Financial firms have to deal with an enormous amount of formats when it comes to data. Some of these are highly standardised, like SWIFT messages, but many aren’t.

Counterparties can all have different ways of sharing and presenting data, from rich-text files to Excel spreadsheets, and even the naming conventions differ from firm to firm.

This is especially true of complex financial products like derivatives, where there really is no limit to the variety or breadth of information that may be included.

And we haven’t even touched upon the fact that 80% of enterprise data is unstructured; e.g. it lives in PDFs, emails, faxes, images and so on.

While the types of data we’ve explained in the previous paragraph could, with infinite resources, all be automated, you can’t get unstructured data into traditional automation technology without an army of staff manually extracting the data and keying it into the systems first.

Change is inevitable in financial firms – in fact, we think it’s best to assume you’ll be dealing with change on a daily basis.

There are many different things that can change, from business-level changes such as new systems, new partners or funds or products, to market changes such as corporate actions, volatility and – of course – regulatory updates.

The traditional operating model in financial services firms just isn’t set up to respond fast to change.

It may seem that dealing with data at large scale is the same as doing so at small scale, only doing more and faster. This isn’t the case, as the typical processes and technology in Finance and Operations aren’t flexible enough to accommodate this.

Think about straight-through-processing (STP). Most firms aim to get a certain level of STP (say, 90%) and then resource for solving the remaining exceptions (10% in this case). At some point in your firm’s growth, that percentage is going to represent a number of breaks beyond the realms of human cognition – and it’s only going to keep growing.

Many firms have tried to keep up with the scale of data through expanding headcount. That has left the largest organisations with literally tens of thousands of Operations workers. When a big change happens, like T+1 cutting a day off settlement windows, doubling those numbers simply isn’t an option.

WHEN SCALING DOESN’T WORK

Post-trade is a great example of an area that is difficult to scale to meet the demands placed upon it.

When market volatility spikes, that creates a lot of extra trades for the post-trade team to process. But the team is the same size and likely already at capacity.

Onboarding new team members would take months, so the firm can’t simply plug the capacity gap with new hires. Besides, when volatility calms down again, suddenly the team is overstaffed.

Data changes and evolves as it travels through your organisation.

In many financial services firms, the same original data is drawn into multiple systems and transformed or enriched by different teams for their own purposes.

This siloed way of working erodes trust in data. Indeed, teams will often perform their own data quality checks because they can’t see what another function has done with the data. It’s too risky to just hope for the best.

On top of this, the chances are that your data’s journey involves multiple systems and more than a few spreadsheet stopovers. All this makes it virtually impossible to track where it is, where it’s been and what has happened to it since it entered your organisation.

Given the high level of audit and regulatory scrutiny placed on financial firms – scrutiny that continues to increase in intensity as the years go by – being unable to answer these important questions about your data can have serious ramifications.

You understandably need to have strict controls in place to protect your organisation and your clients. But, for a lot of firms, these controls can be so restrictive that they slow the business down – or, worse, inject more risk into the business by forcing teams to use workarounds like end-user developed applications (EUDAs).

From a governance perspective, these controls protect the business against unwarranted change. All requests have to be documented and reviewed, so nothing happens without the proper oversight.

Yet many of these changes are required as soon as possible. If the Operations team needs a new reconciliation because the business has launched a new fund, they can’t wait weeks or even months for IT to build one. They need it right now. So they have to resort to some kind of work around, and usually a manual one at that.

So, ultimately, the procedures around governance and control often do exactly the opposite of what they intended. Balancing the need for control with the need to keep the business agile is something that’s only just becoming possible now, thanks to changes in the way businesses deploy and use technology.