Automated, fully sourced indicators
Failed accounting checks
Seconds needed for a full data refresh
About The Central Bank of The Bahamas
Every forecast cycle, the Research Department (the Department) at the Central Bank of The Bahamas (the Central Bank) has to answer one question: where does the Bahamian economy stand, and where is it headed?
For an economy pegged to the U.S. dollar, the answer depends on dozens of indicators, from GDP and inflation to reserves and financial-sector health, and on modeling how external shocks ripple through them. For years, that work ran through an Excel-based Financial Programming Model (FPM), fed by data pasted in by hand from international sources each cycle.
Today, the Department runs a fully coded forecasting system built on Python and the Anaconda Platform. It keeps the same trusted economics, cuts a multi-day forecast round to an afternoon, and gives the whole Department a shared, auditable view of the economy.
The Challenge: A Trusted Model, Built in a Workbook
The FPM is an International Monetary Fund (IMF) style framework that central banks use to turn economic assumptions into policy scenarios. It had served the Department well for years. But the workbook was slow to produce results, and that delayed the work that depended on them.
Each round meant pasting data by hand from sources like the IMF’s International Financial Statistics (IFS), World Economic Outlook (WEO) releases, and the World Bank into a spreadsheet held together by cross-sheet links that could break silently. Dr. Allan Wright, Manager in the Department, describes the old process: “The Financial Programming Model is a useful tool, which has served our department well. However, the process was particularly time-consuming. When you’re using spreadsheets, especially ones with a lot of macros and equations, they can get hung up on you very fast, and you can lose the connections between different sheets in the workbook.”
That Excel-based framework carried compounding risk:
- Manual data entry: Staff re-entered data from multiple external sources every cycle, with no audit trail showing where a number came from.
- Fragile cross-sheet logic: A single broken cell could spread errors silently through the model.
- Slow scenario runs: A full round took days and was hard to reproduce.
- Concentrated knowledge: Mainly whoever maintained the file understood how the model worked, so the process was hard to scale or hand off.
The Approach: Rebuilding the Model in Python
The Department kept the FPM’s economics and rebuilt its mechanics in Python on the Anaconda Platform. The model’s logic is now version-controlled, tested, and reproducible across the Department, with no dependence on a single workbook.
The system has two connected layers. The first is a data pipeline. It pulls from four official APIs (the World Bank, IMF WEO, IMF Statistical Data and Metadata eXchange (SDMX), and Federal Reserve Economic Data (FRED)) to automatically refresh a registry of 257 economic indicators across 14 categories. Every value is tagged with the source that supplied it.
The second is a set of model engines. They run the same nine-step financial-programming logic the Central Bank has always used, from measuring output gaps through the fiscal, external, and banking-sector blocks, in a tested codebase instead of a chain of spreadsheet formulas. A separate engine, a Python implementation of the European Central Bank’s Nowcasting Toolbox, gives the Department a live estimate of current-quarter GDP ahead of official data releases.
Building on the Anaconda Platform gave the team three things: a consistent, governed Python environment for developing and maintaining the codebase, trusted packages for the statistical and data-pipeline work, and a foundation the whole Department can run, test, and extend.
The Results
The new system has changed the pace of the Department’s forecasting work:
- From days to an afternoon: The mechanical work of a full forecast round now runs in an afternoon. A full data refresh takes about 20 seconds, and a multi-scenario projection runs in seconds.
- 257 indicators, fully sourced: Each indicator, across 14 categories, comes from official sources with a recorded audit trail showing where each value came from.
- Zero accounting failures: The system re-checks 15 national-accounting identities on every model run and currently reports zero failures. That replaces manual, late-night verification.
- Near real-time visibility: A nowcasting model combines three independent modeling approaches to give the Department a current read on economic conditions ahead of official data releases, without waiting on the quarterly reporting cycle.
- A shared, testable model: Documentation maps each mechanism back to the original workbook, so the model no longer depends on one analyst’s knowledge.
Looking Forward
The Department plans to extend the platform with GDP-by-industry detail, richer monitoring of data readiness, and more model families, including benchmark time-series models and cross-country comparisons.
Its ambitions reach further. The team wants to bring the same rigor to residential and commercial asset prices, including how hurricane exposure and climate risk could affect property values and insurance costs across the islands. That work ties to the multilateral climate financing the Bahamian government receives. A related effort would track the country’s changing tourism and cruise-passenger economics, as the visitor mix shifts toward a younger, higher-frequency segment with different spending patterns.
Longer term, Dr. Wright envisions a platform where researchers, students, other central banks, and the public can view live Bahamian economic data and run their own scenarios without contacting the Central Bank.
Want to scale AI from development to production on your own terms? Contact our team for a personalized demo.
“At any point in time, we have real-time information on all of the key variables that we have. We are now able to do advanced analysis, as well as forecasting on all of these variables.”
Dr. Allan Wright
Manager, Research Department, Central Bank of The Bahamas