Algo-assistance: The compliance vs profit margin battle | AssetManTech Series

With Annerie Vreugdenhil, head of innovation for wholesale banking, ING Group

By David Beach | 13 May 2019

Caught between the rock of regulation and the hard place of fee reduction, it's currently a difficult climate for banks and asset managers. 

And it is a growing list of rules and regulations for asset managers, from Dodd Frank in the US to Mifid II in Europe and the impacts of SDR, among others. A recent report revealed that banks and asset managers will have to comply with 374 legislative initiatives between now and 2021.

It is imperative that asset managers stay ahead of new rules and offset the cost and resource burden by generating alternative and greater revenues.

On the technology front, asset managers are looking to find a difference. bobsguide caught up with Annerie Vreugdenhil, head of innovation for wholesale banking at ING Group.

 

Transcript -  Annerie Vreugdenhil, Head of innovation for wholesale banking, ING Group

If you look in the financial market space, a lot has been happening there, probably for very good reasons, especially to create more transparency and to make sure that the clients are served the best possible way, which is really good.

But it has a couple of consequences on our side. And one of them obviously, is that with all the transparency that's coming into the system, things commoditized. Which means if you're at our end of it, that margins basically go down.

So on the one hand, you have the increase cost of finding out which regulation there is, complying with it, making sure you can actually deliver the transparency. And that sounds very simple, but you need very good systems to actually do that in which you have to invest.

And then your margins start to crumble, there's a lot of squeeze on what used to be a really good business. And in the end, you have to be able to, in the longer term serve your clients. So you have to be innovative about this.

Traders have a lot of knowledge and they know their markets really well. But they also have to basically look at loads of data sources in very short time to quote something when a client asks for it or when it comes on Bloomberg - basically, that's how it normally works. To digest all this information in such a short timeframe is simply hard. So we thought about how can we actually do that better? And how can we help them improve this?

We created Katana. And Katana is an algorithm that predicts a bandwidth in which it thinks the price will actually be when the trade is done. And we use lots of data sources for that - data sources that our traders are using - but also our data sources, more historical data sources, and we present to the trader the bandwidth in which the Katana algorithm things that will go.

And it leaves the trader with the possibility to pick within the bandwidth or differently, if they think that's appropriate. But helping them really to add something to all the experience to come to better proposal. And the great thing is that it actually seems to work. The hits ratio of the of the traders has gone up quite dramatically with 25%. But also the spread to the next best offer has reduced dramatically.

So our pricing has become better. In the end, you can win every deal. If you price it bit off market, you obviously don't want that. So that was a super interesting development, we're rolling it out now amongst more than one asset class, we started with an emerging market bonds, which is very good for our business.

And it gives us better performance of the whole financial market space, which in the end, you will need to be able to develop new products for our clients. Then the interesting thing was that when we started to publish about this, some of the biggest asset managers in the world, they came to us and they said, 'can we please have this algorithm?' we looked at that, and we said, 'let's see then what your problem is, and how we can play a role in that'.

We found out that their issue is different than for traders, because traders are market makers - they see one side of the deal. But an asset manager, they want to make a series of decisions that match for their purpose. So an asset manager that is responsible for portfolio of emerging market bonds, has to stay within a certain framework.

So say the framework is 100, then if you buy something, you have to sell something as well, because otherwise you're over 100. For an asset manager, the really important thing is that you find the ideal pair to sell and to buy. From the Katana algorithm for our traders, we created a changed algorithm that actually does that for asset managers.

This is where it really gets interesting to use technology because we, do this for emerging market bonds today, we will widen it, but you have to start somewhere. And there are some 20,000 emerging market bonds in the world. But it gives you 2 million possible combinations all the time for a pair of buy and sell.

And a human can simply not look at 2 million possible combinations every morning. And so this is a tool that helps asset managers to select the best pair we're having features in there to help them to tailor it to their risk appetite or their wishes. We're developing it all the time. We co developed it with one of the largest asset managers in the world PGGM and in the end it will bring a new business for ING.

Become a bobsguide member to access the following

1. Unrestricted access to bobsguide
2. Send a proposal request
3. Insights delivered daily to your inbox
4. Career development