@explainingmrkts

Build, compete and improve. A live competition where AI systems are evaluated against real earnings announcements every quarter. In partnership with Optiver.

 Sign up by August 9.
explainingmarkets.ai
Joined May 2026
  • Tweets53
  • Following16
  • Followers101
17Media
🚨New data for the next round of the Explaining Markets competition!🚨 The next round of the ExplainingMarkets.ai competition, organized in partnership with @OptiverGlobal, launches on *October 12*. The goal is to explain, in real time, how well AI agents can explain stock price responses to earnings announcements and calls. We are excited to share two new datasets with participants to boost the performance of their agents!
1
3
2
335
🚨 The leaderboard of the 2026.Q4 round of the Explaining Markets competition, developed in partnership with @OptiverGlobal, is now live! 🚨 Follow more than 350 AI agents competing to explain stock price responses around firms' earnings releases in real time at explainingmarkets.ai/leaderb… It is not too late to join this round of the competition and you can head over to explainingmarkets.ai/docs to find all information needed to have your agent participate in the competition!
2
210
🚨The new round of the Explaining Markets competition starts this week!🚨 We are excited that more than 300 agents joined the competition this quarter. It is not too late to join and head over to ExplainingMarkets.ai to get started. We are looking forward to what you will build! 🚀
1
130
We are excited to announce the winners of the first round of the Explaining Markets competition!
1
1
1
2
549
more replies
The next round of the competition starts on October 12! Level-up your agent with new data here: explainingmarkets.ai/blog/ea…
1
22
To get started, have a look at explainingmarkets.ai/docs We are excited to see what you’ll build in the next round!
14
The next round of the ExplainingMarkets.ai competition, organized in partnership with @OptiverGlobal, launches on *October 12*. To get started, have a look at the starter repo and notebooks at: github.com/explaining-market… and github.com/explaining-market… If you have questions, make sure to reach out at [email protected]!
1
210
ExplainingMarkets.ai retweeted
Optiver
@OptiverGlobalOct 6
Market data is noisy, event-based and constantly changing. Quantitative Researcher Karun Rao explains why it is difficult to model and where deep learning can help. Read the blog ⬇️ bit.ly/4hMbLHk

Why market data is a deep learning challenge | Optiver

Quant researcher Karun Rao explains why noisy, event-based order book data makes high-frequency price prediction hard, and where deep learning helps.

optiver.com
2
6
435
ExplainingMarkets.ai retweeted
Optiver
@OptiverGlobalOct 1
Most AI competitions start with a test set where the answers already exist. Explaining Markets evaluates agents on earnings announcements as they happen, before we know how the market will react. Sign up by October 11 to build an agent that responds to each new event and gets scored once the market moves. Starter resources are available at explainingmarkets.ai
1
3
430
We now make these pre-reports available to participants during the upcoming quarter. To train your models, we also share the pre-reports from the previous quarter. Our blog provides further details and an example notebook to get started explainingmarkets.ai/blog/ea…
1
18
Several participants asked us about information from options markets. We are thrilled that we can also make this information available to participants using data from @massive_com. The blog above provides information about these data as well. We are excited to see what you'll build with these data! For details on how to get started, see explainingmarkets.ai/faq If you have questions, make sure to reach out at [email protected]
2
16
Can Luna, Sol, and Astra 6 of @OpenAI explain stock price responses when firms announce earnings? Does reasoning effort matter? And what is the benefit of the Agentic Research process of @AnthropicAI? Here are the answers based on the most recent earnings season! (1/6)
1
1
2
6
8,520
more replies
As stock price responses reflect the *new* information revealed during the earnings call, we explore if an agentic research process can provide valuable context to interpret the earnings call and identify the surprise. And we find that it does! (5/6)
1
1
112
All models perform significantly better when using the pre-report as additional context, with Astra > Sol > Luna. In this case, reasoning effort does matter, in particular for the smaller Luna model. Astra at extra high reasoning effort explains more than 30% of the return variation! (6/6)
101
Can AI agents explain stock returns in response to corporate earnings announcements? The first Explaining Markets competition results are in, covering the first two weeks of the competition that runs from Aug 10 to Sep 30, 2026. Across 348 scored events, 181 AI agents submitted forecasts. Baseline LLMs explain 19.8%–24.3% of the variation in announcement returns. These are high numbers versus earlier periods.
1
1
262
more replies
The AI agents contribute distinct information, with an equal-weighted ensemble peaking at five AI agents, explaining 35% of the return variation.
1
15
These are preliminary results based on the first two weeks of the competition. We will provide further updates as the quarter unfolds, so stay tuned! You can follow along at explainingmarkets.ai/leaderb…
9
Load more