How to Read a PSP Settlement Statement: iGaming Operator Field Guide

A monthly PSP settlement statement runs 6 to 14 pages on a mid-volume iGaming book and answers four questions in order: how much card volume cleared, how much the PSP and the card networks took out, how much the acquirer withheld against future chargebacks, and how much arrived in the operator's bank. The four numbers reconcile to a single accounting identity, and they never reconcile cleanly the first time the controller checks them. The gap is where unannounced markup, discretionary reserve adjustments, and FX-timing mismatches accumulate. This guide reads the statement the way a payments-side operator reads it before sending the bank-side reconciliation team chasing $40,000 ghosts: section by section, line by line, with the specific field names that ship in production at Adyen, Rapyd, Worldpay, and Nuvei in May 2026.

The accounting identity a settlement statement has to satisfy

The reconciliation identity for any PSP statement reads the same. Gross authorized volume, minus refunds and chargebacks, minus interchange and scheme assessment fees, minus the acquirer's contracted basis-point markup and per-transaction cents, minus the other fees the contract permits (FANF, NABU, MAC, statement, gateway, PCI, monthly minimums), minus the reserve adjustment for the period, plus or minus any FX gain or loss on cross-currency netting, equals the net deposit landing in the operator's bank. Eight inputs to a single output.

The operator who treats the statement as one number, net deposit, and reconciles it against one number, the bank credit, sees only the residual. That residual carries no information about which input drifted between months. Chasing a $40,000 monthly residual on the residual alone is what gets a finance team into a six-week conversation with the PSP about a problem they cannot describe precisely.

The discipline is the inverse: compute each input independently from primary sources (interchange tables, scheme assessment rates, the contract's bps-and-cents schedule), then carry forward to net. When the recomputed net matches the statement's reported net, the statement is internally consistent. When it does not, the gap sits in a specific input the reader can name in the next call.

Statement anatomy: where each line lives in Adyen, Rapyd, Worldpay, Nuvei

PSPs use different names for the same fields. The reader who only knows one PSP's nomenclature reads a second PSP's statement as if it were missing data. The cross-walk for the four PSPs that route most MCC 7995 volume in May 2026:

Per Adyen's settlement details documentation and Rapyd's merchant settlement report reference, the underlying data is the same across PSPs even when the labeling diverges. The reader who builds a normalized internal mapping from PSP-specific labels to the eight identity inputs audits any PSP's statement in the same workflow.

The fields that hide markup most often are the labels with the most PSP-side flexibility: "Markup", "Discount Rate", "Network Access", "Processor Risk Fee". Per Mecca Payments' merchant audit guide and the Merchant Cost Consulting "Other Fees" analysis, six line items can total 81 percent of an "Other Fees" subsection while occupying only 6 of 33 line items. The high-impact lines are not the long ones; they are the ones with ambiguous labels.

Fee creep at month six: when interchange pass-through widens by 30 to 50 bps

Operators sign for "interchange-plus" or "interchange pass-through" pricing and assume the math that produced the contracted rate at signing keeps producing it at month six. It does not. Three independent mechanics widen the effective rate over the first 90 to 180 days, and none requires the PSP to send a formal contract amendment.

The first is rate-tier mix shift on the interchange line itself. Every transaction is assigned an interchange category by the network at clearing. Per Merchant Insiders' rate reference, Visa interchange ranges from 0.05 percent on regulated debit to 2.70 percent on premium consumer credit. Operator volume that skewed to debit during onboarding skews to credit and rewards in steady state, and the interchange line picks up double-digit basis points without any change to the contracted markup.

The second is scheme-fee creep at the network level. Visa and Mastercard publish updated assessment schedules twice a year, and each cycle adds incremental fees the PSP is contractually entitled to pass through under the "interchange-plus" structure. The Visa Fixed Acquirer Network Fee (FANF) bills quarterly rather than monthly per the Merchant Cost Consulting FANF reference, which means the line is invisible on two of every three monthly statements and surprises the operator on the third. Mastercard's TPE program expanded the MAC 03 and MAC 21 fee from retries to original declines starting January 2026 per the Merchant Cost Consulting TPE tracker, which on an MCC 7995 book at 13 percent CNP decline rate adds $0.03 per CNP decline that did not previously carry a fee at all.

