This Request for Proposal (“RFP”) is part of a collaborative initiative between JICA DXLab and the National Bank of Cambodia (“NBC”) and aims to select a Digital Partner (“DP”) to develop an AI-based liquidity forecasting model Proof of Concept (“PoC)” solution. The AI-based liquidity forecasting solution will test and validate the effectiveness of different machine learning models to forecast liquidity. The DP will be responsible for development and testing of the AI-based liquidity forecasting solution.

The objective of the PoC is to validate that an AI-based solution can improve the accuracy of liquidity forecasting.

The DP is expected to provide a workplan in response to this RFP. However, an illustration of the implementation effort and the phases can be found below:

NBC is the central bank of Cambodia. The mission of the NBC is to determine and direct the monetary policy aimed at maintaining price stability to facilitate economic development within the framework of the kingdom's economic and financial policy. As the monetary authority, the NBC is the sole issuer of the Cambodian Riel, the national currency. This helps maintain monetary stability. As the supervisory authority, the NBC has the authority to license, delicense, regulate, and supervise banks and financial institutions in Cambodia. The NBC also conducts regular economic and monetary analysis, publishes various publications, oversees the nation's payment systems, establishes balance of payments, and participates in the management of external debt claims.

Cambodia’s monetary landscape is shaped by a high degree of dollarization, with foreign currency deposits accounting for over 80% of total money supply. .Within this environment, the NBC maintains a managed float exchange rate regime for the Cambodian Riel against the US Dollar. These structural characteristics significantly limit the traditional levers of monetary policy (i.e., exchange rate and domestic interest rate adjustments).

Liquidity management is a routine function of the central bank aimed at achieving the intermediate targets of monetary policy. Accurate liquidity forecasting helps align liquidity conditions with these targets and minimizes volatility that could result from the central bank relying solely on ex-post reactions.

Improving the forecasting of local currency liquidity is therefore a critical step towards enabling more autonomous and responsive monetary policy. Enhancing NBC’s technical capabilities in liquidity forecasting will allow for better informed data-driven decision-making, and more effective policy communication. These improvements are also expected to support the development of a short-term Cambodian Riel-denominated money market and promote broader use of local currency.

This joint initiative between NBC and JICA DXLab aims to develop an AI-based liquidity forecasting model. This model will support NBC’s efforts to enhance the accuracy and timeliness of its liquidity projections, thus contributing to improved monetary policy operations and resilience of Cambodia’s financial system.

Candidate DPs must fulfill all the following criteria to participate in the RFP (please refer to the RFP document for full details).

1. Japanese Company: The company must either be headquartered in Japan, including consolidated subsidiaries, or have a majority-Japanese management team.

2. Legal History: The company’s management must have no criminal record, and the company must not be included on any sanction list.

3. Financial and Organizational Stability: The company must have a stable revenue stream, with a minimum annual revenue of USD 1,000,000 in each of the last two fiscal years, and the company must have been in operation for at least five years.

4. Debarment: The company must not have been debarred/blacklisted by JICA in the past.

5. Data Security: The company must implement and maintain robust data security measures in alignment with industry standards to preserve data confidentiality throughout the PoC (e.g., access controls, device security measures).

Only proposals from qualified candidates will be considered. Proposals will be evaluated on technical and commercial criteria. The winning candidate will be selected based on combined technical and commercial evaluation scores. This structured approach ensures a transparent and merit-based selection of the most suitable bidder.

Please note that the Scope of Work may be negotiated after the award notice to accommodate the selected solution/proposal. The selected DP will be responsible for the development, testing, and validation of an AI-based liquidity forecasting solution following the scope of work (please refer to the RFP document for full details):

1. Target forecast variables: Daily projections for forecasting variables over a 7-day horizon

2. Data inputs: Expected that 5-10 years’ worth of sanitized data with multiple entries per day will be provisioned for model development and testing

3. Model universe: Three types of models will be required to be developed during PoC:
- a. Statistical time-series:
- b. Pure AI
- c. Hybrid models

4. Validation requirements
- a. Model cross-validation
- b. Forecast performance
- c. User/operational fit

5. Infrastructure: Hybrid cloud setup where NBC sanitized data is stored on NBC’s on-premises infrastructure and streamed to a cloud environment for computation

6. Development tools: Python is the prescribed development environment for the PoC and proprietary intellectual property and licensed tools that NBC cannot freely access should not be used

The maximum budget for the services is JPY 55,000,000 including all taxes and expenses. Proposals that exceed this budget ceiling will not be considered.

An online information session was hosted for prospective bidders on 7th August 2025 in Japanese to offer clarifications on the RFP and address received inquiries.