Media Releases
Published Date: 14 March 2018

MAS Moves Towards Zero Duplication of Data Requests to Financial Institutions

Singapore, 14 March 2018… The Monetary Authority of Singapore (MAS) today unveiled the roadmap to transform its data collection approach from financial institutions. The roadmap includes measures to progressively reduce duplication and automate data submission by financial institutions. This will help financial institutions reduce the resources and preparation time needed to produce data requested by MAS. It will also make it more efficient for MAS to process and analyse the data collected. The measures will take effect from 31 March 2018.

Non-Duplication of Data Requests

2 To ensure that the data collected is managed and used effectively, MAS will enshrine data reusability1 in its data collection approach and ensure that financial institutions need not submit the same data to MAS twice.

3 Financial institutions will be allowed to decline any request from MAS for structured data2 that they have previously provided in their regular regulatory submissions to MAS. MAS will extend this treatment in subsequent phases, to other data submitted to MAS through surveys and one-off requests. The aim is to eliminate all duplication in data requests by the end of 2019. (Further details are provided in Annex 1.)

Machine Readability of Data Submissions

4 All new regulatory returns from financial institutions to MAS from 1 April 2018 onwards will have to be in machine-readable formats, i.e. in a form that allows data to be automatically read and processed by computers. The machine-readability format will be extended to new surveys and ad-hoc data requests from 2019 onwards.

5 Collecting machine-readable data is a key element of MAS’ plans to automate its data collection process. Automated data collection will increase efficiency by removing the need for manual processes and reduce the risk of human errors. MAS will seek feedback from financial institutions on providing data in the machine-readable templates prior to 2019 and take into account this feedback before the second phase of implementation. (Further details are provided in Annex 2.)

Granular Data Collection

6 MAS will also change how it defines its data requirements. For data that can be aggregated in different ways, MAS intends to collect more detailed data on the underlying transactions instead of the aggregate statistics. For example, instead of asking banks to submit separate statistics on loans extended to each industry and in each country, MAS might ask for each loan labelled with the industry and country that the loan is associated with. This enables MAS to manipulate the datasets internally according to its analytical needs while reducing the reporting burden on financial institutions. MAS is working with the industry to determine the appropriate levels of granularity in its data collection.

7 Dr David Hardoon, Chief Data Officer, MAS, said, “MAS is doing a fundamental review of its data collection approach. We will implement these measures in close partnership with the financial industry within a reasonable timeframe. MAS will also upgrade our data collection infrastructure to support these changes. This is an opportunity for both MAS and financial institutions to co-create an industry data collection platform that not only benefits MAS as a regulator, but also allows financial institutions to leverage the data collected to improve their operations.”



Information on non-duplication of data submissions

1 From 31 March 2018, regulated entities will be granted the right to decline any request from MAS for structured data that they have previously provided to us in an MAS-administered regulatory return. The right to rejection will be limited as follows:

(a) It only applies to data requested and given to MAS after 31 March 2018. If regulated entities had provided the data to MAS prior to 31 March 2018, they will have to submit it again if so requested.

(b) It does not apply to data requested under existing regulatory returns. MAS will phase out the collection of overlapping data in regulatory returns as they are updated. In the interim, regulated entities may not decline to fill out certain sections of existing regulatory returns that they think overlap with other returns.

(c) It only applies to requests for exactly the same data3.

(d) If MAS is prohibited from using the data for the purpose of the new collection by law or contractual clauses, MAS may have to collect the data again.

(e) If regulated entities wish to exercise the right to decline MAS’ data request, they must prove4 within 14 calendar days of the submission deadline that they had previously submitted the data.


Information on machine-readable format

1 From 31 March 2018, submission templates in machine-readable format will apply to new regulatory collections5 that meet all of the following criteria:

(a) Contains only structured, alpha-numerical data6.

(b) Contains only data that exists in electronic form7.

(c) Issued under an MAS-administered legislation on or after 31 March 20188.

2 Machine-readable templates will include both partially machine-readable formats, such as MS Excel-based files, and more fully machine-readable formats, such as CSV files. Regulated entities may only submit data using these templates. This will not apply to data collected under existing regulatory collections.

Annex 3




1Data reusability is the ability to use data collected once for multiple purposes. It facilitates consistency in analyses and promotes greater efficiency in the data collection process.
2Structured data refers to information that is arranged in an organised manner that allows for it to be easily searchable and analysed.
3For example, MAS may have previously requested for gross profits from 2016Q1, 2016Q2, and 2016Q3. MAS then issues a new request for gross profits from 2016Q1, 2016Q2, 2016Q3, and 2016Q4. The regulated entity may decline to provide the gross profit number for 2016Q1, 2016Q2, and 2016Q3. However, they must still provide the gross profits for 2016Q4.
4Evidence that the data was successfully submitted to MAS previously. This would include: the return that the data was submitted under, the date that the data was submitted, and the means through which the return was submitted (e.g. email, MASNET).
5Mandatory data collections issued under MAS-administered legislation.
6Unstructured data, such as an email or annual reports, need not be submitted in machine-readable formats.
7Data that only exists in hardcopy as of 31 March 2018 will be subject to machine-readability requirements at a later date.
8New collections not issued under any MAS-administered legislation, such as surveys and ad-hoc data requests, need only be submitted in machine-readable formats from Q1 2019.