Benefits, Limitations Abound for ‘as-a-Service' Platforms When it Comes to Risk

I've been a technology reporter for about six-and-a-half years, all here at WatersTechnology. There's one thing that I've learned very well in that time: When someone from a bank or asset manager that doesn't work in technology starts talking about technology, it's usually in the most general terms. The word "technology" is the catch-all for any tool that someone from the business or non-tech C-level uses.

This is obviously not true for everyone ─ some of the savviest hedge fund traders in the world are also elite programmers ─ but they simply don't have the same breadth of knowledge as a chief technology or information officer. And nor should they have that level of ability; it's their job to make money, not patch networks.

Smarter Risk

I bring this up because this week I attended the Smarter Risk Summit in New York, which was held in a beautiful hall inside the New York Stock Exchange. The event was sponsored by IBM ─ about as hardcore a technology company as you can find ─ and Risk magazine, which is a sibling publication of WatersTechnology. Risk touches on technology all the time, but more from a business, regulatory and modeling perspective.

Broadly speaking, the conference aimed to highlight the power of analytics when managing risk. I was there for the opening panel, Transforming Risk With Next Generation Data Management. The moderator, Amiel Goldberg of IBM, brought up the idea of turning to risk-as-a-service (RaaS) solutions to help firms better flatten their data architectures and to ensure that the finance department is working with the same information as the risk department.

Now, as you may remember, IBM bought the highly-respected risk analytics vendor Algorithmics back in 2011 for $387 million. Since then, IBM has been able to bundle its cloud-based Algo Risk Service with its IBM Algorithmics Managed Data Service. The solution won last year's Best Buy-Side Risk Management Initiative at the Buy-Side Technology Awards 2015.

It's a risk-data-as-a-service play, and it's proven quite effective on the buy side. But it also got me wondering about the idea of an "as-a-service" model when it comes to managing risk.

Hesitation? 

The Smarter Risk panel featured three industry experts: Jennifer Agnes, who had just left GE Capital after 20 years with the organization; Marc Alvarez, chief data officer at Mizuho Securities; and Thomas Mavroudis, global head of data management at HSBC. These three were perfect for this kind of panel ─ data people that understand data management platforms/challenges, and work closely with the business-side of the firm. But none of the three specifically took the bait on RaaS.

Mavroudis had this to say: "It's a noble cause, but very difficult. You don't own most of the data. As a chief risk officer, 90 percent of the data, you're a consumer and maybe you create 10 percent of that data, highly aggregated data. ...To control that data and to build a service out of that data ─ and to have that data be fit for purpose ─ it's a cultural transformation to be able to work with the businesses, have SLAs, be able to understand the lineage, and build all of that is a very difficult thing to do."

Alvarez said that "there's a big role for technological capabilities within the firm" when it comes to risk management, but said these programs are difficult because of the "culture shock" that these projects tend to create amongst the data owners and creators. When it comes to risk as a service, he concluded, "I think we'll be hearing a lot about metadata, a lot about control, but we are nowhere near there yet."

And Agnes noted something that should be obvious, but that is too often ignored: "The big stumbling block is the different silos ─ finance teams, risk teams, data teams, modeling teams," adding that "it's not OK to be siloed anymore," but that still comes down to a cultural change, and those are never easy.

What's Appropriate? 

No one seemed willing to really delve into which specific functions are ripe for RaaS, and which areas they would never in a million years turn over to a third party. To my earlier point, maybe that's because most of the people in attendance weren't technologist and this wasn't a technology conference. [Shameless Plug: The next two technology conferences on our agenda are the North American Trading Architecture Summit on April 21 in New York and the European Buy-Side Technology Summit in London on May 17.]

So that's what I'm interested to hear about: Where do RaaS platforms fit into the technology landscape? As was noted in my February feature looking at risk analytics for alternative investments, you need a robust set of models, approaches and measures for managing risk. Some of that will be done in partnership with third parties; some risk operations will always stay in house.

Everything is seemingly going down the "as-a-service" path. Risk is already on that path. But how much, still remains to be seen.