TLDR
- The Caribbean receives over US$18 billion in remittances annually. Jamaica alone received approximately US$3.2 billion in 2023, representing roughly 18% of GDP.
- Not one Caribbean credit bureau incorporates remittance receipt data as a scored credit signal. The data sits locked inside money transfer operator systems, producing zero signal for applicants who depend on it as their primary income.
- A household receiving reliable monthly remittances for five years has demonstrated more consistent financial behaviour than many formally employed applicants, yet scores lower on every standard credit model.
- Maestro AI Labs' Data Archaeology product recovers exactly these non-traditional signals. Credit Garden's 302-point score uplift for credit-invisible Caribbean applicants is built on this data foundation.
- The window to build the Caribbean credit data infrastructure is open. The remittance flows already exist. What is missing is the structuring layer that converts them into scores.
What Remittances Actually Are as a Data Signal
A remittance is not just a money transfer. It is a packet of behavioural information about two people, a financial relationship, and an economic corridor.
On the sender side: consistent remittances over 24 months demonstrate reliable income, disciplined saving behaviour, and sustained relationship maintenance. These are the same attributes that consumer credit models try to infer from payment histories and utilisation rates. A Jamaican diaspora member in the UK or US who has sent money home every month for three years is demonstrating creditworthiness. No bank has ever scored that fact.
On the recipient side: regular remittance receipt demonstrates address stability, family relationship networks, and predictable income flow. In economies where formal employment is irregular, remittances may be the most stable income a household has. A Haitian household in Port-au-Prince receiving consistent transfers from a family member in Miami is demonstrating income stability that a credit model should interpret as a positive signal. Instead, it produces silence.
The reason for that silence is structural. Remittance data sits inside the closed systems of money transfer operators: Western Union, MoneyGram, and a range of smaller regional providers. These companies do not share transaction data with Caribbean credit bureaus. Caribbean credit bureaus have not historically sought it. The result is a predictable information gap with a concrete cost.
The Numbers Behind the Gap
| Country | Remittances (2023 est.) | Share of GDP | Formal credit penetration |
|---|---|---|---|
| Haiti | ~$3.8B | 26% | <15% of adults |
| Jamaica | ~$3.2B | 18% | ~28% of adults |
| Dominican Republic | ~$10.4B | 8% | ~40% of adults |
| Guyana | ~$600M | 5% | ~25% of adults |
| Barbados | ~$230M | 4% | ~55% of adults |
| Trinidad & Tobago | Net sender | N/A | ~48% of adults |
The table makes the problem concrete. Haiti and Jamaica, two of the Caribbean's largest economies by population, are simultaneously the most remittance-dependent and the most credit-excluded. These are not coincidental facts. They are connected: when remittance income is not counted as credit-qualifying income, the households most dependent on it are structurally excluded from formal credit.
The World Bank's 2021 Findex survey found that across the Caribbean and LATAM, approximately 40 to 60 percent of adults do not have access to formal credit. The IDB estimates the SME credit gap for the region at over US$380 billion annually. These numbers are not primarily a banking infrastructure problem. They are a data problem. The credit signals exist. They are just in the wrong place.
Why the Existing System Produces This Outcome
Credit bureau infrastructure in the Caribbean was built by financial institutions, for financial institutions. It captures what banks can see: loan repayments, credit card utilisation, mortgage performance, utility bill payments where these are reported. It was not designed to capture the informal and semi-formal financial behaviour that characterises how a large share of Caribbean households actually manage money.
SUSU participation records are not scored. Informal lending circle repayment histories are not tracked. Mobile money transaction patterns are not incorporated. And remittance receipt histories, which represent the most regular, most verifiable non-bank income flows in the region, are entirely absent from every credit model deployed in the Caribbean today.
This is not a technical failure. It is a decision failure. The data exists. The decision not to recover it and structure it as credit signal was made by default: credit bureaus did not pursue it, and money transfer operators had no commercial incentive to offer it. The result is a market with US$18 billion in annual financial flows that has produced zero basis points of credit score improvement for any of its participants.
"Every month a Caribbean household receives a remittance without it counting toward their credit file is a month of evidence destroyed. Five years of reliable receipts is 60 months of destroyed evidence. That is the cost of the status quo."
