Github Repository

All the source code is in the Master branch except the Subgraph. Subgraph Code is in the Subgraph branch.


The prediction markets are not popular financial products in DeFi due to their risky nature. Some of the existing solutions are based on the "winner takes all" model. The inspiration for the "No-loss" prediction markets was mainly due to the Aave lending protocol. So there is a low barrier for users who want to try out these financial products with essentially no risk.

What it does

Currently, There are two types of markets available:-

  1. Normal Market (High Risk) It's a market where correct predictors will get rewards proportional to their stake plus the interest earned from the Aave lending protocol. And, incorrect predictors will lose their stake.

  2. No Loss Market (Less Risk) Here the interest earned from the Aave lending pool will only be distributed among the correct/winning predictors. So, It's is a less risky market as even an incorrect/losing predictor will get their stake back.

We also have two types of market outcomes:-

  1. Conditional Outcomes: Here one can predict only "YES or NO", based on condition.

For eg. Will the price of ETH go above $500?

Here, there will be only two options YES (For predicting that it will go above $500) and NO (For predicting that it will not go above $500).

  1. Strict Outcomes: Here one has to predict the exact value. For example:

Q. Who will win the 2020 U.S. Presidential election?

1. Joe Biden
2. Donald Trump

At the time of the market resolution (sometime in the future), the data is fetched from the chainlink oracles and the market is resolved with the correct results. These results are fetched from a trusted source.

How we built it

We have implemented prediction markets in solidity and deployed the contracts on the Kovan testnet. We make use of chainlink and Aave in our contracts. We are using Chainlink oracles to fetch external data to resolve the market. We believe that Oracle is the only secure way to fetch data into a smart contract.

We are using Aave for maximizing the profits of the users. The user can stake for a limited time span after that the funds are deposited in the Aave lending pool where it earns interest until the market is resolved. At the time of market resolution, the funds are withdrawn from the market to the smart contract. The eligible users can then withdraw their respective rewards anytime they want.

Once the market reaches the market close timestamp, the system automatically displays the resolve market button. On the click of the button, all the funds from Aave get withdrawn and get deposited to the market contract. At the same time, the smart contract calls the chainlink function to get the market result. In few seconds chainlink provides the result and the market gets resolved.

We are using The Graph protocol, to visualize market data like predictions, stake, and market resolution results. This data can be analyzed by the user for risk assessment or due diligence. The subgraph can also be used as an oracle for other markets to supply useful data but such use cases are not explored in this project.

The subgraph is deployed on kovan:


  1. You have to connect Metamask and choose the network as Kovan. The system will show all the available markets. It will display both No Loss and Normal prediction markets. One can also see if the market is already closed.
  2. You can click on any market and the system will direct on the market page. On that page, one can see all the details of the market.
  3. You can create your position by clicking Want to Predict button. You have to choose the outcome and need to fill the stake amount, We only support DAI for now.
  4. The system will show you metamask pop-up to approve the token. It will call ERC20 to approve the function.
  5. After approving, you will get another metamask popup to sign the submit prediction transaction.
  6. You can see the result once the market gets resolved at the market close timestamp. If you have any position in the market and made a correct prediction, You can click to claim reward button. For the No Loss market, also the incorrect predictor will get the button Get Stake to get their stake back.

Challenges we ran into

  1. Calling the API from chainlink.
  2. Making the conditional market. Creating a generic market for assets like ETH, BTC, etc.

Accomplishments that we're proud of

  1. We are able to create No Loss as well as Normal prediction markets.
  2. We successfully created a condition-based market, which is fully decentralized.
  3. Able to integrate Aave to maximize user profits.

What we learned

  1. We learned to leverage chainlink oracles to fetch data securely.
  2. We learned about Aave and Chainlink.
  3. We got to know the power of The Graph protocol.

What's next for Chainlink Prediction Market

  1. We want to make our system developer friendly so that anyone can deploy the markets easily. For example, anyone can easily create a market for: "Will Trump win general elections of 2020?"
  2. We want to incentivize the market creators with some fees.
  3. We want to use The Graph for providing results to Chainlink using External Adapters in the future to get the result of any type of questions.
  4. Providing interest to the incorrect predictors also in the case of Normal Markets (High Risk).
  5. We want to use Gelato For automation of Depositing stake on Aave and Resolving Market.
Share this project: