Free Intensive Nov 21–22Free 2-day hands-on Vibecoding Intensive on Nov 21–22 at 16:00 MSK (Sat–Sun) · Live Cohort 3 starts Nov 23Join Free →Cohort 3 (Nov 23)
ReleaseAPI / Product · ActionableRB.RU→ original

Moscow Exchange Launches AIMOEX, an AI-Selected Index: A Neural Network Will Pick 20 Stocks from 100

Moscow Exchange has launched AIMOEX, the first index where a neural network chooses the stocks rather than company size. From 100 securities in the broad-market index (MOEXBMI), the AI evaluates financial reports, price dynamics and news, assigns each stock an AI Score, and takes the top 20. No single company can account for more than 8%, and the composition is revised once a month. This is a public test of whether the algorithm can beat the classic market-capitalization selection.

AI-processed from RB.RU; edited by Hamidun News
Moscow Exchange Launches AIMOEX, an AI-Selected Index: A Neural Network Will Pick 20 Stocks from 100
Source: RB.RU. Collage: Hamidun News.
Listen to article
Article Fact Sheet & Key Figures
ParameterDetails
Primary sourceRB.RU ↗
Editor-in-Chief
Key metric #1The price index is recalculated every 15 seconds during trading hours
Key metric #2The total return index accounts for dividends and is calculated once a day
Key metric #3Previously, stocks were included in indices based on free-float market capitalization

Moscow Exchange has launched AIMOEX, the first Russian equity index whose composition is determined by a neural network. Usually, the largest companies make it into an index, and no one asks which of them is more promising. This index works differently.

From the 100 securities included in the broad-market index MOEXBMI, the algorithm selects the 20 most attractive. It assigns each stock an overall AI Score, and the top twenty go into the basket. No single company can exceed 8% of the index, and the exchange reviews the composition once a month.

Moscow Exchange CEO Viktor Zhidkov announced this at the NAUFOR conference "Russian Stock Market — 2026." The indicator's full name is "Moscow Exchange Index — Artificial Intelligence Financial Services," and its trading ticker is AIMOEX.

For developers and product managers in fintech, this is an interesting case. Here the AI does not answer in a chat window. It makes decisions that affect real money and products built on top of the index.

How the neural network selects stocks

The selection is made by a multi-agent system: several AI components jointly analyze the same dataset. The input is the 100 stocks from MOEXBMI. The system evaluates each security and assigns it an AI Score, a measure of attractiveness. The twenty stocks with the highest score enter the index.

The evaluation draws on three groups of data:

  • Company financial metrics: profitability, business growth rates, debt load, dividends, and the company's market value relative to its financial results.
  • Stock behavior on the market: how quotes have moved and which direction the price is heading.
  • News: how significant an event related to the company is and when it occurred.

The last point is telling. A fresh piece of news and one from a month ago carry different weight, so the system considers not only the fact of an event but also when it appeared. This is closer to the reasoning of a working analyst than to a fixed formula.

The 8% cap and two versions of the index

To keep the algorithm from assembling a portfolio around a single favored stock, the rules set a ceiling: no company's share can exceed 8%. For comparison, previously the main criterion for inclusion in Moscow Exchange indices was free-float market capitalization, the value of shares that investors can actually buy.

The exchange calculates two measures:

  • The price index reflects only changes in the value of the shares. It is recalculated every 15 seconds during the main and additional trading sessions.
  • The total return index additionally accounts for dividends. It is calculated once a day, after the main session ends.

The difference matters for anyone who wants to compare results. AIMOEX should be compared with a classic index of the same type: price with price, total return with total return. Otherwise, dividend-paying stocks will distort the picture.

Why it matters and what to expect

According to Zhidkov, AIMOEX should show how the results of the traditional approach to index construction differ from the algorithm's selection. The exchange first reported plans to launch such an index back in December 2025, and the goal was the same: to place the classic methodology and the AI's recommendations side by side.

It is worth staying sober about this. The index has no trading history yet, so it is not possible to say whether it will beat the market. Twenty stocks out of a hundred is a small sample, and because the composition is revised monthly, the result depends heavily on the quality of the evaluations in each particular month. In addition, the algorithm's output depends on its data: if news or financial reporting is distorted, the AI Score will be inaccurate too. This is not an investment recommendation but a public experiment that will produce verifiable numbers.

The value for the industry lies elsewhere. An open benchmark appears: a fund manager, a fintech startup or a developer of trading products will be able to compare "AI selection" with ordinary market-capitalization weighting using real data. If AIMOEX shows a sustained advantage, funds and structured products could be built on it. If it does not, the market will get a straightforward answer to a question that marketers often raise today.

In parallel, AI is also making inroads into the regulatory side of the market. The Bank of Russia plans to test an assistant for analyzing contracts for financial products, and a pilot version has already been shown on the platform of the Fintech Association. For those building AI products in finance, the signal is clear: major market participants are moving from presentations to working systems, and requirements for transparency in evaluations will grow.

Story

Key Questions Answered

How the neural network selects stocks?

The selection is made by a multi-agent system: several AI components jointly analyze the same dataset. The input is the 100 stocks from MOEXBMI. The system evaluates each security and assigns it an AI Score, a measure of attractiveness. The twenty stocks with the highest score enter the index. The evaluation draws on three groups of data: Company financial metrics: profitability, business growth rates, debt load, dividends, and the company's market value relative to its financial results.

What are the key details on The 8% cap and two versions of the index?

To keep the algorithm from assembling a portfolio around a single favored stock, the rules set a ceiling: no company's share can exceed 8%. For comparison, previously the main criterion for inclusion in Moscow Exchange indices was free-float market capitalization, the value of shares that investors can actually buy. The exchange calculates two measures: The price index reflects only changes in the value of the shares. It is recalculated every 15 seconds during the main and additional trading sessions.

Why it matters and what to expect?

According to Zhidkov, AIMOEX should show how the results of the traditional approach to index construction differ from the algorithm's selection. The exchange first reported plans to launch such an index back in December 2025, and the goal was the same: to place the classic methodology and the AI's recommendations side by side. It is worth staying sober about this. The index has no trading history yet, so it is not possible to say whether it will beat the market.

ZK
Zhemal Khamidun — CPO at Alpina Digital, product manager of AlpinaGPT. Building AI products since 2019. Hamidun News is a pet project that grew into a production aggregator with 6,000+ readers.
✦ Private AI Builders Community · Vibe Club

Ship AI Faster with Founders, Engineers & Weekly Architecture Reviews

Inside Vibe Club (hamidun.com/club) — weekly live masterminds with Zhemal Khamidun, Tech Radar, full LMS access, and 370+ production Claude Code skills.

Join Vibe Club →Explore Hamidun Academy
What do you think?
Loading comments…