Skip to content

8 Benchmark Rate Setting

Benchmark integrity depends on the full chain from evidence and submission to calculation, governance, publication, surveillance and audit. Conflicts and influence are recurring themes.

Benchmarks are reference rates or prices used in valuation, contracts, settlement and risk management. Their credibility depends on reliable inputs, transparent methodology, independent controls and ethical conduct.

Surveyed benchmarks use contributor judgement or quotations; traded benchmarks use eligible transaction data. The evidence, controls and risks differ, but both require integrity and a reasonable basis.

IOSCO principles provide an international reference for benchmark governance, methodology, quality, accountability and oversight. They inform good practice but do not replace Singapore law or the relevant administrator’s rules.

Benchmark governance separates oversight, administration, calculation and contribution responsibilities. Clear role allocation reduces conflicts and makes accountability traceable.

An oversight committee monitors the benchmark framework, methodology, conflicts, complaints and changes. It provides governance challenge rather than making unauthorised dealing decisions.

The administrator owns or oversees the methodology and process, sets governance and controls, manages contributors and publishes or arranges publication of the benchmark.

The calculation agent applies the published methodology to eligible inputs and calculates the benchmark. It needs reliable data, controls, auditability and appropriate escalation for anomalies.

A submitter contributes rates or transaction information on behalf of a market participant. Submissions must be authorised, representative, evidence-based, controlled and protected from influence.

Contributions should be based on the applicable hierarchy of evidence and the published methodology. Judgement is permitted only where supported, documented and not used to manipulate the result.

A reasonable basis may include actual transactions, relevant market data, quotes and documented expert judgement in the permitted hierarchy. The submitter should retain evidence for the contribution.

Traded-benchmark inputs should be bona fide eligible transactions with accurate timestamps and an audit trail. A trade executed to move or later cancel the benchmark is prohibited manipulation.

When liquidity is limited, a submitter may use the next permitted evidence level, such as related transactions, proxy assets, indicative prices or expert judgement, with adjustments and documentation.

Submitters must act with integrity, professionalism and ethical standards, follow the methodology, substantiate contributions, manage conflicts and protect the contribution process from influence.

Policies should exclude unsuitable data, support objective evidence, mitigate conflicts, preserve electronic trade integrity and prevent inappropriate influence on staff involved in submissions.

A person responsible for contributions should not comment before publication where possible. If commentary is permitted, actual, potential or perceived conflicts must be prominently disclosed in writing.

Collusion, wash trading, deliberately inappropriate prices, cancelling trades after the calculation window and withholding bona fide trades to affect a benchmark are prohibited. Genuine competitive trades remain legitimate.

For a benchmark not overseen by the administrator, the participant should follow the relevant owner’s laws, regulations and practices. If none exist, the Blue Book requirements may guide conduct.

Best practice combines fit-and-proper staff, supervision, information management, records, confidentiality, segregation, escalation, surveillance and independent assurance.

Contributors should have appropriate experience, seniority, fit-and-proper character, training and formal authority. Responsibility should not be delegated to an inexperienced person simply because the task is inconvenient.

Supervision includes dual controls, pre-submission validation and a checker senior enough to challenge the submitter. Oversight should be documented rather than assumed.

Historical submissions, trends and exceptions should be analysed and presented to the appropriate risk forum. Information management supports detection of drift, bias and unexplained changes.

Market data, submissions, supporting information, policies, deviations, staff roles and queries should be retained for at least five years or the longer period required by law.

Contribution information must not be sent to people outside the contribution process except regulators or another permitted recipient. General market information may be exchanged if it does not reveal desired benchmark prices or protected contributions.

Policies should minimise conflicts through appropriate seating, access and reporting separation among submitters, sales and trading and reviewers. Segregation should be proportionate and documented.

Suspicious, inappropriate or breached contributions should be escalated through whistleblowing and management channels and, where appropriate, to the administrator and MAS.

Electronic communications involving submission parties should be monitored under appropriate surveillance controls, with alerts, review, retention and escalation for suspicious conduct.

The benchmark process should be covered annually by independent audit or assurance, with results available to MAS on request. Product control or valuation functions may perform more frequent checks for surveyed benchmarks.

  • Surveyed vs traded: surveyed uses contributor judgement/quotations; traded uses eligible transaction data. Both need integrity and a reasonable basis.
  • Parties: oversight committee (governance challenge), administrator (owns methodology), calculation agent (applies it), submitter (contributes evidence). Clear role separation reduces conflicts.
  • Evidence hierarchy: actual transactions → related/proxy transactions → indicative prices → documented expert judgement; only the next permitted level where liquidity is limited.
  • Prohibited: collusion, wash trading, deliberately inappropriate prices, cancelling trades after the calculation window, withholding bona fide trades. A trade executed to move and later cancel the benchmark is manipulation.
  • Commentary: avoid public comment before publication; if permitted, disclose actual, potential or perceived conflicts prominently in writing.
  • Records: retain for at least five years or longer if required by law.
  • Trap: genuine competitive trades remain legitimate — the prohibition targets intent to affect the benchmark, not trading itself.

Practise this chapter →