Article map
A mark-to-market number can appear precise while depending on stale rates, incorrect trade terms or an unapproved model. Treasury needs valuation not only for accounting, but also for counterparty limits, collateral, hedge performance, liquidity and decision-making. The number must therefore be reproducible and its movement explainable.
A controlled valuation framework connects trade data, market data, curves, models, independent checks, sensitivities and P&L explain. This article sets out the core components.
1. Start with complete trade terms
Valuation depends on legal entity, counterparty, instrument, currency pair, buy and sell amounts, rate or strike, trade date, value date, premium, exercise style, settlement and amendments.
Trade terms should come from the approved deal and matched confirmation. An unconfirmed amendment or incorrect sign can reverse value.
The model should preserve original and current versions.
2. Govern market-data sources
Spot rates, forward points, discount curves and volatility should come from approved sources with timestamp, market, contributor and quality status.
Hierarchy and fallback should be defined. A missing primary source should not silently use an old value.
Market cut-off should align with reporting purpose. Different time zones can create unexplained differences between treasury and counterparty valuations.
3. Build currency curves consistently
Forward valuation requires spot and interest or forward curves for relevant currencies and tenors. Interpolation and extrapolation should be documented.
Curve construction should be consistent with the instrument and collateral or discounting framework where material. Treasury should avoid mixing vendor points and internally derived rates without control.
Curve versions should be retained for reproduction.
4. Value forwards and swaps transparently
A forward can be valued by comparing contracted cash flows with current forward economics and discounting appropriately. An FX swap values the near and far legs and any settled component.
The system should show signed value, currency of value and reporting conversion. Users should understand whether the number includes accrued or settled cash.
Independent sample calculations can validate model implementation.
5. Value options with appropriate models
Currency options require volatility, time, rates, strike and model assumptions. Simple vanilla options may use standard models; exotic features require additional capability.
Volatility can vary by strike and maturity. The methodology should state whether a surface, smile or simplified input is used and whether the effect is material.
Model limitations and reserves should be documented where relevant.
6. Control calendars and time
Value depends on trade and settlement calendars, day count and time to expiry. Option value can change materially near cut-off.
The system should use approved holiday calendars and time zones. A one-day error can affect short-dated trades.
Valuation timestamp and market-data timestamp should both be visible.
7. Perform independent price verification
Front-office or treasury-system valuation should be compared with independent market data or counterparty statements according to materiality.
Differences should be decomposed into trade terms, market cut-off, curves, credit, model and rounding. A counterparty value is not automatically correct.
Tolerance, frequency, owner and escalation should be policy-driven.
8. Reconcile trade population
Valuation control begins with completeness. Active trades should reconcile to confirmations, counterparty statements, settlement and accounting.
Matured, terminated or novated trades should not remain active. Missing trades should not be added only through aggregate adjustment.
Population reconciliation should occur before valuation differences are analysed.
9. Calculate sensitivities
Useful measures include spot sensitivity, forward-point or rate sensitivity, option delta, gamma, vega and time decay where appropriate. Simpler portfolios may use scenario value changes.
Sensitivities should be aggregated by currency, entity, maturity and counterparty with sign conventions defined.
They support limit monitoring and help explain why value changes.
10. Use stress and scenario analysis
Scenarios can shift spot, rates, forward basis and volatility, including historical or hypothetical shocks. They should consider non-linear option payoff and concentrated maturity.
The output should show derivative value, underlying exposure and net risk. A derivative stress in isolation can misrepresent the economic position.
Liquidity effects such as collateral and settlement should be included.
11. Produce daily or period P&L explain
Value movement can be decomposed into spot, carry or forward points, rates, volatility, time decay, new deals, settlements, cash flows, amendments, model changes and residual.
Large unexplained residuals indicate data or model issue. The tolerance should reflect portfolio complexity.
P&L explain also improves management understanding and accounting reconciliation.
12. Measure counterparty exposure
Positive derivative value represents a current claim, subject to legal netting and collateral. Potential future exposure may be relevant for limits.
Exposure should aggregate across trades and products at legal counterparty and banking-group level. Collateral and netting should be recognised only when enforceable and operationally reliable.
Pending settlement and close-out timing can be material.
13. Connect valuation to hedge performance
Economic reporting should compare derivative value with the underlying exposure or benchmark. Hedge-accounting calculations may use designated components and specific methods.
The same market data and trade terms should be reconciled, even if calculations differ by purpose.
A derivative loss can be expected when the underlying gains; the net outcome matters.
14. Control manual adjustments and reserves
Valuation adjustments for model limitation, liquidity, credit or data should have methodology, owner, approval and effective period.
Manual plugs should not conceal unresolved trade or curve errors. Adjustments should be separately reported and reviewed.
Changes should preserve history and impact analysis.
