Skip to main content
Belief Index is a family of rules-based indices published by Belief Systems. Each index – called a series – tracks a themed basket of prediction market contracts and condenses them into a single published number: a benchmark of how markets are pricing a coherent category of real-world event risk, such as the trajectory of U.S. monetary policy or the outcome of a national election. If you have used the S&P 500 to read the equity market, or the VIX to read volatility, the concept is familiar: rather than inspecting hundreds of individual instruments, you read one rules-based number whose construction is published, systematic, and independently verifiable.
This documentation covers the full production methodology: market selection, weighting, NAV computation, index levels, reconstitution, and known limitations. Belief Systems publishes index data and analytics to the general public; separately, a small number of invited participants hold interests linked to certain series through a private, invitation-only Alpha Program (documented in its own section). Nothing in this documentation is an offer to sell or a solicitation of an offer to buy any security. See Disclosures.

​
The Measurement Problem

Prediction markets are emerging as one of the most efficient mechanisms for aggregating beliefs about future events – elections, economic data releases, policy decisions, and more. But as a data source, they present the same problem equity markets did before the first stock indices: hundreds of individual instruments and no summary statistic.
A single venue hosts hundreds of individual markets, each with its own order book, liquidity profile, and resolution terms. The aggregate signal – “how do markets price rate-cut risk?” – is scattered across dozens of related contracts with no principled way to read them together.
There is no single number that captures a thematic question – such as “how is the market pricing the 2026 easing cycle?” or “what are aggregate Republican electoral prospects for 2028?” – across multiple related events at once. Every market must be found, read, and mentally weighted by hand.
Individual prediction markets settle at exactly $1 or $0. Read in isolation, a single market is a noisy, discontinuous series that can jump to zero overnight on one surprise. A diversified basket turns a set of binary readings into a continuous, interpretable signal.
Each market has its own resolution date, settlement rules, and oracle. Tracking a theme over time means constantly replacing resolved contracts with live ones – and doing so without introducing artificial jumps into the series. This is precisely the problem index reconstitution and chain-linking solve.

​
What a Belief Index Does

Belief Index applies the core disciplines of index construction – published selection criteria, systematic weighting, transparent valuation, and rules-based maintenance – to prediction markets.
1

Curated, themed baskets

Each series bundles 5-20 related prediction markets around a coherent theme, selected according to published criteria including liquidity, resolution clarity, and thematic relevance.
2

Rules-based weighting

Constituent weights follow a published formula with concentration caps. Where judgment enters – theme definition, judgment screens – it is exercised once, at composition time, under published criteria; it is never applied continuously or invisibly.
3

Transparent valuation

The basket is valued at periodic NAV windows – currently every 30 minutes, subject to change – from observable order book midprices, using a published methodology. Every formula is disclosed; every computation is independently verifiable. No black boxes.
4

Published index levels

Each series publishes an index level, rebased to 100 at inception, that moves proportionally with the underlying basket. Resolved markets fold their settlement values into the level; reconstitutions are chain-linked so composition changes never create artificial jumps.

​
Series Examples

Each series tracks a coherent theme, so the index level answers a category-level question rather than an event-level one: An election-focused series answers “what are aggregate Republican electoral prospects for 2028, right now?” – one number, computed by published rules, comparable across time, for a question no listed instrument prices at all. Themes like elections, policy decisions, and geopolitical events have no futures curve or swap market; a Belief Index series is often the only continuous, citable measurement of them. See Where the Signal Is Irreplaceable.

​
Reading an Index vs. Reading Individual Markets

​
The Order of Operations

In equities, the index came decades before the index fund: Dow Jones began publishing averages in 1896; the first index fund arrived in the 1970s. The benchmark is what makes everything downstream possible – research, journalism, risk measurement, and eventually investable products built by issuers who license the index. The sequence matters for a second reason: a benchmark’s authority accrues with its track record, and a track record cannot be created retroactively. Prediction market order books are not archived the way equity prices are – every valuation window published today is a point-in-time data point that no one, including us, could reconstruct later. See A History That Cannot Be Rebuilt. Belief Systems is built in that order. The indices are published, informational, and free to cite. Issuers who want to reference a series as the underlying for a product can do so under license through Index Services, and a small invitation-only Alpha Program operates privately under the conditions described in Disclosures.
Belief Index levels are informational benchmarks computed from observable market data – theoretical values, not execution prices. See Risk Factors for methodology limitations.

​
Key Principles

  • Transparent – The valuation methodology is fully published. Anyone can replicate any published level using public market data and a spreadsheet.
  • Systematic – Index composition and weighting follow published, rules-based methodology. Where human judgment enters – theme definition and the documented judgment screens – it is exercised once, in public, under published criteria, not continuously and invisibly. That is what makes the resulting number citable.
  • Theme-driven – Each series tracks a coherent risk domain, so the index level reads as an answer to a specific macro or event-driven question.
  • Point-in-time – Levels are recorded as of each valuation window and are never silently restated; methodology revisions are dated and prospective-only. The archive is built live, window by window – see Benchmark Governance.
  • Investable-grade discipline – Valuations are computed with forward-pricing windows, double-entry fund accounting, and continuous reconciliation against on-chain holdings: the operational standard an investable product would require, applied to a published index from day one.

​
Getting Started

How It Works

How an index is produced, from market data to published level.

Why Belief Index?

The case for rules-based benchmarks of event risk.

Index Series

How themed baskets are constructed, weighted, and maintained.

NAV Methodology

The transparent, verifiable methodology behind every published level.