R&D Assurance Framework eligibility at a glance

IRAS currently requires applicants to meet three core conditions.

Requirement

IRAS requirement

Internal R&D process

Good internal evaluation process and supporting documentation demonstrating that projects included in the claim satisfy the R&D requirements

Project volume

At least 5 R&D projects undertaken in-house in Singapore in the YA of application

Qualifying expenditure

At least S$500,000 of qualifying R&D expenditure for those projects in the YA of application

For this purpose, the S$500,000 threshold relates to expenditure qualifying for the Section 14D deduction.

Meeting the numerical thresholds does not by itself establish that the company is ready for the Framework.

IRAS also examines the quality of the company's:

  • project assessment;
  • internal evaluation process;
  • supporting documentation; and
  • ability to substantiate the R&D claims included in the application.

Five projects and S$500,000 are entry thresholds—not evidence that an application will automatically succeed.

What does the R&D Assurance Framework provide?

Under the Framework, a successful applicant can obtain upfront certainty on its R&D claims for up to three YAs, including the YA of application, subject to the requirements and ongoing responsibilities set by IRAS.

For a business with recurring material R&D claims, that can improve predictability around an area of tax that otherwise depends heavily on technical project assessment and supporting evidence.

But "upfront certainty" should not be interpreted as an exemption from R&D substantiation.

Companies still need to:

  • identify their R&D projects properly;
  • maintain detailed project documentation;
  • apply the R&D requirements consistently;
  • calculate qualifying expenditure appropriately; and
  • support sample projects selected by IRAS for review in subsequent YAs.

A better way to view the Framework is:

Who is the Framework most relevant for?

The Framework is particularly relevant to companies that already treat R&D tax as a recurring part of their tax and innovation strategy.

That can include businesses that:

  • make material R&D claims every year;
  • operate several R&D projects concurrently;
  • maintain an established Singapore engineering or R&D function;
  • expect qualifying R&D expenditure to remain substantial;
  • have increasing management focus on tax governance;
  • want greater predictability around recurring R&D claims; or
  • want to institutionalise their R&D claim process rather than rebuild it at each filing deadline.

Typical company profiles can include:

  • multinational groups with Singapore R&D centres;
  • semiconductor and electronics businesses;
  • advanced manufacturers;
  • established software and technology companies;
  • engineering-intensive businesses;
  • biotech and life-sciences companies; and
  • larger Singapore technology companies with recurring R&D portfolios.

When the Framework may be premature

The Framework may be less appropriate where:

  • the company has fewer than five in-house Singapore R&D projects;
  • qualifying Section 14D expenditure is below S$500,000;
  • the company is preparing an isolated or one-off R&D claim;
  • technical documentation remains weak;
  • projects are identified only retrospectively at tax-filing time;
  • project boundaries are not consistently defined; or
  • there is not yet a reliable financial methodology for connecting qualifying R&D to expenditure.

In those cases, the better first step may be to strengthen the underlying annual R&D claim process before considering an Assurance Framework application.

The key question is readiness—not just eligibility

A company can meet the numerical thresholds and still not be operationally ready for the Framework.

The more useful question is therefore not only:

Are we eligible to apply?

It is:

Do we have an R&D tax process that can withstand detailed review across a portfolio of projects?

That distinction is important.

The Framework asks the company to demonstrate that its R&D claims are supported by a repeatable system rather than a collection of individually reconstructed projects.

What does IRAS review?

To apply, the company provides:

  1. the R&D Assurance Framework application; and
  2. a complete list of the R&D projects proposed for inclusion.

IRAS then selects sample projects for detailed review.

This means the application is testing two related things.

1. The R&D projects

The selected projects need to support the R&D positions being taken.

That means the underlying project files should be capable of explaining matters such as:

  • the R&D objective;
  • novelty or technical risk;
  • the systematic investigative and experimental study;
  • project boundaries;
  • relevant project timing; and
  • supporting technical evidence.

2. The company's R&D claim process

The broader company process should also demonstrate an ability to consistently:

  • identify potential R&D projects;
  • distinguish R&D from routine development;
  • assess technical eligibility;
  • maintain evidence;
  • calculate qualifying expenditure;
  • apply the correct tax treatment; and
  • prepare supportable annual claims.

The Framework is therefore not merely a review of several good projects.

It is also a review of the system that produced those claims.

How many projects does IRAS review?

IRAS currently states that it will select:

for detailed review.