15. Govern models and releases
Models should have documentation, owner, validation, version, approved use and change testing. New products should not be booked before valuation and risk capability exists.
Code or configuration change should be tested on representative trades and reconciled with prior results.
Independent review should be proportionate to complexity and materiality.
16. Report for different decisions
Dealers need position and sensitivity; finance needs valuation and accounting support; risk needs limits and stress; management needs residual exposure and material movement.
Views should use common data while presenting relevant aggregation. Users should be able to drill to trade and market input.
A single unexplained MTM total is inadequate for all audiences.
Build valuation governance and independent controls
Valuation should have a documented hierarchy of market data, curves, volatility surfaces, credit adjustments, calendars and conventions. The process should record source, observation time, fallback and any manual input. Independent price verification can compare selected positions with counterparty statements or alternative sources, while recognising that a counterparty value is not automatically an independent fair value.
Tolerance should reflect product, tenor, materiality and model uncertainty. A difference should be explained by inputs, methodology, timing or trade terms rather than overwritten. Model changes and valuation adjustments should be approved, impact-assessed and versioned so that period-to-period movement remains explainable.
Maintain a model-risk inventory
Each instrument should be mapped to the model used, key assumptions, known limitations and validation status. Plain forwards may be operationally simple, but options and structured products introduce volatility, correlation, barrier and interpolation risk. The inventory should identify products outside standard capability and require specialist review before dealing.
Validation should test implementation against specifications, benchmark results, sensitivities and edge cases. Production monitoring can detect unusual valuation jumps, stale data, zero sensitivities or breaks between trade economics and accounting output. This turns model governance into a continuing control rather than a one-time implementation document.
Connect analytics to liquidity and limits
Market risk cannot be managed only through mark-to-market. Treasury should translate sensitivities and scenarios into potential settlement, collateral, credit and liquidity needs. A hedge that protects the forecast value of a transaction may still create an interim cash requirement when market value moves adversely.
Dashboards should therefore connect exposure, derivative value, stress loss, collateral terms, counterparty capacity and settlement calendar. Decision-makers can then distinguish economic protection from funding risk and avoid adding a trade that solves one dimension while creating an unacceptable constraint in another.
Reconcile valuation movement before reporting risk
The valuation process should explain movement between opening and closing value through market rates, forward points, volatility, time decay, new trades, settlements, cash flows and model or data changes. An unexplained residual should be investigated before the number enters accounting or risk reporting. This bridge helps distinguish genuine market exposure from operational noise and provides a common basis for treasury, finance and independent review.
The bridge should be reproducible by trade and portfolio, with material manual adjustments separately approved. It is particularly useful when a net portfolio value appears stable because large offsetting movements conceal concentration or liquidity risk.
Practical illustration: explaining a forward loss
A company hedges a future USD receipt by selling USD forward. The domestic currency weakens, creating a negative value on the forward. A standalone report labels the hedge a loss.
P&L explain shows the movement arises primarily from spot and forward change, while the forecast receipt increases in domestic-currency value. The net economic position remains within policy. Management can distinguish expected hedge offset from execution or model error.
Implementation checklist
FX valuation and analytics should include:
- confirmed and versioned trade terms;
- approved, timestamped market-data hierarchy;
- controlled curve construction and interpolation;
- transparent forward and swap methodology;
- appropriate option models and volatility inputs;
- calendars, time zones and valuation cut-off;
- independent price verification;
- complete active-trade reconciliation;
- defined sensitivities and aggregation;
- market and liquidity stress scenarios;
- P&L explain with residual thresholds;
- counterparty exposure, netting and collateral;
- underlying and hedge performance linkage;
- approved adjustments and reserves;
- model inventory, validation and release control; and
- audience-specific reporting with drill-down.
Common valuation failures
Common failures include using stale rates, inconsistent sign conventions, missing novations, accepting counterparty values without review, valuing options with flat volatility without assessing materiality, ignoring time-zone differences and reporting derivative MTM without underlying exposure.
Another failure is changing a curve or model and treating the resulting P&L as market movement.
Closing perspective
Valuation is a controlled explanation of what a financial instrument is worth and why that value changed. It depends on accurate trade population, governed market data, appropriate models and independent review.
When sensitivities, P&L explain, counterparty exposure and underlying risk are connected, treasury can use valuation for decisions and reporting rather than treating it as a black-box period-end output.
Frequently asked questions
What drives the value of an FX forward?
Value depends on contracted rate, current spot and forward rates, remaining tenor, currencies, discounting and settlement terms.
What additional data is needed for currency options?
Option valuation also requires volatility, strike, expiry, exercise style and model assumptions, with attention to smile or surface effects where material.
Why is P&L explain important?
P&L explain decomposes value movement into market changes, passage of time, new trades, settlements, cash flows, model or data changes and residuals, improving control and management understanding.