Positions, accounting balances and risk exposures
A position records a holding or obligation, an accounting balance records amounts under defined posting and valuation policies, and a risk exposure measures sensitivity or potential loss under specified assumptions. These quantities can describe the same activity without being numerically equal.
Quantities, values and modeled effects
A position can be expressed as a number of securities or a contractual amount. A valuation assigns a monetary value using a price, curve or model. An accounting balance incorporates the policies and events relevant to the book in which it is recorded.
Risk calculations add another transformation. Market risk concerns changes in market factors. Counterparty exposure also depends on the legal counterparty, eligible netting and collateral. A security quantity, its carrying amount and its modeled exposure are therefore different outputs.
Take a position of 100 shares and a selected price of 12 dollars per share. The position quantity is 100 shares and its simple marked value is 1,200 dollars. A scenario price of 11 dollars implies a 100-dollar decline from that mark under the stated calculation. The three numbers have different units or meanings; forcing them to agree would destroy the distinctions.
Investment and accounting books of record
An investment book of record supports a defined view of investment activity and positions. An accounting book of record supports specified accounting policies and reporting. Their interfaces need named timestamps, valuation rules and correction authority.
Portfolio construction introduces an earlier state: a target allocation. A target does not become a holding until the required execution and booking steps occur. An optimizer’s output still needs to pass through executable quantities, available cash, fills and account allocations.
Integrated software can carry several of these states. Integration reduces some handoffs but does not remove the need to define which state a field represents.
Reconciliation between purposeful representations
Reconciliation compares records after establishing the mapping between their purposes. It can compare like quantities directly, or explain differences arising from timing, currencies, valuation policies and lifecycle stages.
Take a trade-date position and a settled position during the interval before settlement. Their difference can represent an identified unsettled trade. A useful reconciliation ties the difference to that transaction and its expected state transition. Labeling either number wrong solely because the figures differ omits the timing basis.
A risk result also needs the position population, market-data version and model configuration that generated it. A customer name alone does not establish an eligible netting set.
What stays fixed is the need to name quantity, unit, timestamp, population and valuation or model basis. What changes is the representation required for the financial task.
Questions about position
Should the position and risk report contain the same number?
No. A quantity, a valuation and a response to a scenario have different units and meanings.
Does an explained difference remove the need to reconcile?
No. The explanation must connect the records under their stated timing and policy.
Sources and method
- Investment Accounting SimCorp
- Welcome to CVXPY CVXPY Project
- Aladdin Risk BlackRock
- Principles for effective risk data aggregation and risk reporting Basel Committee on Banking Supervision
- decimal — Decimal fixed-point and floating-point arithmetic Python Software Foundation
- Charles River Trader Charles River Development / State Street
Read next
- Counterparty exposure: from trades to netting sets and collateral
Trade values become counterparty exposure through agreement scope, netting and collateral. Follow the calculation boundaries.
- Decimal arithmetic does not choose a rounding policy
Decimal arithmetic cannot decide when to round. A worked example shows why line-item and aggregate rounding produce different balances.
- Data lineage, quality checks and reconciliation establish different facts
Knowing where data came from does not prove it is complete or correct. Compare lineage, quality checks and reconciliation.
