Direct answer: what data you need
To assess “USD concentration,” you need data that lets you estimate how much of your relevant exposure is denominated in US dollars (USD), plus enough metadata to verify the estimate. In practice, collect (1) the underlying positions or cashflows, (2) the currency of each item, (3) amounts and units, (4) the timestamps that define “as-of,” and (5) any conversion inputs you use when translating to a common measurement currency.
Because concentration is a measurement, not a guaranteed outcome, the key is data provenance (where the numbers came from), timeliness (when they were observed), and quality (whether the inputs are complete and internally consistent).
Mechanics: define USD concentration and map inputs
USD concentration is usually expressed as a share: the proportion of total exposure that is denominated in USD.
A common high-level approach is:
- Define the scope of “exposure” (examples: spot and forward holdings, receivables/payables, debt obligations, margin balances; pick one and state it).
- For each exposure item, record the currency of denomination (USD or non-USD).
- Convert each item into a common base measure if needed (e.g., all items valued in USD-equivalent or in a reporting currency).
- Sum USD-denominated exposure and total exposure within the same scope and time window.
- Compute USD concentration as: USD exposure ÷ total exposure.
Data inputs to make that work:
- Position-level or cashflow-level detail: instrument/ledger item identifier, amount, and currency.
- Reporting convention: whether amounts are nominal, notional, marked-to-market, or booked values.
- FX conversion inputs (only if conversion is required): the exchange rate definition you apply, the rate timestamp, and whether you use mid, bid/ask, or a specific policy.
- “As-of” timestamp: when the positions were measured and when the FX rate snapshot applies.
- Consistency rules: how to treat netting, offsets, collateral, and internal transfers (state the rule you apply).
Assumptions must be explicit. For instance, if you convert non-USD items, you must assume the chosen FX rate is appropriate for the valuation moment you are analyzing. If you cannot justify an assumption, you should report the limitation.
Evidence or example: what a usable dataset looks like
A dataset that supports independent verification typically includes:
- A table of exposure items with columns: item ID, currency, amount, measurement type (e.g., notional vs. valued), and effective date(s).
- A valuation or aggregation definition: which items are included/excluded (e.g., only open positions, or also committed cashflows).
- A separate FX table (if conversions are used): rate value, rate source, and rate timestamp.
- An audit trail: how amounts were extracted (system/export name), and whether any transformations were performed.
Material example assumption (keep it testable): “As-of date is 2026-08-17 00:00 UTC; FX rates are applied from the same timestamp; concentration is computed using open exposures only; no collateral is netted.” In real use, you should adapt the “as-of” definition to your context, but the structure should remain reproducible.
If your scope is broad, missing exposures become a dominant risk. For example, exposures that are off-ledger, embedded in derivatives valuation, or captured only in separate systems may cause the USD share to be overstated or understated.
Limitations and risks: failure modes to plan for
At least one major limitation is typically “incomplete or mis-scoped exposure.” Common failure modes include:
- Scope mismatch: mixing spot holdings with different effective dates or different measurement bases.
- Hidden currency exposure: instruments where USD exposure is embedded but not recorded as a simple “currency” field.
- Incorrect conversion policy: applying FX rates with a different timestamp than the “as-of” positions, or using inconsistent rate types.
- Netting/collateral handling errors: treating offsets differently across items can change the numerator and denominator.
- Regime shifts: historical currency relationships (or past behavior) do not make future concentration stable; concentration can change when positions roll, mature, or get rebalanced.
Uncertainty should be treated as part of the result. You can often reduce uncertainty by narrowing scope, improving timestamp alignment, and documenting conversion and netting rules.
Verification or next question: how to check the assessment
Independent verification requires that someone else can redo the calculation from your data with the same inputs and rules.