Most brokerage back offices in India will tell you they’re automated. Ask what happens when a trade doesn’t match the first pass, and the answer often changes: someone pulls a report, opens a spreadsheet, and starts checking manually. That one exception is where the automation quietly ends.
It didn’t matter much. Settlement cycles left enough room to catch up. It matters now.
T+1 is the standard across Indian capital markets, and T+0 is already moving through pilots.
A back office that still treats reconciliation as an end-of-day cleanup task is running on a clock that no longer matches the market it’s serving.
The cost of that mismatch rarely shows up on a balance sheet. It shows up as an analyst headcount that never stops growing, exceptions that surface a day after they should have, and an audit trail that has to be pieced together instead of pulled on demand.
The visible cost is headcount; analysts matching rows a system should already know. The hidden cost compounds faster, especially in institutional and retail broker back-office environments, where trade volumes are higher, and the margin for a day’s delay is thinner.
Delayed exception detection A break found at end-of-day review has already compounded for a full day.
Settlement risk Every hour spent on manual reconciliation raises the odds of an unresolved trade break as settlement windows keep compressing.
Audit exposure A reconciliation trail stitched together from spreadsheets isn’t an audit trail; it’s a reconstruction, assembled under pressure, after the fact.
None of this is a people problem. It’s a systems problem.
Post-trade settlement software gets used loosely, which is why so many brokers assume manual reconciliation isn’t really a gap in their setup. To actually eliminate it, the software has to cover the full journey from execution to settlement:
Covering four of six and patching the rest with spreadsheets isn’t post-trade settlement software; it’s manual reconciliation with better branding.
The distinction that matters isn’t the feature list. It’s whether those features talk to each other in real time.
Reconciliation delays are rarely a process failure on their own; they’re usually a symptom of what the underlying broker back-office software was built on.
Monolithic systems treat trade matching, settlement, corporate actions, and fee calculation as one tightly bound unit. Fixing reconciliation in isolation doesn’t work when the architecture forces every function to move at the pace of the slowest one.
Most “automated” back offices are really just faster spreadsheets; the manual step didn’t disappear; it just got a shorter turnaround
Regulatory compliance in financial markets has always assumed a clean, provable trail from trade to settlement. A break that’s caught and fixed off-system, in a spreadsheet outside the platform of record, leaves a gap in that very audit trail. Here’s the test that matters: if a regulator asked for a full audit trail on a single trade, right now, without notice, could your team produce it directly from the system, or would someone need to rebuild it first?
Those gaps don’t always surface at the trade level, either. Often, they show up downstream, in balance tracking, corporate action processing, and collateral positions at the depository. A trade that reconciled cleanly on the brokerage side can still create a mismatch at the depository if balance updates and custody communication aren’t running on the same real-time footing. Small, unreconciled discrepancies quietly become audit findings weeks later.
Dolphin 2.0 is KGiSL’s post-trade settlement platform for Indian capital markets, built around one premise: reconciliation shouldn’t be a task a system helps you do faster. It should be a step the system removes entirely.
Straight-through processing gets called “automation” a lot, but the more accurate word is elimination; real STP doesn’t make reconciliation faster, it removes the step altogether. Under Dolphin 2.0, trades match against counterparty and exchange data the moment they occur, not at end-of-day. Exceptions are flagged the instant a mismatch happens. Settlement instructions generate themselves the moment a trade is confirmed clean; matching, exception handling, and instruction generation running as one continuous flow instead of a queue of manual handoffs.
Dolphin 2.0 runs matching, exception management, settlement instructions, corporate actions, collateral and margin calculation, and regulatory reporting as a single system; not six tools quietly passing files to each other. That’s what lets a break in one area surface instantly everywhere else it matters.
Dolphin 2.0 generates its audit trail as trades settle, not after someone goes looking for one. The trail exists as a byproduct of the process itself; nothing to reconstruct under pressure before a regulator asks.
The same real-time reconciliation logic runs across the depository side too: balance tracking, corporate action adjustments, and custody communication, catching the last blind spot at the point it occurs, not at the next scheduled review.
| Features | Manual Reconciliation | Straight-Through Processing (STP) |
|---|---|---|
| Matching | End-of-day, by hand or semi-automated scripts | Real-time, as trades occur |
| Exception detection | Hours to a full day later | At the moment of mismatch |
| Settlement instructions | Manually generated after matching | Auto-generated on confirmed match |
| Audit trail | Reconstructed from multiple sources | Built into the process itself |
The difference between these two columns isn’t the hours spent matching. It’s everything a day’s delay lets go unnoticed.
None of this works as five separate upgrades bolted onto an existing system. The real-time matching, the six-component coverage, the architecture underneath, the audit trail, and the depository extension only deliver on the reconciliation-elimination premise because they run as one platform, not five features competing for the same legacy infrastructure. That’s the difference between a system that automates reconciliation and one that was built from the ground up to make it unnecessary; and it’s the design choice Dolphin 2.0 was built around from day one.
Remove manual reconciliation and the back office doesn’t just get faster; its shape changes. Matching, exceptions, settlement, and compliance stop operating as separate checkpoints that catch up to each other at end-of-day, and start operating as one continuous state that’s always current.
The real test of brokerage back-office software isn’t whether it can produce a clean reconciliation report. It’s whether reconciliation, as a distinct manual task, has any reason to exist inside it at all.
For brokers weighing that question, the honest starting point isn’t “how fast is our reconciliation.” It’s “why are we still reconciling”; and it’s the question Dolphin 2.0 was built to answer.
