Index education

Like markets, index investing is constantly changing. Hence, we have created a short educational guide for everything from index construction and customisation to index operations and technology. We hope you find it useful. 

Index construction

What is index design? 

Index design is the process of defining the rules that determine what an index tracks and how. It sits upstream in the index process and covers things like investable universe, eligibility and selection criteria, constituent weighting, rebalancing frequency, and treatment of corporate actions. The output is a precise methodology document that a calculation agent can use to compute the index.  

Modern index design increasingly happens on software platforms like PANTA that let investment teams test rule variations against historical data, rather than building calculation logic from scratch for each new potential index. 

What is index backtesting?  

Index backtesting is the process of testing the proposed rules to see how an index would have performed in the past. This lets investment teams check historical returns, volatility, turnover, sector exposures, and how the index would have responded to specific market events – all before committing to a final methodology. 

It should be noted that backtesting is a design-stage tool, not a guarantee. Past performance under a set of rules doesn’t ensure the same behaviour going forward. For this reason, robust backtesting typically involves testing across multiple time periods and market regimes, and being transparent about the assumptions and data used.  

What is an index methodology?  

An index methodology is the published rulebook that defines how an index is constructed and maintained. It is what makes an index reproducible.  

Methodology documents also underpin an index’s transparency and governance. There should be sufficient detail that the rationale behind, for example, constituent inclusion or weighting decisions can be understood and verified independently. Changes to a methodology are typically announced in advance and take effect on a defined date, so that strategies tracking the index can prepare for the resultant turnover. 

Index customisation 

What is index blending? 

Index blending is the practice of combining two or more existing indices into a single composite index according to defined weights or rules. Rather than designing a new index from raw constituent-selection rules, a blended index is built on top of underlying indices. For example, mixing a developed market equity index with a global bond index.  

Blending is often used to create custom benchmarks or strategy indices quickly, since the underlying indices have already been calculated and validated. The design work shifts to defining the blend weights, rebalancing rules, and how changes in the underlying indices should flow through. 

What are custom indices? 

Custom indices are indices built to a specific client’s requirements rather than standard, off-the-shelf benchmarks. They are typically commissioned by an asset manager, asset owner, or bank to reflect a particular investment view, strategy, or set of constraints. For example, a sector tilt, an ESG screen, or factor combination.  

Custom indices still require the same rigour as standard indices: a documented methodology, independent calculation, and ongoing governance. What differs is that the client, often working with an index provider or using a technology platform like PANTA, drives the design decisions, and the index may not be publicly distributed beyond the parties using it. 

What is self-indexing? 

Self-indexing is when an asset manager creates and maintains its own index, using in-house intellectual property to define the methodology, rather than licensing an index from an external index provider. This approach is typically used when a manager wants closer control over strategy and cost, or wants to launch a product built around proprietary research.

Self-indexing raises the same, if not greater, governance questions as custom indices. For example, to avoid conflicts of interest, regulators typically want clear separation between the entity designing the index and the one calculating it.  

Self-service platforms like PANTA have emerged that let clients take an existing index as a starting universe and adjust selection rules or weighting parameters directly, lowering the barrier to building custom or proprietary indices without a large in-house quant team. 

Index operations 

How is an index calculated?  

An index is calculated by taking its constituents, their weights, and the prices or data inputs its methodology specifies, and applying the published rules to turn that raw market data into a single index level. A calculation agent like PANTA typically handles this on a daily or intraday basis, applying corporate action adjustments, currency conversions, and the specified weighting formula.  

For most indices, calculation has to be accurate, auditable, and repeatable. Any two parties applying the same methodology to the same data should arrive at the same number. Increasingly, regulators demand independence between the entity designing an index’s rules and the one calculating its values. 

What are corporate actions and how do they affect an index? 

Corporate actions are events initiated by a company – dividends, stock splits, mergers and acquisitions, spin-offs, rights issues, delistings, etc – that alter something about an index constituent. Because these events originate in the market rather than the index itself, the index methodology must predefine how each action is identified, adjusted for, and reflected in the index. 

Corporate actions affect an index in several ways: price-affecting events like splits require an adjustment factor so the index level stays continuous rather than jumping artificially; changes in free-float shares can shift a constituent’s weight outside the normal rebalancing cycle; mergers, delistings, and similar events can trigger constituent removal or replacement; and dividend treatment determines whether an index is calculated as price return or total return. Because corporate actions are frequent and often announced with short notice, accurate and timely processing is a core, ongoing part of index calculation rather than a one-off adjustment. 

What is index rebalancing? 

Index rebalancing is the periodic process of updating an index’s constituents and their weights so the index continues to reflect its stated methodology. This might mean adding or removing securities that no longer meet eligibility criteria, or adjusting weights back to target levels after market movements have caused drift. 

Rebalancing frequency varies by index – monthly, quarterly, or annually are common – and is set out in the methodology document alongside rules for handling corporate actions, buffers to reduce unnecessary turnover, and the effective date on which changes take effect. Well-designed rebalancing rules aim to balance two competing goals: (1) keeping the index representative of its target market or strategy, and (2) minimising turnover costs for strategies tracking it. 

Index technology  

What is an index calculation agent? 

An index calculation agent is the party/technology responsible for computing an index’s value on an ongoing basis by applying the published methodology to live market data.

To avoid conflicts of interest, the calculation agent is typically kept separate from the entity that designed the index and owns its methodology. 

What is an index operating system? 

An index operating system is a single platform that unifies the full index lifecycle. The term draws a parallel to computing: just as an operating system gives disparate applications a common interface and shared infrastructure, an index operating system gives index design, calculation, and governance a common interface and shared infrastructure, instead of requiring purpose-built tooling (or spreadsheets) for each function. 

At PANTA, this is the model behind our own platform: PANTA Lab supports index design and testing, while PANTA Engine handles institutional-grade calculation, corporate action processing, and compliance controls, with seamless connectivity between the two. The goal is to digitise the index ecosystem, giving clients one platform for index design, blending, testing, calculation, rebalancing and more.