Alternative Data

Alternative data turns information created outside conventional financial reporting into an analytical input. For investment research, examples include satellite imagery, aggregated card transactions, online activity and mobile-location data; related uses in public statistics include measuring travel and spending. These are distinct applications, not evidence that every dataset can be licensed or used for securities trading. (sec.gov)

The commercial question is whether a source improves a specific decision after accounting for its coverage, lawful access, processing cost and continuity. A timely signal can be useful without being representative or exclusive. This report therefore treats alternative data as a global market for usable data and derived insights, rather than equating the market with one buyer type or claiming a market size.

From acquisition to defensibility. Source novelty alone is a weak purchasing test. OECD research on privately generated tourism data identifies coverage bias, acquisition costs, dependence on providers and changes to methodology or access; SEC examiners have separately observed inconsistent diligence of alternative-data providers used by advisers. Together, these observations point to a commercial advantage for products with documented provenance, stable historical definitions and clear usage rights. That advantage is an inference, not a measured change in vendor rankings. (oecd.org)

More access does not mean unrestricted reuse. The EU Data Act addresses access to data from connected devices, while the European Data Protection Board says GDPR requirements apply when scraping processes personal data. Data suppliers should distinguish the right to obtain a feed from the right to combine, retain, train models on or redistribute it. (digital-strategy.ec.europa.eu)

AI can make alternative data more usable by extracting patterns from large structured and unstructured sources and combining signals for review. ESMA's survey of EU securities-market firms records that perceived benefit, but also reports concern about data and model vulnerabilities, cybersecurity and dependence on third parties. It is evidence about surveyed financial firms' AI use—not a measurement of global alternative-data spending. (esma.europa.eu)

The effects accumulate across collection, labelling, modelling and distribution: an upstream coverage gap or misclassified observation can be carried into a downstream signal and then into a decision. ESMA separately warns investment-services firms about bias, data quality, opaque decisions and overreliance on AI. The practical implication is to preserve source-level audit trails and human challenge even when automated analysis becomes cheaper; model output cannot repair an unusable licence or an unrepresentative input. (esma.europa.eu)

North America has a clearly documented investment-adviser compliance use case in the United States and tourism-data experimentation in Canada and Mexico. Latin America has a distinct payments context: IMF research discusses Brazil's and Mexico's digital payment systems, which could generate useful transaction signals but do not establish that those signals are commercially licensed. Mexico is treated under North America in the featured-country coverage below. (sec.gov)

Europe combines documented tourism and transaction-data applications with EU rules relevant to connected-device access and personal-data processing; legal access must be assessed by dataset and intended use. Middle East evidence includes a Saudi Arabian mobile-positioning tourism application and a GCC programme for cooperation on data governance. Neither establishes region-wide commercial uptake. (oecd.org)

Africa offers a case for complementary geospatial signals: World Bank-described work in Malawi pairs satellite features with household observations rather than replacing survey ground truth. Asia-Pacific spans Australian mobility-data use, South Korean tourism-data integration and India's examination of non-conventional statistical sources. These examples demonstrate possible applications, not comparable regional market sizes. (blogs.worldbank.org)

ASEAN provides voluntary model clauses for cross-border personal-data transfers, subject to applicable national requirements. BRICS has articulated a data-economy governance understanding covering flows, privacy and interoperability while acknowledging fragmented regulation. The European Union has applicable connected-device data-access rules, but those rules are not a blanket permission to resell data. Each framework is relevant to product design; none by itself proves a unified market for alternative-data feeds. (asean.org)

The G7 has endorsed cooperation on data free flow with trust. The GCC has reported joint work on data governance and AI through 2026; a cooperation programme should not be mistaken for completed common legislation. NATO calls for secure data labelling, metadata and access controls within its defence-oriented digital strategy. NATO is therefore a distinct potential context for governed information sharing, not a civilian alternative-data licensing regime. Group initiatives overlap geographically and cannot be added together as separate demand estimates. (gov.uk)

North America: In the United States, SEC staff examinations document compliance concerns for advisers using alternative data. The OECD records tourism-source exploration in Canada and Mexico; IMF research adds context on Mexican digital payments. These observations identify use cases, not the countries' relative investment-data spending. Latin America: Brazil appears in the OECD's tourism-source survey, while IMF research discusses its digital payments as a possible source of transaction-derived insight. (sec.gov)

Europe: In France, INSEE describes card-transaction data for monitoring conditions and nowcasting. Germany has a documented North Rhine-Westphalia tourism data hub. OECD tourism research also records source exploration in Italy and the United Kingdom, and describes transaction and mobility applications in Spain. For Russia, the inspected material does not adequately establish the scope of its commercial alternative-data supply or buyer demand; no comparative market finding is made. These national examples should not be read as adoption rankings. (insee.fr)

Asia-Pacific: Australia uses mobile-positioning information to complement domestic tourism measurement; a South Korean tourism data hub brings together public and private sources. China promotes big-data methods in official statistics, while India has examined satellite, telecommunications and other non-conventional sources alongside established statistics. Japan's documented 2020-base consumer-price methodology used scanner data and selected web-scraped prices; that historical document does not establish every current collection method. The South Korean, Chinese, Indian and Japanese examples likewise do not quantify commercial feed sales. (oecd.org)

1. Approve sources for a stated use. Record the origin, collection method, onward-use rights, retention terms and potential material-nonpublic-information exposure of each feed. Reassess when a provider changes collection practices; SEC staff have observed failures in consistent onboarding and repeated diligence. (sec.gov)

2. Prove incremental value. Compare each signal with an appropriate conventional baseline using historical data available at the decision date. Check geographic, cash-spending and population-coverage gaps before promoting a tourism or transaction proxy into a broader economic claim. (oecd.org)

3. Set separate privacy and AI controls. Examine whether scraping or location records process identifiable people, and test models for input drift, bias and supplier failure. Keep a reviewer able to challenge consequential outputs. These are proposed operating controls informed by regulatory concerns, not a claim that regulators endorse this checklist. (edpb.europa.eu)

4. Localise contracts and delivery. Specify permitted transfer, processing, model-training and redistribution for each jurisdiction instead of assuming that a geographic grouping supplies one transferable licence. (asean.org)

This is a qualitative desk review of inspected primary regulator and government material, an OECD cross-country tourism study, and institutional financial and development research. It uses those sources to distinguish documented applications and policies from commercial inferences. No vendor interviews, contracts, transaction-level datasets or comparable country-by-country procurement records were examined. The OECD country examples concern tourism, the SEC alert concerns US adviser examinations, and ESMA's AI survey concerns EU securities-market firms; none measures the global alternative-data market. (oecd.org) Geographic coverage follows the supplied reporting scope, not a claim that featured countries or groupings are exhaustive. Evidence is uneven: Russia lacks a sufficiently supported commercial assessment here, and regional or intergovernmental policy statements cannot establish local sales, implementation or lawful rights in a particular feed. Accordingly, the report makes no market-size, share or numerical-growth claim.

Alternative data's opportunity is broader than an early investment signal: the inspected examples also show uses in tourism measurement, economic monitoring and geospatial analysis. Its constraint is equally broad: coverage, supplier continuity, permission to use the data and downstream model quality all affect whether an insight survives operational scrutiny. (oecd.org) Editorial assessment: products that pair useful signals with reproducible methods and defensible rights are better positioned for demanding buyers than products relying on novelty alone. AI may increase the value of well-governed inputs, but the evidence reviewed does not support a quantified forecast or a ranking of regions or suppliers. (esma.europa.eu)