On Thursday 24 September a colleague at the trading company where I work asked whether he should send his month-end money to India early. Khaleej Times had just carried a statement from Ali Al Najjar, CEO of Al Ansari Exchange, saying the proposed strike “is not expected to affect our remittance services or cause delays to customer transfers.” The next day The National put the scale on it: the UAE sends 19.2 per cent of India’s inward remittances, and 76.1 per cent of UAE transfers went through digital channels in 2024.

Then, at 9.30 pm on Sunday 27 September, the United Forum of Bank Unions met the Indian Banks’ Association and deferred the strike, hours before it was due to start. So nothing happened. The question is still worth answering, because the same Khaleej Times piece reported the unions were threatening an indefinite strike from 26 October, and because the answer is the clearest map I know of how a remittance actually moves.

What an Indian bank strike stops, and what it does not

The Indian government described the stakes bluntly. Khaleej Times quoted the statement: “The strike, together with the preceding weekend, could effectively result in a five-day closure of banking services, affecting customers, businesses, government transactions and international banking operations.” Five days is the branch network. It is not the payment rails.

State Bank of India said as much on 25 September, per Business Today: UPI, NEFT, RTGS, IMPS, internet banking, mobile banking and ATMs would keep running through the strike, and branches would open on Sunday the 27th ahead of it. NEFT has settled in 48 half-hourly batches between 00:30 and 00:00 every day, holiday or not, since 16 December 2019. A batch does not need a teller.

Vijay Valecha of Century Financial gave The National the sentence that matters for anyone sending money from the UAE: transfers to India “are unlikely to be affected, as these systems run on computerised back-end systems that require no manual processing by branch staff.” His warning was for the other direction of exposure: an NRI expecting a cheque “should not assume that depositing the cheque immediately before the strike will ensure normal settlement.”

That is the whole split. Electronic credits run on infrastructure. Cheques, counter cash and passbooks run on people.

The three legs of a transfer from the UAE to India

At the trading company, treasury reads every outgoing wire as three timestamps: the debit date, the value date, and the beneficiary’s confirmation. A supplier in India has not been paid when the dirhams leave the account; they have been paid when their bank says so. The same three moments exist inside a AED 5,000 family transfer. The exchange-house receipt prints only the first one.

Leg one is the exchange house. Al Ansari’s Key Facts Statement for remittances prices an India bank-account transfer at a service charge of “Minimum AED 0.00 - Maximum AED 25.24”, with beneficiary back-end charges listed as “NIL”. The digital share is smaller than the industry figure suggests: Al Najjar told Khaleej Times “Around 30 per cent of transactions are completed digitally”, the rest across 290 branches. A UAE strike would touch this leg. An Indian one cannot.

Leg two is the rail. The same statement gives the delivery time as “Instant, Standard (1-2 Working days)”. Instant is an immediate-credit rail like IMPS. Standard is a batch, which is NEFT territory. Neither involves a branch employee in India, which is why Al Najjar could promise continuity on a strike his company had no control over.

Leg three is the receiving branch. This is where cheques clear, where cash is withdrawn over a counter, where a passbook gets printed. It is the only leg with a person in it, and the only leg a strike can stop. The person exposed last week was not the one sending AED 5,000 through an app. It was the one being paid by cheque in India, or paying the exchange house by cheque here: the Key Facts Statement says a cheque-funded transaction “will be valid and released subject to cheque realization”.

The dirhams leave on the debit date. The rupees arrive on the rail’s timetable. Only the third leg has a person in it, and that is the only leg a strike can stop.

What the receipt commits the sender to if a transfer does come back

The receipt looks like a proof of payment. Under the terms it is a proof of instruction. Three lines in Al Ansari’s statement decide what a delay costs if it turns into a return.

The first is the claim window. Transfers “are considered executed and delivered unless a claim is received within 3 months for cash payout and 6 months for bank transfer”. Separately, “Any transaction-related complaint should be lodged within a maximum period of fourteen (14) days from the date of the transaction.” Fourteen days is short when the beneficiary is not sure whether the credit landed.

The second is the refund rate. A cancelled or returned transfer “will be refunded to the sender at the prevailing market buying rate or transaction rate whichever is lower, excluding the sending charges, any deductions from the beneficiary bank, cancellation fees and any other expense incurred”. So a AED 5,000 transfer that bounces does not come back as AED 5,000. The fee stays with the exchange house, and the rupees are converted back at whichever of the two rates gives fewer dirhams.

The third is the trigger. A refund needs “confirmation of non-payment and cancellation of the original instruction” from the correspondent bank or agent before anything is paid. If the correspondent’s staff are the ones on strike, that confirmation is the part that waits. The dirhams sit inside the system in the meantime, belonging to nobody the sender can call.

None of this is a reason to avoid exchange houses. It is a reason to read the receipt as a treasury clerk would: the debit is done, the value date is the rail’s, and the confirmation is the beneficiary’s job. Legs one and two were never at risk last week. Leg three was, and it would be again on 26 October.

For my colleague the answer was the instant option, sent before the weekend, with the beneficiary’s screenshot of the credit as the close of the transaction, not the receipt. In my own tracker, money sent home leaves on the debit date and the line stays flagged until that screenshot exists. The strike was called off. The habit is worth keeping.