For example, a company submitting:

  • 5 projects should expect a material proportion of its portfolio to be exposed to review;
  • 20 projects can still have up to 5 projects selected; and
  • a significantly larger portfolio may result in more projects being reviewed where 20% exceeds 5 projects.

The practical point is straightforward:

An Assurance Framework application should be prepared on the assumption that selected projects may receive detailed technical scrutiny.

The application should not rely on a small number of carefully prepared projects surrounded by a weaker project population.

R&D Assurance Framework application timeline

The Framework has a substantially earlier timetable than the normal corporate tax filing process.

The application form and complete project list should be submitted at least nine months before the tax-return filing deadline for the relevant YA.

The process currently operates broadly as follows:

Stage Timeline
Application and complete R&D project list submitted At least 9 months before the relevant tax filing deadline
IRAS identifies selected sample projects Within approximately 2 weeks
Company provides detailed information on selected projects Within 2 months of IRAS' request
IRAS meeting with taxpayer Within 2 months after IRAS receives the complete information
Additional information, where requested Generally within 2 months of IRAS' request
IRAS communicates outcome Within 2 months after the meeting or receipt of additional information

If the Framework review is not completed before the relevant filing deadline, the company still needs to comply with the applicable R&D claim requirements for that YA.

The practical consequence

The company needs to prepare its project population, technical files, financial methodology and internal process sufficiently early for the application and detailed IRAS review.

For companies considering the Framework, readiness work should therefore begin well before the nine-month application deadline.

What should be in place before applying?

We would assess readiness across six areas.

1. A complete R&D project identification process

The company should have a method for identifying potential R&D across its Singapore activities.

That process should not depend solely on:

  • an "R&D" department;
  • a specific accounting code; or
  • projects nominated by Finance shortly before tax filing.

It should capture technical work across relevant functions and then apply the R&D criteria consistently.

A recurring process might include:

  • annual or quarterly project identification;
  • nominated technical project owners;
  • defined project information requirements; and
  • review by the appropriate R&D tax stakeholders.

2. A consistent technical eligibility methodology

The company should be able to demonstrate how it distinguishes qualifying R&D from ordinary product or process development.

A robust methodology should address:

  • the technical objective;
  • the state of knowledge at project commencement;
  • novelty or technical risk;
  • the investigative and experimental study;
  • project start and end points; and
  • exclusions and routine activities.

The objective is consistency.

Two technically similar projects should not be treated according to entirely different internal standards simply because different teams prepared the claim.

See what qualifies as R&D ->

3. Contemporaneous technical evidence

IRAS expects R&D claims to be supportable through documentation.

For a mature R&D claimant, that should increasingly mean collecting evidence while the project is happening, rather than reconstructing everything when the tax return becomes due.

Depending on the work, evidence can include:

  • project plans;
  • technical research;
  • design documents;
  • architecture records;
  • test protocols;
  • experiment results;
  • simulation outputs;
  • technical issue records;
  • source-control history;
  • engineering notes;
  • failed approaches;
  • redesign decisions; and
  • technical conclusions.

The company does not necessarily need a tax-specific document for every engineering action.

What matters is that the underlying technical record can demonstrate the substance of the R&D position.

4. A repeatable financial methodology

The technical project then needs to connect to a consistent expenditure calculation.

That can require processes for:

  • employee allocation;
  • payroll extraction;
  • consumables;
  • outsourced R&D;
  • cost-sharing arrangements;
  • local and overseas activities;
  • Government grants;
  • relevant project expenditure; and
  • reconciliation to the company's financial records.

The financial methodology should follow the qualifying technical work rather than rely simply on broad R&D cost centres.

The relationship should remain traceable:

project -> activity -> resource -> expenditure -> tax treatment

Understand qualifying R&D expenditure ->

5. Clear governance and responsibilities

A mature R&D claim process normally involves several functions.

These can include:

  • R&D or Engineering;
  • Finance;
  • Tax;
  • senior management;
  • external tax advisers;
  • auditors; and
  • specialist R&D advisers.

The company should know who is responsible for:

  • identifying projects;
  • confirming technical facts;
  • collecting evidence;
  • approving project eligibility;
  • producing financial data;
  • validating expenditure allocations;
  • reviewing the tax position; and
  • approving the final claim.

Without clear ownership, R&D claims can become dependent on one individual or on repeated last-minute reconstruction.

6. An annual review and control process

The final element is quality control.

