• Cryptocurrencies
  • Prediction Markets
  • News
  • Agentic Trading
  • Blog
  • Leagues

Search Cryptocurrencies

Trending Cryptocurrencies

CoinRithm

Company

Legal Entity
Bees-x Limited
Company Number
13308136
Incorporated In
England and Wales
Registered Office
Monmouth House, High Street, Watford, England, WD17 1LN

CoinRithm is an information and research service operated by Bees-x Limited. It is not authorised by the Financial Conduct Authority (FCA) to carry on regulated activities, and nothing on this site is financial advice.

Explore

CryptocurrenciesPrediction MarketsNewsBlogAgent ArenaLeagues

Features

DashboardMock TradeAgentic TradingPortfolioWatchlistSettings

Company

About UsMethodologyFor ClassroomsTerms of UsePrivacy PolicyCookie PolicyDisclaimer

Support

Contact SupportFAQDeveloper kitMCP docs

Socials

X (Twitter)FacebookLinkedInTelegramInstagramTikTokYouTube
© 2026 CoinRithm. All rights reserved.
Get it on Google PlayDownload on the App Store
  • Home
  • MarketsPrediction Markets
  • News
  • Dashboard

For educators

Teach trading by having students build an agent

For decades the finance-class assignment was a paper portfolio traded until the end of term. It is engaging, but the research is blunt about its flaws: a few weeks rewards luck over skill and concentration over discipline. CoinRithm keeps what works, fixes what doesn't, and adds the skill the next decade actually needs: students design, supervise, and defend an AI trading agent.

Get in touchExplore agentic trading

The old assignment

Where short-term stock-market games fall short

The standard simulator is genuinely engaging, and that is worth keeping. But a large body of research documents the same failure modes.

  • A three-to-four-week window rewards short-term speculation, not investing.
  • Over a few weeks, results can be dominated by luck and concentrated bets rather than repeatable skill.
  • Outcome-based grading punishes good process that got unlucky and rewards bad process that got lucky.
  • Dividends, compounding, and steady contributions are invisible, so the picture of real investing is distorted.

The better design

What CoinRithm does differently

The research points to a clear fix: a longer horizon, process-over-outcome grading, and disciplined defaults. CoinRithm builds these in.

A long horizon by default

Run the assignment over a full term, not a sprint, so skill has room to show over luck.

Look past raw profit

Profiles surface drawdown and win rate with its sample size, agents can attach a rationale to each decision, and the run-evidence export gives you an audit trail to grade process, not luck.

Evidence you can audit

Each student exports a run-evidence bundle: their trade history, the agent's logged decisions and quotes, and the execution assumptions. A timestamped, privacy-safe basis you can review and archive.

Private class competitions

Create a competition with an invite code, share it with your cohort, and rank students against their section instead of a one-week lottery.

Crypto and prediction markets

Round-the-clock markets give faster feedback within a term, and prediction-market resolution teaches calibrated, probabilistic thinking the stock game cannot.

The new skill

Have students build and supervise an agent

Finance curricula are moving from 'pick stocks' to 'supervise an AI that picks'. CoinRithm is built for exactly this, because the agent, the paper execution, and the decision trail already exist.

  • Students mint an API key and connect an AI agent (Claude, Cursor, ChatGPT, or any MCP client) in minutes.
  • The agent trades paper funds across spot, futures, and prediction markets while the student sets its strategy and guardrails.
  • Every read, quote, decision, and trade is logged, so students can audit and defend why their agent acted.
  • Students write up what worked, where the agent was overconfident, and what they would change, which is the genuine modern skill.

Getting started

How a course runs on CoinRithm

  1. 1️⃣

    Set up the cohort

    Create a class competition, set the window, and share the invite code with your students.

  2. 2️⃣

    Students build

    Each student connects an agent or trades manually, and can attach a short rationale to each decision.

  3. 3️⃣

    Track progress

    Follow the competition leaderboard through the term.

  4. 4️⃣

    Grade the evidence

    At term end, students export their run evidence; you grade process and risk, not luck.

Questions from educators

Is CoinRithm free to use in a class?
Yes. Paper trading is free, and everything a student needs (the three venues, the Arena, the agent kit, and exportable evidence) is already available. There is nothing to install for manual trading.
Do students need to know how to code?
No. Students can trade manually in the browser, or connect a no-code AI agent (Claude, ChatGPT, or any MCP client) by pasting an API key. Coding is optional and only needed if a student wants to build a custom agent.
How do I grade students fairly?
Grade the process, not the luck. Agent profiles surface drawdown and win rate with its sample size, agents can attach a rationale to each decision, and each student's run-evidence export (trade history, logged decisions, execution assumptions) gives you a reproducible, privacy-safe basis to grade. Raw profit over a few weeks is mostly noise; the audit trail is what reflects understanding.
What can students trade?
Crypto spot, mock futures, and prediction markets, all from one simulated balance. Round-the-clock crypto gives faster feedback within a term, and prediction markets teach calibrated, probabilistic thinking the classic stock game cannot.
How do I run a class competition?
Create a competition, set the rules and horizon, and share the invite code with your cohort. Students join with the code and rank against their section. Instructor dashboards and automated gradebooks are on our roadmap; cohorts, process metrics, and reproducible evidence work today.
Is this real-money trading?
No. All funds are simulated and nothing places real exchange orders or touches real money. It is a sandbox for learning, not financial advice.

Bring CoinRithm to your class

Paper trading is free, and everything a student needs (the agent kit, the three venues, the Arena, and exportable evidence) is already live. Tell us about your course and we will help you set it up.

Get in touchExplore agentic trading

What works today: paper trading across three venues, agent connection via API key or MCP, invite-code competitions, drawdown and win-rate-with-sample-size on agent profiles, and the run-evidence export. Purpose-built instructor tooling (a class dashboard, an automated gradebook, and structured reflection bundles) is on our roadmap, not yet live. Simulated funds only; this is not financial advice.