A client funded an account in February. In May the acquirer sends a notification: the transaction has been charged back, the money has already been taken out of the merchant's settlement, and there are days rather than weeks to respond. Nobody asked the merchant's permission. That asymmetry is the design of the system, and every operational decision around card payments follows from it.
The chain, and who moves money at each step
Five parties are involved. The cardholder, the issuer that gave them the card, the card scheme, the acquirer that holds the merchant account, and the merchant. A dispute starts when the cardholder contacts the issuer, not the merchant. Many never contact the merchant at all, because calling the bank is easier than finding a support email.
The issuer assigns a reason code and raises the chargeback through the scheme. The acquirer debits the merchant, typically for the transaction amount plus a fixed dispute fee, and this happens before any argument is heard. The merchant then has a defined window to submit representment, meaning evidence that the transaction was valid. The issuer reviews it and either accepts the defence or escalates to pre-arbitration, after which the scheme itself can arbitrate for a fee that makes fighting small transactions economically pointless.
The reason code decides the whole case
Reason codes are not labels, they are the rules of that specific argument. A fraud code asks whether the genuine cardholder authorised the transaction. A service code asks whether what was paid for was delivered as described. A processing code asks whether the transaction was handled correctly, for example a duplicate charge or a wrong currency.
The evidence that wins one loses another. Proving delivery is irrelevant to a fraud claim, where the question is authentication. Proving the card was authenticated is irrelevant to a cancelled-subscription claim, where the question is what the terms said and when the cancellation was requested. The first thing to read on any dispute notification is the code, and the second is the exact evidence list the scheme publishes for it.
Representment is graded on documents, not intent. A dispute you should win on the facts will be lost if the file is incomplete, late, or answers a different question from the one the reason code asked.
What goes in a file that wins
For a trading firm the strongest defence is a single continuous record tying one verified person to one payment and one usage trail. That means the identity documents collected at onboarding and when they were verified, the account holder's name matched against the cardholder name, the IP address and device at sign-up and at deposit, the timestamped acceptance of terms, the deposit confirmation shown on screen and emailed, the authentication result including any 3D Secure data, and the activity that followed the deposit.
The last item matters more than people expect. Trades placed, withdrawals requested and support conversations after the payment are all evidence that the account holder was present and knew what they had funded. Firms that keep this material scattered across three systems assemble it slowly and submit late, which is the most common way a defensible dispute is lost. Keeping verification records in the same place as payment records is an unglamorous investment that pays for itself the first time a batch of disputes arrives.
Ratio programmes make a bad month expensive
Individual chargebacks cost the transaction and a fee. The real damage sits at the portfolio level. The card schemes run monitoring programmes, Visa's dispute monitoring programme and Mastercard's excessive chargeback programme among them, which track a merchant's chargeback count and ratio against thresholds. Cross them and the merchant enters a remediation programme with escalating monthly costs and reporting obligations.
Acquirers move first, usually by raising the rolling reserve, tightening settlement, or exiting the account. For a firm in a category acquirers already classify as higher risk, losing a processing relationship is harder to replace than the disputed money, which is the practical reason high-risk merchant accounts price the way they do. Two months of carelessness can undo a year of building banking relationships.
The dispute that is really about a loss
The pattern every broker and prop firm recognises: a client deposits, trades, loses, and files a dispute. Nobody writes "I lost money" on the form, because that is not a valid reason. It arrives as unauthorised use, or as a service not provided, or as a claim that the client did not understand what they were buying.
Some of these are genuine, involving a stolen card or a family member. Many are what the industry calls friendly fraud, and the honest position is that a firm cannot always tell which is which. What it can do is remove the ambiguity in advance: verify before the first deposit rather than before the first withdrawal, refuse third-party cards, keep the descriptor on the statement recognisable so a client does not dispute a name they do not recognise, and answer support messages fast enough that the bank is not the easier option. Our existing note on chargeback and fraud prevention goes further into the operational side.
Prevention beats representment, every time
Winning a dispute returns the money and still counts toward the ratio in most programmes, so the win is partial. Prevention is the only move that helps both numbers.
Authenticate transactions where the rules allow it, since 3D Secure shifts liability on the fraud codes even though it does nothing for service disputes. Publish refund terms plainly and honour them quickly, because a refund is cheaper than a chargeback in every dimension. Use scheme alert services if the volume justifies them, refunding a flagged transaction before it becomes a formal dispute. Track disputes by acquisition source, since a single affiliate or a single geography is often responsible for a disproportionate share, and that is a marketing decision rather than a payments one. Firms that watch this alongside approval rates tend to catch the problem in the month it starts.
"Everyone wants to argue disputes better. The firms with a clean ratio are not better arguers. They verified the client before the first deposit and made refunds easy, so the bank never got the call."
— Roman Onta, Executive Director, SINGUARD
Key Takeaways
- The acquirer debits the merchant before any argument is heard, so cash leaves first and the defence comes second.
- The reason code defines which evidence counts; a strong file answering the wrong question still loses.
- Scheme monitoring programmes punish the ratio, and acquirers react by raising reserves or ending the relationship.
- Verification before the first deposit, a recognisable statement descriptor and fast refunds prevent more disputes than any representment template.
Frequently Asked Questions
How long does a cardholder have to raise a chargeback?
Card scheme rules give windows measured in months rather than days, and the exact limit depends on the reason code and the scheme. For a merchant that means a payment is never truly final at the moment of authorisation, and funds should be treated as reversible for a long period after the client has spent them.
Does 3D Secure stop chargebacks?
It shifts liability for one category. When a transaction is authenticated under 3D Secure, fraud-related disputes generally move to the issuer rather than the merchant. Disputes raised under service reasons, such as a client claiming a service was never provided or that a subscription was cancelled, are unaffected and still land on the merchant.
Can a client charge back a trading deposit because they lost money?
They can raise the dispute, usually filed under an unauthorised or service-not-received reason rather than as a complaint about losses. The merchant's defence is a documented record showing the same verified person opened the account, passed verification, made the deposit and used the funds. Without that record, the dispute is difficult to defend even when the transaction was genuine.