Before the annual claim is finalised, the process should be capable of identifying:

  • projects with insufficient evidence;
  • inconsistent eligibility positions;
  • unsupported employee allocations;
  • incomplete project periods;
  • local / overseas issues;
  • outsourced or CSA complexities;
  • grants not reflected in the calculation;
  • unexplained year-on-year movements; and
  • discrepancies between technical and financial records.

This is what turns claim preparation into a controlled recurring process.

From R&D claim preparation to R&D tax governance

For an ordinary annual claim, the company may ask:

What R&D did we perform last year and what can we claim?

For an Assurance Framework application, the more important question becomes:

Do we have a system that can consistently identify and substantiate our R&D claims every year?

That is a meaningful shift.

Retrospective model

Year ends -> Finance identifies R&D costs -> technical teams reconstruct projects -> claim is prepared

Governance model

Projects are identified -> eligibility is assessed -> evidence is retained -> expenditure is captured -> annual claim is validated

The second approach is more aligned with what an assurance framework is designed to evaluate.

It also reduces dependence on retrospective reconstruction even independently of the Framework itself.

How the Assurance Framework relates to the normal R&D claim process

Our normal R&D tax methodology follows six stages:

For an Assurance Framework applicant, those same stages need to become a repeatable internal operating model.

The company should be able to demonstrate that:

  • potential projects are identified consistently;
  • technical assessment follows a defined methodology;
  • expenditure is connected to qualifying activities;
  • technical and financial evidence is retained;
  • claims undergo internal review; and
  • the process can operate again in the following YA.

The Framework therefore sits above the normal annual claim process rather than replacing it.

See the R&D tax claim process ->

What happens after a successful application?

The Framework can provide upfront certainty for up to three YAs, including the YA of application.

However, the R&D process continues.

IRAS may continue selecting sample projects for review during the second and third YAs to ensure that the company continues complying with the R&D requirements.

Successful applicants therefore still need to:

  • identify new projects;
  • assess technical eligibility;
  • maintain detailed documentation;
  • calculate expenditure appropriately;
  • comply with the responsibilities set by IRAS; and
  • support subsequent sample reviews.

What if the company has fewer than five projects later?

IRAS states that a successful application is not automatically affected if the company undertakes fewer than five in-house Singapore R&D projects in the second or third year.

However, companies are encouraged to consider their expected future R&D pipeline before applying because sample reviews continue during the Framework period.

This reinforces an important principle:

What happens if the application is unsuccessful?

IRAS indicates that unsuccessful applicants will be informed of the reasons and recommended areas for improvement.

An unsuccessful Framework application does not necessarily mean that the business performs no qualifying R&D.

Depending on the facts, improvement areas could potentially relate to matters such as:

  • project selection;
  • technical substantiation;
  • internal evaluation procedures;
  • supporting documentation;
  • financial methodology; or
  • overall application readiness.

The appropriate response is therefore to understand the reason for the outcome and determine whether the underlying R&D claim process should be strengthened before a future application.

Is the Framework worth applying for?

Meeting the IRAS thresholds does not automatically mean that an application is commercially worthwhile.

The decision should consider:

Size and recurrence of the claim

A company with significant annual R&D expenditure and a substantial recurring tax benefit has more to gain from increased predictability than a company making a small or occasional claim.

Number and stability of projects

A stable pipeline of Singapore R&D projects makes the Framework more relevant than a one-off burst of technical activity.

Current documentation quality

Where project documentation is already strong, the incremental work required to prepare for the Framework may be manageable.

Where substantial remediation is first required, the company may be better served by strengthening its normal annual process before applying.

Governance objectives

For some larger businesses, greater tax governance and management certainty can be valuable independently of the immediate tax saving.

This can be particularly relevant where R&D tax benefits are material to:

  • forecasts;
  • effective tax rate;
  • investment decisions;
  • group tax reporting; or
  • Singapore R&D planning.

The commercial question is therefore:

Would greater certainty over our recurring R&D claims justify the governance and application effort required to obtain it?

How Reeve | R&D supports an Assurance Framework application

Our role is not limited to completing the application form.

The more important work is determining whether the company's R&D claim process is ready for IRAS review and strengthening it where required.

Phase 1 - Framework readiness assessment

We assess the company against the core application requirements and practical readiness factors.

This can include:

  • number of in-house Singapore R&D projects;
  • qualifying expenditure;
  • existing project-assessment process;
  • technical documentation;
  • financial methodology;
  • governance;
  • annual claim controls; and
  • expected future R&D pipeline.

