A FINOV8 Signals series examining the assumptions, technology architectures and operating models shaping institutional trading.
During a global trading-platform transformation at a large global private wealth manager, a senior fixed-income trading executive in Asia asked me to recreate her Excel-based blotter in the new platform.
It had nearly 200 columns. Yes, 200!
But this wasn't simply a trader wanting everything on one screen.
She and her desk had spent years refining that spreadsheet to support trading, post-trade workflows, compliance reporting and record-keeping.
All while battling to get their enterprise-grade systems to do what they actually needed.
The spreadsheet wasn't the problem. It was their solution.
What traders say, and what they actually do
One thing I've learnt over the years: asking traders what they need and watching them do their jobs can produce two very different answers.
You can spend hours discussing requirements, pain points and workflows.
Then spend half a day shadowing a trader and discover more about how the desk actually operates than you did in weeks of requirements workshops.
The manual steps they forgot to mention. The spreadsheets they quietly maintain. The information they copy between systems. The assumptions they've learnt to make and the workarounds they've developed around their OMS, EMS, trading platforms and venues.
All while juggling four, five or six screens across their trading desk, mentally piecing together information that their technology should be helping them interpret.
Not because they're withholding information.
Because, to them, that's simply how trading works.
When workarounds become requirements
And that's where transformation projects can go badly wrong.
We ask traders what they want from a new platform, then risk rebuilding everything they've invented to compensate for the old one.
Of course, the underlying data, controls, records and workflows matter.
But does that mean recreating all 200 columns?
Or should we be asking why those columns, manual interventions and disconnected processes were necessary in the first place?
What if we designed around what traders are trying to achieve, rather than the systems they've learnt to work around?
Understand the decisions. Observe the workflows. Identify the exceptions. Automate what should never have required manual intervention. Preserve authoritative records and controls without forcing traders to manage them through spreadsheets or, in some cases, handwritten notes in a notebook.
Digitising a workaround doesn't make it a good operating model.
Keep the orders. Keep the controls. Keep the records.
But perhaps it's time the 200-column blotter retired with honours.