Automating performance fee calculations for accuracy and efficiency

Automating performance-fee calculations cut an asset manager's calculation time by 30% and reduced discrepancies and calculation errors by 15%, while an improved forecasting suite gave the business a clearer, more current view of future revenue. What used to depend on third-party vendor outputs, internal checks and manual Excel reconciliation now runs as a single, self-checking engine.

The problem

Before the engagement, the asset manager's performance-fee calculations relied on a mix of third-party vendor outputs, internal verification and manual Excel work. Every calculation cycle meant collating figures from more than one source, checking them against each other, and resolving mismatches by hand.

Performance-fee calculations of this kind are rarely simple: mechanics such as hurdle rates or high-water marks are a common source of complexity in the industry generally. The source material describes the manager's own arrangements only as carrying "complex fund nuances", without specifying which mechanics were involved, but it was this kind of complexity that made reconciliation genuinely time- and resource-intensive.

Because fee figures only became final after manual reconciliation, the finance team could not forecast future revenue with confidence. The inputs to any forecast were themselves provisional until reconciled, so revenue planning lagged behind the reality of the business and undermined the accuracy of future forecasting.

What QuantSpark built

The approach centred on replacing the manual reconciliation loop with one automated engine that could ingest, calculate, check and report in a single pass, built in seven stages:

1. Mapped the existing process end to end, tracing how figures moved between vendors, internal checks and spreadsheets, to identify exactly where time and accuracy were being lost.
2. Automated data ingestion, so fee-relevant data could be pulled in from multiple source systems without manual re-keying.
3. Encoded each fund's fee rules into a single calculation engine, so fees are computed consistently across the fund range rather than fund-by-fund in Excel.
4. Added automatic cross-checking of the engine's output against the third-party vendor's own figures, flagging discrepancies as they arise rather than at the end of the cycle.
5. Generated interim calculations partway through the cycle, giving the reconciliation team a running check rather than a single end-of-cycle result.
6. Produced standardised, comprehensive per-fund reports directly from the engine, removing a separate manual reporting step.
7. Used the same clean, reconciled data to strengthen the existing forecasting suite, so the business could project fee revenue forward rather than only report it backward.

How it flows

In practice the new process runs as four stages. Ingestion pulls fee-relevant figures automatically from the manager's multiple source systems. A calculation engine applies each fund's rules to produce a fee figure. A reconciliation step checks that figure against the third-party vendor's own calculation, surfacing interim results and flagging discrepancies before they reach the end of the cycle. Finally, reporting and forecasting take the same reconciled data and turn it into both a per-fund report and a forward revenue forecast.

The systems underneath

The engagement is best understood as four functional layers rather than named products: a data-ingestion pipeline pulling from multiple source systems; a rules-based fee-calculation engine encoding fund-specific mechanics; a reconciliation layer that checks the engine's output against third-party figures and flags discrepancies; and a reporting and forecasting layer turning the same reconciled data into per-fund reports and forward revenue projections. QuantSpark did not need to replace the manager's existing vendor relationship; it added an internal capability that made oversight of that relationship faster and more accurate.

What it delivered

The measured results are calculation time down 30% and discrepancies and calculation errors down 15%, both against the prior manual process. The improvement to the forecasting suite is a qualitative addition on top: because fee outputs are now reconciled and reported automatically rather than manually, the finance team has a more current, more trustworthy input for revenue forecasting and can run a more advanced revenue strategy as a result.

The natural way to think about return on an engagement like this is time saved multiplied by the cost of the people-hours previously spent on manual reconciliation, multiplied by how often that reconciliation cycle runs. The source material does not give the salary, headcount or cycle-frequency figures needed to turn that into a specific pound value, so this case study reports the two measured percentages as delivered and leaves any monetary translation to be modelled against a manager's own figures.