What Recovery Looks Like
Recovering remittance data as credit signal requires three things: access to the data, a methodology for structuring it as a scored signal, and a legal framework that permits its use under Caribbean data protection law.
On data access: the pathway is not forcing money transfer operators to share data. It is building consent-based portability. A credit applicant who consents to sharing their remittance receipt history with a lender can authorise the transfer of that data. This is the same model used in open banking contexts in the UK and EU, where account transaction data can be shared with third-party lenders with explicit customer consent. Jamaica's Data Protection Act 2020 and Barbados's Data Protection Act 2019 are both compatible with this approach.
On scoring methodology: raw receipt amounts alone are not sufficient. What matters is regularity (consistent monthly receipts carry more signal than irregular large transfers), duration (a three-year consistent receipt pattern is more predictive than a six-month one), and corridor stability (transfers from the same sender in the same country suggest a stable relationship, which correlates with income stability for the receiving household). Maestro AI Labs' Data Archaeology work has produced the methodology and historical data to calibrate these weightings for Caribbean market populations.
On legal frameworks: Jamaica's Data Protection Act 2020 and Barbados's Data Protection Act 2019 both provide for consent-based data processing. Using remittance data in credit scoring, with explicit applicant consent, is permissible under both frameworks. There is no legal barrier to building remittance-inclusive credit scoring in the Caribbean's main markets right now.
Credit Garden's Position in This Market
Credit Garden is the product that sits at the intersection of Data Archaeology's data recovery work and the Caribbean credit gap. Its 302-point score uplift for credit-invisible applicants, documented in the Caribbean credit frontier article, comes from exactly the kind of non-traditional signal recovery described above.
The 2.3 million proprietary records in Credit Garden's data foundation include remittance receipt patterns recovered through consent partnerships, SUSU participation histories obtained through cooperative partnerships, mobile money transaction summaries, and utility payment records not captured in standard bureau data. These signals are not peripheral to the credit model. They are the model's primary advantage in markets where traditional bureau data covers fewer than 30 percent of adults.
The commercial opportunity is clear. Any Caribbean or LATAM lender that wants to serve the credit-invisible market needs a data advantage to do so responsibly. Lending to people you cannot score is not financial inclusion; it is unsecured guessing. Lending to people you can score with alternative data, because you have invested in building the infrastructure that makes scoring possible, is the business model.
The Remittance Fee Opportunity
The credit scoring story is half of the remittance data opportunity. The other half is the fee.
Caribbean remittance corridors are expensive. The World Bank's Remittance Prices Worldwide database reports average costs of 6 to 7 percent to send US$200 to Caribbean destinations as of 2024. The UN SDG 10.c.1 target is under 3 percent globally by 2030. The gap between 6.7 percent and 3 percent represents roughly US$630 million in excess fees paid by Caribbean recipients annually on the $18 billion in inflows.
AI-native remittance services can narrow this gap through better fraud detection, more accurate real-time exchange rate management, and reduced manual compliance processing. The same data signals that improve credit scoring (regularity of remittance patterns, sender reliability scores, corridor risk profiles) also improve the risk management that drives pricing. A lender or remittance platform with three years of consistent remittance history for a corridor can price that corridor's fraud risk more accurately than one relying on static rules, and pass the savings on as lower fees.
What Builds Next
The Caribbean remittance credit data opportunity is not a research project. It is a buildable product. The data exists. The methodology to score it is established. The legal frameworks that permit it are in place. The market that needs it is the majority of the Caribbean adult population.
StarApple AI's network, which includes Maestro AI Labs, CreditGarden, and the StarApple AI platform, is the organisation with both the data infrastructure and the Caribbean market understanding to build it. The Caribbean's first AI company, founded by Adrian Dunkley in Kingston in 2023, has built the data assets and scoring methodology over three years of Caribbean market work that no external entrant has replicated.
The Caribbean AI community is building the infrastructure to support this. AI Jamaica, Caribbean AI Association, AI Guyana, and AI Barbados each track development in their respective markets. The Caribbean AI Risk Management Council has published guidance on alternative data governance for Caribbean lenders. The regional knowledge network exists. What it needs is the data product that converts remittance signals into scores, and that product is Credit Garden.
US$18 billion flows into the Caribbean every year. Every dollar of it is a data point. It is past time to score them.