How a Scrip Desk Actually Runs Today

Brokers get written about badly by people building software to replace them. That is not this page. A scrip desk exists because there is no exchange, no published price, and no way for a stranger to check a stranger, and it solves all three with the only instrument available, which is judgement. This is an account of that work, the five jobs it quietly bundles together, and the one of them that no amount of skill can make safe.

The state built a rail and stopped there

This is the fact everything else follows from, and the government has been unusually direct about it. When DGFT re-operationalised the scrip transfer recording module for the older schemes, the trade notice said the owner transfers the scrip as per the negotiated terms and conditions between the buyer and seller, and added that DGFT and Customs are not responsible for any lapse by the old or new owner, or any dispute between them.

Read plainly, that is the State providing issuance and a transfer rail and formally declining the commercial layer. Someone has to find the counterparty. Someone has to judge whether they are good for the money. Someone has to arrive at a price. Someone has to stand between the scrip moving and the payment landing. For two decades that someone has been a broker, and nothing in the design of the schemes suggests anybody planned it that way.

The five jobs inside one seat

It is worth separating them, because they are usually discussed as one thing and they have completely different risk profiles.

1. Sourcing and matching. Knowing which exporter has credit sitting idle and which importer at which port has duty to pay this month. This is genuine market knowledge and it is not written down anywhere. One desk that publishes its method says it watches daily demand at the major ports, which is exactly the right input and also exactly the kind of thing that cannot be looked up.
2. Credit judgement. Deciding whether a counterparty will actually pay, or actually deliver. In a market with no clearing house, this is the substitute for one.
3. Pricing. Arriving at a number with no benchmark behind it. DGFT publishes the scheme rates that decide how much credit an export earns; nobody publishes what the credit then sells for.
4. Settlement. Standing in the gap between the transfer and the money. This is the job that is invisible from the outside and the one that ends desks.
5. Paperwork. The tax invoice, the transfer letter, the courier, and the reconstruction of all of it when someone asks a year later.

Four of those are skill. The fifth is a balance sheet risk being carried by a business that does not have a balance sheet designed for it.

Why a desk cannot get leverage on its own book

The operational ceiling here is not effort, it is the design of the rail, and it is worth being precise about it because it explains why a desk with forty clients does not feel four times easier than a desk with ten.

A transfer is initiated from inside the seller's own ICEGATE e-scrip account. A one-time password goes to the seller's own registered email and mobile, and it is valid for fifteen minutes. The scrip then sits at Transfer Pending until the buyer, who must already hold an e-scrip account of their own, approves it with a one-time password of their own. Two people, two phones, one narrow window, per trade.

Nothing about that batches. There is no bulk mode and no delegated mode inside the portal. A desk running many client IECs is therefore running many separate manual sequences, each of which stalls if one person is on a flight, or in a factory, or simply not looking at their phone. The cost per trade does not fall as the book grows, which is the definition of a business that cannot scale on its own effort.

Then there is the shape of the inventory. Under Regulation 7(2) of the Electronic Duty Credit Ledger Regulations, duty credit transfers at a time for the entire amount in the e-scrip, and transfer in part is not permitted. Lot sizes are whatever the exporter happened to generate. A desk cannot split a large scrip to fit a small buyer or top up a small one to fit a large buyer. Desks handle it by pooling small credits into larger lots, which works, and which also means small scrips wait, and waiting is expensive when the instrument has two years of life and no way to get it back.

Pricing without a reference

Every desk sets a number. Almost none of them shows its working, and in fairness there is nothing to show it against.

What is publicly visible is narrow. One desk publishes a live two-way board on its own site, quoting a buying and a selling price for RoDTEP and RoSCTL as percentages of face value, roughly one and a half to two points apart. That is the clearest single window into intermediary economics anyone in this market has opened, and it is one venue.

Set against it, an independent trade policy research group described large importers buying scrips from small exporters at discounts of up to ten per cent of scrip value. Both numbers can be true, of different trades, in the same market, on the same day. That gap is not evidence of anyone behaving badly. It is what price dispersion looks like when there is no reference rate, and it is felt hardest by the smallest seller, who has the least idea what their credit is worth and the least ability to walk away.

For a desk, that dispersion cuts both ways. It is where the margin lives, and it is also why every client conversation starts from scratch and why a good quote and a poor one look identical to the person receiving them. How brokers price scrips goes further into the mechanics.

Where the risk sits, and why it sits on you

This is the section written for the person who actually runs a desk, because it is the part clients never see.

You are the one who checked. There is no public lookup on ICEGATE that lets a non-holder verify a scrip number, its balance, its expiry, or whether it has already been transferred. The authoritative view of a scrip appears only once it is in a ledger you control. So everything done before that point, reviewing a copy of the scrip, reading the shipping bill chain, knowing the exporter, is judgement standing in for a check that does not exist. When it is right, nobody notices. When it is wrong, you are the one who said it was fine.

You are exposed on both legs at once. In the version of this trade where funds move through the desk's own account, the desk owes the seller for a scrip it has not yet been paid for, and owes the buyer a scrip it has not yet received. Those two exposures are live simultaneously, against different parties, for a margin measured in tenths of a percent.