The third is line-item insertion under contractual catch-all clauses. "Network Access Fee", "Compliance Fee", "PCI Service Fee", and "Regulatory Fee" are the four common entries that surface in month two or month three of a contract, frame as pass-through, and carry processor margin. Per Mecca Payments' audit guide, the typical PCI non-compliance fee runs $20 to $50 per month even on PCI-compliant merchants who failed to return a self-assessment questionnaire on time.

A representative six-month creep pattern on a mid-volume MCC 7995 book that signed at "interchange + 0.95 percent": interchange line drifts up 25 bps as portfolio mix rebalances toward credit, scheme fees pick up 10 bps as FANF surfaces quarterly and the MAC expansion takes effect, and a "Network Access" line at the second-quarter mark adds 5 bps of pass-through margin. The aggregate effective rate at month six runs 40 bps above the headline. On a $5M monthly card book that is $20,000 a month the operator did not budget.

The audit catches it month-over-month, not as a single big number. Effective rate is total fees divided by gross volume, computed once per month and tracked across three to six months minimum. A 5 to 8 bps drift on consecutive months is the signal; a one-time spike usually traces to a portfolio mix change in the underlying volume rather than a markup change.

Rolling reserve vs ReserveAdjustment: same dollars, different obligation

Rolling reserves are the contracted withholding mechanism. The merchant agreement specifies the percentage (typically 5 to 10 percent on MCC 7995 books per Merchant Maverick's reserve reference and Tailored Pay's 2026 guide), the holding period (90 to 180 days), and the release schedule (typically rolling daily). A 10 percent reserve at 180 days on a $5M monthly book sees $500,000 enter the reserve over the first month, the steady-state balance climb to roughly $3M over the holding period, and the first day of the seventh month produce the first dollar release.

That mechanism is auditable because the formula is in the contract. The reserve line reads predictably month over month, and any deviation from the predicted reserve balance points to a single non-contractual action.

The non-contractual action sits on a different line. Adyen's settlement details report records it under the "ReserveAdjustment" journal type per the Adyen reference. Rapyd categorizes equivalent activity under "reserve_hold" entries that fall outside the rolling-reserve schedule. Worldpay and Nuvei use "Risk Hold" or "Exception Hold" on the merchant statement face. The naming matters less than the mechanic: each is a withholding the PSP applies under contractual catch-all clauses that allow discretionary increases on observed risk events.

Per TFM Law's analysis of withheld reserves and GambleHub's PSP reserve guide, the events that trigger discretionary holds in May 2026 are documented on the contract side: a single-month chargeback ratio above the acquirer-internal threshold (typically the 0.5 to 0.7 percent acquirer-level ceiling rather than the published 1.5 percent merchant cutoff under VAMP), a sudden volume spike the underwriter classifies as outside the projected book size, an inbound regulator inquiry, or a single high-ticket dispute the PSP wants to fund. None of those events requires the operator to be notified before the hold appears on the next statement.

The diagnostic that distinguishes a contractual rolling reserve from a discretionary hold is the math, not the label. The reserve schedule produces a daily contribution equal to the contracted percentage of the day's settled volume; any deposit into the reserve account that exceeds that formula on the day is a discretionary action. The operator who computes the contractually predicted reserve daily and compares it to the statement's reported reserve balance catches the discretionary holds on the day they appear, not 90 days later when working capital tightens unexpectedly.

The contractual recourse on a discretionary hold sits in the merchant agreement. Before the call to the relationship manager, the operator should know which clause the PSP cited (the "additional reserve" or "reserve increase" clause in most MGAs), what the trigger event was on the operator's side, and what the documented release condition is. Discretionary holds without a documented trigger or release condition are negotiable; discretionary holds with a clean trigger record (a chargeback spike, a regulator letter) typically are not.

PSP-net minus bank-deposit: four root causes for the gap

The PSP says it sent $4,892,310 net for the month. The bank credits $4,847,205 across the same period. The $45,105 gap is the operator's monthly mystery. Four mechanics produce that gap, and each resolves to a specific reconciliation step. The webhook-handler architecture that ingests the events upstream of this gap is the companion read for the operator wiring both layers from scratch.

FX timing on cross-currency settlements. A book that takes EUR card volume but settles to a USD bank account converts at a rate the PSP picks at settlement time. The transaction-level data the operator stored at authorization carried a different EUR/USD rate. Per Aurum Solutions' multi-currency reconciliation analysis, the spread between the two rates is material on cross-currency volume during normal market days and widens during volatility. The resolution is to log both rates (Adyen exposes them in the "Exchange Rate" column and the corresponding "Net Currency" column), book the spread as FX gain or loss separately from reconciliation breaks, and stop counting it as a reconciliation gap.

