The Forecasting Gap: Why Most Treasury Models Underestimate FX Impact

Every finance leader managing a global operation has experienced a version of the same conversation. The quarterly treasury model projected an FX impact within acceptable range. The actual P&L showed something materially worse. The post-mortem identified the usual suspects which include unexpected currency volatility, a rate movement that fell outside the model's assumptions and a market event that nobody anticipated. 

These explanations are not wrong. But they are incomplete. Unexpected volatility and market events are real, but they account for a smaller proportion of the FX forecasting gap than most treasury teams acknowledge. The larger proportion is structural — the result of modelling assumptions that systematically underestimate FX exposure, undercount the transactions through which that exposure is realised, and undervalue the operational decisions that determine the cost at which exposure is converted. 

The forecasting gap is not primarily a volatility problem. It is a modelling problem. And modelling problems are fixable, but only once they are correctly diagnosed. 

The Four Structural Sources of Treasury Model Underestimation 

Understanding where the forecasting gap comes from requires examining the specific modelling assumptions that produce it. Four structural sources account for the majority of the discrepancy between projected and actual FX impact in most treasury models.

The Execution Gap — What Models Assume and What Actually Happens 

Beyond the structural modelling sources, there is a second category of forecasting gap that deserves specific attention: the execution gap which is the difference between the conversion rate the model assumes and the rate the business actually achieves. 

Most treasury models assume conversion at or near the mid-market rate, sometimes with a modest adjustment for known banking spreads. This assumption is optimistic in ways that compound with transaction volume. 

In practice, the execution rate on any given conversion is determined by four factors that models rarely capture with sufficient precision: 

• Timing of conversion 
FX rates vary throughout the trading day based on liquidity and market activity. Conversions executed during low-liquidity periods such as early morning, late afternoon, or outside major market hours, consistently achieve worse rates than conversions executed during peak liquidity windows. For businesses converting reactively when a payment falls due rather than when conditions are favourable, the timing penalty is systematic and persistent. 

• Corridor-specific spreads
The spread applied to a USD/SGD conversion is different from the spread applied to a USD/IDR or USD/KES conversion. Treasury models that apply a uniform spread assumption across all currency pairs underestimate the cost of conversions in less liquid corridors, precisely the corridors where growing businesses are likely increasing their payment volumes. 

• Tranche size and market impact
Large conversions executed as single transactions can move the market against the converting party — achieving a worse blended rate than smaller tranches executed across a window. This is well understood in FX markets but frequently ignored in treasury models that treat each conversion as a single execution event at a point-in-time rate. 

• Banking relationship terms
The specific terms of the business's banking relationships, which are rarely uniform across all currency pairs, determine the actual spread applied to each conversion. Treasury models that use published or indicative rates rather than the specific terms applicable to the business's actual banking relationships introduce a systematic optimism bias into execution rate assumptions. 

The execution gap is not random. It is directional and actual execution rates are consistently worse than modelled rates, for structural reasons that repeat across every conversion cycle. This means the execution gap does not average out over time. It accumulates. 

What Better Treasury Modelling Requires 

Closing the forecasting gap requires addressing both the structural modelling sources and the execution gap, which in practice means changing several deeply embedded modelling conventions. 

• Model gross flows, not just net exposure
The model should capture every conversion event, not just the net exposure position, including the conversion costs incurred on gross flows that are netted at the position level but not at the P&L level. This requires more detailed cash flow forecasting but produces a model whose output is comparable to what the P&L will actually record. 

• Model intraperiod exposure, not just period-end positions
For businesses with significant intraperiod cash flow variability, the model should capture exposure at the point of conversion, which requires payment timing data, not just expected period-end positions. This is more operationally demanding but significantly more accurate. 

• Model execution rates, not expected rates
Treasury models should incorporate realistic execution rate assumptions, based on the business's actual banking relationship terms, historical spread data by currency pair, and conversion timing patterns, rather than mid-market or forward rates. The gap between indicative and execution rates is not noise. It is signal. 

• Capture all conversion events
The model should include the full population of conversion events, including card transaction conversions, platform payment provider conversions, and implicit conversions embedded in banking fee structures, not just the major, visible treasury transactions. Building this comprehensiveness requires finance teams to audit their actual conversion event population, which consistently reveals categories of FX cost that were previously untracked. 

The Role of Real-Time Data in Closing the Gap 

The structural modelling improvements above are necessary but not sufficient. They produce a better model — one whose inputs more accurately reflect the business's actual FX exposure and conversion cost structure. But a better model built on lagged data will still produce forecasts that diverge from outcomes as conditions change. 

Closing the forecasting gap to its irreducible minimum, the portion attributable to genuine uncertainty rather than modelling inadequacy, requires real-time data feeding the model continuously rather than historical data updating it periodically. Real-time FX exposure data changes what treasury modelling can do: 

• Intraperiod exposure tracking becomes possible when transaction data flows in real time rather than at end-of-day batch intervals — allowing the model to reflect exposure as it builds rather than after it has already been realised

• Execution rate monitoring becomes actionable when the gap between expected and achieved rates is visible in real time — allowing treasury teams to adjust conversion timing and routing before the pattern becomes a quarterly variance

• Conversion event capture becomes comprehensive when payment data from all channels flows into a unified view in real time — closing the undercount that leaves significant FX cost invisible in period-end modelling

The businesses whose treasury models most consistently predict actual FX impact are not the ones with more sophisticated modelling methodologies. They are the ones with more complete, more current data — and infrastructure that makes that data available to the model continuously rather than periodically. 

The Gap Is Measurable. The Causes Are Structural. The Fixes Are Available. 

The forecasting gap between treasury model projections and actual FX P&L impact is not an unavoidable consequence of currency market uncertainty. It is a predictable consequence of modelling conventions that systematically underestimate gross exposure, miss intraperiod flows, assume execution rates that are not achievable in practice, and undercount the full population of conversion events. 

To get started and partner with a solutions provider that can help your business optimise payments and help you scale both locally and globally, open a SUNRATE account today or contact our sales team.