The tail is longer than the trade. A settled trade is not a closed one. Provenance questions surface when someone else's export chain is examined, months later, and the first call goes to the name on the invoice.

And the thing you are checking for is narrower than it looks. Indian case law draws a hard line between two kinds of bad scrip. An instrument that was never validly issued, a forgery, leaves whoever used it liable for the duty, and the Supreme Court has held that good faith goes to the penalty rather than to the demand. An instrument that was validly issued but obtained by fraud has been treated differently: voidable rather than void, good until the issuing authority cancels it, with buyers who used it while it was subsisting held not liable. That distinction is what a desk's diligence is really aimed at, whether or not anyone has ever phrased it that way in a client call. And it is not answerable from a transfer letter.

It is worth knowing what the enforcement record actually contains, too, because it is not what most people assume. The documented cases in this market are fabricated exports and inflated invoices producing real scrips, and, more recently, hijacked digital signatures used to take over exporter accounts and move credit out of them. Those are origin failures, and they are exactly the failures a desk is trusted to catch with the tools it has, which are documents and acquaintance.

The software layer has been thin, and it is worth saying why

Products have been launched at this market. Looking at them honestly is more useful than pretending we are first.

Of the online scrip venues we could find, one is a live app with a small user base that solicits competing quotes from agents, which is brokers on tap rather than an order book. One publishes live rates and carried a notice, dated July 2026, that it had placed pending RoDTEP purchases on standby over portal synchronisation problems. Two more are gone: one domain no longer resolves, another serves a hosting suspension page. Several intermediaries describe holding funds in the middle, and we could not find one that publishes the terms of that arrangement or names who holds the money.

The reason is not that the people building them were careless. It is that the hard part of this market is not a listing screen. It is the settlement leg, and the settlement leg touches a government rail that was never designed to be driven by a third party. That is the wall everyone hits, and it is the wall a desk has been climbing manually, client by client, for years.

What we could not verify, and are not going to invent

• What a desk uses internally. No intermediary publishes how it tracks client positions, matches lots or reconciles commission, so we will not characterise it. It would be easy to write the word spreadsheet here. We have no evidence for it.
• Broker rate-broadcast groups. Desks publish WhatsApp numbers as a customer channel, which is verifiable. We found no public evidence of the rate-broadcast groups that get described second hand, and closed groups would not leave any, so we make no claim.
• Commission levels. Not one intermediary we found publishes a percentage. Any figure we quoted would be invented.
• Whether a transferee can transfer again. Regulation 7 is silent, and the ICEGATE advisory does not address it. We do not know, and neither does anyone writing confidently about it.
• How often desks actually get burned. We searched Indian judgments, tribunal orders and trade press for a reported case of a scrip trade failing on payment, in either direction, and found none. Every documented failure in this market is revenue-side, meaning fraudulent or forged instruments, or theft-side, meaning hijacked digital signatures and stolen scrips. Nothing on counterparty payment.

We are not going to spin that. An empty litigation record is the single strongest thing anyone can say in this trade's favour: on the available evidence, desks have largely been paying. The argument for rails is not that the market is full of defaulters. It is that a risk which has not yet cost you is still a risk you are carrying personally, priced at nothing, on every trade.

What changes with ScripX

The premise is that the four jobs a desk is good at should stay with the desk, and the fifth should stop being a personal risk. Outcomes only, because the outcomes are the part that belongs to you.

• You stop standing in the settlement gap. Settlement is delivery versus payment. The buyer's funds are locked before the scrip moves, and the seller's payout is released against confirmation of the transfer, the same business day, with a UTR. A settlement that fails unwinds and refunds in full the same day. The exposure that used to sit on your account stops being yours to carry.
• You keep the client and the commission. ScripX runs on an agency model and never takes title to a scrip, so your arrangement with your client is untouched. What you have earned is read from the same record the money posted to, so the figure on your desk and the figure in the bank cannot drift apart.
• The whole book is one book. Every client IEC in one place, each trading as itself under its own signed, revocable mandate, with ledgers, trades and audit packs kept separate client by client. Sell side and buy side on the same desk.
• Your quote has something behind it. Offers are benchmarked to a firm price for the specific scrip, held while you decide, by scheme and expiry band, and a firm offer holds for sixty seconds. The client stops asking you to justify a number that had nothing public to stand on.
• Provenance is screened before a scrip lists, so the judgement you were making alone has a documented check underneath it rather than resting entirely on your name.
• It can run inside your own portal. Firm quotes, orders and a per-trade audit pack over REST, so your clients see your product and you see the rails. The desk and the API covers the seat in detail.

That is the bridge. Not a replacement for the person who knows which importer at Nhava Sheva has duty to pay in March, but a floor underneath them, so the knowledge earns and the risk does not have to be personal.

Common questions

Where this came from

The transfer sequence and the whole-scrip rule are from the ICEGATE e-scrip advisory and the Electronic Duty Credit Ledger Regulations, 2021 as amended. The position that Customs and DGFT take no responsibility for the commercial terms is from DGFT's own trade notice on the scrip transfer recording module. Descriptions of intermediary practice come from what intermediaries publish about themselves, and the ten-per-cent discount figure is from independent trade policy research published in January 2024. What we could not source is named above rather than smoothed over.