Erasure is a decentralized marketplace for stock-trading data, built on Ethereum and IPFS thanks to Numerai.

It’s designed to enable anybody to make market predictions, stake on them with its native NMR token and build a verifiable track record.

Erasure is home to dApps such as ErasureBay and Quant – a stock-picking tournament which uses Erasure to timestamp stock picks, then reveal them at a later date to build reputation scores.


Erasure Protocol has its roots in Numerai, a hedge fund from San Francisco which was founded in 2015 by Richard Craib – a member of the Forbes 30 Under 30 Finance list of 2017.

The company had its first funding round in April 2016, raising $1.5 million from a variety of independent investors including Naval Ravikant. This was closely followed by a larger Series A funding round the same year, which raised $6 million led by Union Square Ventures.

Even before launching the Numeraire token, the company was focused on crowdsourcing stock market prediction models using encrypted data sets, from data scientists around the world. These data scientists were originally incentivized via competition to win Bitcoin for their efforts.

About a year after its creation, Numerai released their own ERC20 cryptocurrency token, Numeraire (NMR), designed to incentivize data scientists to make higher-quality predictions.

Why Erasure Protocol?

Erasure Protocol is the first platform of its kind, which helps to solve a problem that they call “trapped information”. Trapped information is valuable market data which is never used to its full potential because its quality cannot be verified.

According to Numerai, this most often results in a dead-end for this valuable information. Since there is no fail-proof way to verify its credibility, it is glossed over by potential buyers such as hedge funds and left to waste. Meanwhile, the bearer of the information or insight will often not have the means or capital to take full advantage of it.

Erasure – and apps like Quant – provide a solution to this by building verifiable track-records for individuals with unique market insights, as well as the means to buy and sell insights in the future.

On top of Quant, Erasure also offers the ErasureBay product – an easy way to collect verifiable data by posting to a decentralized forum.

How Does It Work?

Quant provides an interface on top of Erasure Protocol for users to:

  1. Create stock market predictions
  2. Encrypt and upload them to IPFS
  3. Timestamp them by saving the IPFS file hash in Ethereum
  4. Reveal and score predictions against market data, once they come to fruition
  5. Reflect such results in user reputation scores.

The Erasure Protocol also leverages the Numeraire token (NMR) for staking on predictions, which improves incentives for high-quality data.

This is effected both via limiting the amount of stock-predicting data submitted by users (preventing spam and reducing the occurrence of “lucky” predictions), as well as providing a mechanism through which data buyers can have recourse on sellers of inaccurate data.

This mechanism involves a buyer being able to destroy tokens staked by the seller, in exchange for destroying some of their own – in a ratio called the “griefing factor”.

A large griefing factor will improve confidence in a stock market prediction feed, due to the seller putting more of their stake at risk.

How to Partcipate

Getting started with Quant and the Erasure Protocol is easy – all it takes is an email address and authentication method such as MetaMask.

To sign up and begin using Quant, visit

Users will enter their desired username and associated email address, then click “Sign Up”.

After signing a transaction to create a feed, users will then receive their very own user summary page. This summarizes their uploaded predictions and performance, as well as the amount (of NMR tokens) they have staked on their predictions.

Pictured above and below is the user summary page for “The Signal” – currently the best-performing trader on the Quant leaderboard.

A chart and table of past performance is also visible, showing a very clear indication of a user’s historical accuracy.

For users to submit their own stock predictions, this will be done using a CSV format, which is submitted using the “Upload Data” button on the left-hand side of their summary page.

For more information on how to create and format a CSV file for Quant, visit the Erasure World documentation webpage.

Stay Up to Date!

There are a number of ways to stay up-to-date and get involved in the Numerai and Erasure Protocol community. For all the latest, follow them on Twitter, or check out their official blog on Medium for long-form write-ups on news and releases.

They also have a forum on their official website, as well as a community chat where anyone can participate.