to pay fees, post collateral, and serves as the currency used for wagering.

2. User Interface/Service: Veil (failed) and Guesser

· Veil built on top of Augur and using 0x to speed up trading and settlement. Launched in January 2019, Veil announced the suspension of the company in July 2019. The Co-founder, Paul Fletcher-Hill attributes a difficult onboarding experience to its “failure”. The code, however, continued to be open-sourced and they have done a great job for improved user experience.

· Same as Veil, Guesser is an application built on top of Augur and they optimized user experience. They aim to build an open prediction platform for non-technical users. Apart from UI/UX improvements, they are providing bet indicators on events, advanced filters by subtopic, a probability chart and prediction list and a bunch of package tools for users.

3. Prediction Market for Insurance

· Although the most obvious use cases for prediction markets are gambling and betting, we believe both insurance and hedging have huge potential. For example, Companies heavily exposed to a particular type of natural disaster (i.e. earthquakes, hurricanes, or wildfires) could buy shares in the outcome “yes”. People can use prediction markets to hedge their positions to limit their downside.

· Decentralized insurance solution could cut 18% cost (i.e., human cost and operational expenses) in the traditional insurance market according to Nexus Mutual (Nexus Mutual is a decentralized insurance platform). In addition to the cost saved, decentralized insurance could enhance flexibility with customized insurance products available. Big traditional insurers are hesitating to provide crypto insurance due to compliance concerns and high volatility of crypto value. Therefore, we haven’t seen large amounts of complex products especially for hedging purposes yet. This gap is captured by projects such as Upshot and Opyn. Opyn introduces an insurance and hedging instrument that protects against both technical and financial risks that threaten users of DeFi.

· However, just as the prediction market, it is suffered from an unattracted profit and low liquidity. Take Upshot as an example, to obtain the premium, the counterparty must be fully-collateralized, which means collateralizing the equivalent amount of what policyholders’ insurance contract covering. The full-collateralization for a relatively low insurance premium means a low return, which is unattractive for counterparties and could lead to low liquidity.

Insights for the future

A prediction market is deployed whenever we want to know the market and to discover what people truly believe. Through the token economy, the best prediction gets rewarded and further incentivized to make better decisions.

Projects that provide valid solutions solving current market pain points deserve more attention. We believe there are three potential investment opportunities.

1. The prediction market is still at its early stage; yet has the potential to capture the mass market for greater usage. We believe finanicial attractive decentralized protocol with a friendly user interface would be designed in the near future. Currently, the prediction market has only 0.08% total value locked dominance in the whole DeFi field, while Maker accounts for ~50%. Augur as the first player is upgrading itself by cutting resolution time and improving user experience. Prediction market as one of the earliest and promising DeFi applications, there is still huge potential to be discovered. Also, the protocol layer might be less risky in terms of legal regulations as it does not interact with end users much. That’s why we are looking more at protocols in this field. But even with protocols, projects need to consider its business model with public chains and DApps, and most importantly, its monetary incentive for users. How to charge when others deploy your protocol? How to attract users to increase trading volume? How to incentivize users to stay?

2. The prediction market to attract crypto and traditional users will need robust UI/UX and useful automated tool providers. As most players are focusing on improving protocols itself, UI/UX and value-added tools are heavily ignored. Obviously, the market still needs more education, but current platforms do not make it easier for junior users to get started and the learning cost is high as well. The traditional potential customer base is huge. For the political prediction market, about US$277 million was traded on the 2016 U.S. presidential elections on Betfair. For illicit betting, a Wired article published in 2018 claims that the size of wagers in illicit sports betting alone ranges from US$80b to $150b annually. For insurance, an article of Coindesk in 2018 estimated the total amount of coverage is willing to provide to crypto custodians and exchanges at around US$6 billion. We think projects that focus on these details can be a large-scale user entrance. Efficiency, resolution time and liquidity still remain as major obstacles in the prediction market. We are expected to see more innovative methodologies to conquer the difficulties mentioned.

3. Regarding a more specific field, we see products dealing with insurance & hedging could be a killer DApp in the prediction market. Needless to say the huge market potential, decentralized insurance gonna bring flexibility and customization to contribute to the open economy. We believe there will be more complex financial instruments that appear as DeFi matures and prediction platforms can be a suitable environment where especially insurance & hedging products receive attraction.

References

1. Prediction markets by Circle Research
2. Content Review: Augur, Veil and Decentralised Prediction Markets by Anthony Faccenda
3. https://www.theblockcrypto.com/linked/47176/bitfinex-to-make-foray-into-prediction-markets
4. Crunchbase
5. The Ultimate Guide to Prediction Markets by Cultivate Labs
6. http://blog.oddhead.com/2006/10/30/implementing-hansons-market-maker/
7. All websites of projects mentioned above
8. Augur Protocol Overlays: Veil and Guesser. Prediction Markets Made Easy by MARTIN BANOV
9. https://zhuanlan.zhihu.com/p/37205364
10. Nexus Mutual white paper
11. Friedrich Hayek (September 1945). “The Use of Knowledge in Society” (PDF).
12. https://www.coindesk.com/crypto-prediction-market-and-derivatives-platform-veil-is-shutting-down
13. https://www.paddypowerbetfair.com/\~/media/Files/P/Paddy-Power-Betfair/documents/annual-report-2016.pdf