Output: Framework Readiness Assessment

The output identifies:

  • whether the company appears suitable for the Framework;
  • readiness gaps;
  • priority remediation actions; and
  • an application plan.

Phase 2 - R&D portfolio review

We review the complete project population proposed for inclusion.

This can involve:

  • reviewing project boundaries;
  • validating the consistency of eligibility assessments;
  • identifying higher-risk projects;
  • reviewing available technical evidence;
  • checking financial linkages; and
  • identifying documentation gaps.

The objective is not to prepare only a small number of showcase projects.

It is to ensure that the overall project population follows a consistent R&D methodology before IRAS chooses its sample.

Phase 3 - R&D governance design

Where required, we help establish or strengthen the process used to manage R&D claims.

That can include:

  • project-identification procedures;
  • technical eligibility templates;
  • evidence requirements;
  • interview processes;
  • expenditure-allocation methodologies;
  • responsibilities and sign-offs;
  • annual review controls; and
  • claim timetables.

The objective is to create a process the company can operate repeatedly rather than a one-time application exercise.

Phase 4 - Application preparation

We prepare and coordinate the Framework application, including:

  • application information;
  • complete R&D project list;
  • supporting methodology;
  • project information;
  • technical explanations; and
  • relevant financial analysis.

The application should reflect the same methodology used for the company's underlying annual R&D claims.

Phase 5 - IRAS detailed review

Once IRAS selects the sample projects, we support the company through the detailed evaluation.

This can include:

  • preparing selected-project files;
  • validating technical narratives;
  • preparing financial substantiation;
  • coordinating technical stakeholders;
  • preparing for the IRAS meeting;
  • supporting explanations during review; and
  • responding to additional information requests.

Phase 6 - Ongoing assurance cycle

For successful applicants, the process continues during the remaining Framework period.

We can support:

  • identification of new projects;
  • technical assessments;
  • documentation quality;
  • annual expenditure calculations;
  • sample-project readiness; and
  • ongoing alignment with the company's tax filing process.

Is your company ready for the R&D Assurance Framework?

A practical readiness assessment should answer five questions:

  1. Do we have at least five in-house R&D projects undertaken in Singapore?
  2. Do those projects generate at least S$500,000 of qualifying Section 14D expenditure?
  3. Can we substantiate the technical eligibility of the project portfolio consistently?
  4. Can we trace the qualifying projects into a supportable financial calculation?
  5. Do we have an internal process capable of repeating this each year?

If the first two answers are yes but the remaining answers are uncertain, the priority should be readiness work before application.

If your company has a material recurring Singapore R&D portfolio, we can assess whether the Framework is appropriate, identify gaps in the existing claim process and determine what should be strengthened before an application is made.

R&D Assurance Framework FAQs

How many R&D projects do we need to apply?

At least five R&D projects undertaken in-house by the company in Singapore in the YA of application.

The project threshold is only one part of the eligibility requirements. The company must also satisfy the expenditure and internal-process requirements.

How much qualifying R&D expenditure is required?

The qualifying R&D expenditure for the relevant projects must be at least S$500,000 in the YA of application.

For this threshold, IRAS specifies expenditure qualifying for the Section 14D deduction.

How long does the R&D Assurance Framework last?

Subject to IRAS' requirements, successful applicants can obtain upfront certainty on their R&D claims for up to three Years of Assessment, including the YA of application.

Does successful acceptance mean IRAS will no longer review our R&D projects?

No.

IRAS continues to select sample projects for review in subsequent YAs, and companies are still required to maintain detailed supporting documentation for their R&D projects.

How many projects can IRAS select for detailed review?

IRAS currently states that it will select up to five projects or 20% of the total number of projects, whichever is higher for detailed review.

The company should therefore prepare its full R&D project population on a consistent basis rather than focusing only on a small number of selected projects.

When should we apply?

The application and complete R&D project list should be submitted at least nine months before the tax-return filing deadline for the relevant YA.

Because selected-project information and an IRAS review follow the initial application, readiness work should start well before this deadline.

Do we still need detailed R&D documentation after acceptance?

Yes.

IRAS requires businesses to continue maintaining detailed documentation of their R&D projects. Sample projects can continue to be reviewed during the second and third years of the Framework.

What if we have fewer than five projects in the second or third year?

IRAS states that this does not automatically affect a successful application.

However, companies are encouraged to assess their expected future R&D pipeline before applying because IRAS continues to select sample projects for subsequent review.