Intermediate fees that bypass the gross fee subtotal. Some PSP-internal fees (FX conversion margin, instant-settlement uplift, expedited-payout charges) book below the "Total Fees" line on the merchant face but above the "Net Deposit" line in the journaled data. The merchant who reads only the visible total fees and adds it to the gross volume cannot match the bank deposit because the journaled deductions land somewhere the eye does not. The fix is to pull the transaction-level export rather than the PDF summary and sum every debit against the merchant account, not only the lines appearing in the "Fees" section.

Batch-cut timing across the month boundary. The PSP's settlement batch closes on a specific UTC time, not midnight in the operator's timezone. Authorizations on the 31st that settle in the early hours of the 1st belong to the next month's report. Per Webgility's settlement-discrepancy reference, this is a recurring cause of monthly mismatch on books that grew faster than their finance process. The diagnostic is to anchor reconciliation to the PSP's batch-close UTC time rather than the operator's calendar month, and to expect the monthly report to lag the calendar month by hours to a day on the close date.

Reserve increment that does not match the contracted formula. The structural case sits in the prior section. The reconciliation case is mechanical: compute the contracted daily reserve contribution, compare to the statement's reserve line, and flag any positive deposit into the reserve account beyond the formula as a discretionary hold. The "missing" cash is in the reserve account, the operator owns it on a future date, and the conversation with the PSP is about the trigger and the release condition, not about missing money.

The four-step debug walk that decomposes the gap:

1. Reconcile the PSP's transaction-level export to gross volume. Pull the row-level data, sum the gross authorization amounts in the merchant-account currency, compare to the statement's gross volume. A mismatch here is rare and points to PSP-side reporting error.
2. Reconcile the PSP's net-of-fees subtotal to the recomputed identity. Sum interchange, scheme fees, contracted markup, and other-fees lines using the contract's bps-and-cents schedule. A mismatch here is fee creep, and the gap names which line drifted.
3. Reconcile the post-reserve subtotal to the contracted reserve schedule. Compute the contracted daily reserve contribution and release for the period, apply to the prior-month reserve balance. A mismatch here is a discretionary hold.
4. Reconcile the PSP-reported net to the bank deposit. A mismatch at this final stage, after the prior three reconcile, is FX timing, batch-cut timing, or intermediate fees the merchant face hides. The transaction-level export resolves the first two; the contract's fee schedule resolves the third.

The gap that the operator initially saw as a single number is now four named gaps, each with a specific root cause and a specific contractual recourse.

The audit bundle that gets the PSP relationship manager to engage

The relationship-manager call goes one of two ways. The operator who walks in with "the deposits are wrong, can you check them" loses the conversation in the first ten minutes. The operator who walks in with the four-step decomposition and a one-page audit bundle gets a different call.

The bundle that lands well at the PSP:

• The reconciliation worksheet for the month, four columns: identity input, contracted formula, statement-reported value, recomputed value. Each row points to a specific contract clause or a specific scheme-fee table.
• The effective-rate trend across three to six months, one row per month, computed as total fees over gross volume. The drift is visible at three or four months and quantifiable at six.
• The reserve balance projection from the contracted formula, day by day for the month, against the statement-reported reserve. Discrepancies above the contracted daily contribution surface on this view.
• The list of new line items that appeared in the current month and were absent in the prior 90 days. Each is matched against a contract clause or flagged as unilateral.
• The PSP-net minus bank-deposit gap broken down into FX timing, intermediate fees, batch-cut timing, and reserve increment. Each component carries a number and a source.

The conversation that follows is no longer about whether the operator is reading the statement correctly. It is about which contract clause covers which line item, which discretionary action is being negotiated, and which fees are pass-through versus markup. PSPs that hold the relationship-manager call accountable to specific contract language and specific scheme-fee tables move on the negotiable lines. The same PSPs route the operator who cannot anchor the conversation to specific lines into a six-week loop of "we'll look into it" with no resolution.

The audit reads the same statement every month. The investment is the build, not the recurring cost: a normalized internal mapping from PSP-specific labels to the eight identity inputs, a script that pulls the transaction-level export and recomputes the identity, and a standing dashboard that tracks effective rate, reserve balance, and PSP-net minus bank-deposit gap month over month. With that build in place, the next anomaly surfaces on the day it happens, not on the day a contract-renewal cycle forces the finance team to reconstruct six months of statements under time pressure.