Why two heads are better than one – the hedge fund perspective

Eileen Murray, co-CEO of Bridgewater Associates, helped launch the industry’s first dual-outsourcing model using two custodians. Joe Parsons speaks to Murray about how the world’s largest hedge fund has utilised this model.

By Joe Parsons

In 2016, Bridgewater Associates became the pioneer of the dual middle- and back-office outsourcing model. The unique system introduced a new “double-do” concept, whereby all middle- and back-office functions were performed by BNY Mellon, and were coordinated with Northern Trust as an independent verifier.

The new model enabled Bridgewater to concentrate on its core strengths as a hedge fund manager, while at the same time gaining greater oversight and assurance over hundreds of trade reconciliations and reports.

Two years after the model emerged, Global Custodian spoke to Eileen Murray, co-CEO of Bridgewater Associates, on how she feels this new model has gone, and how the dual-outsourcing model can be evolved.

Global Custodian: How have you found the dual outsourcing model experience? Have there been any complications?

Eileen Murray: This has been a terrific experience for us. What we have is an independent, dual-party verification model of client assets, ranging from trade all the way to custody. I came up with this by looking at what was going on in the world and seeing three major issues we needed to address. When I looked at what was going on in the world in terms of business continuity and potential disruption in different parts of the market, I identified that we needed to have a back-up to cover those different timelines. Secondly, we also needed a model that could provide independent valuations of our inventory. Finally, we deal with a tremendous amount of data and there were a number of regulatory changes coming out just before we adopted this model, so I thought why don’t we outsource our middle-office and our fund administration to a single counterpart and then have that re-verified by a second counterpart that is not in the same region.

Our goal was to go after client safety and security, increase accuracy with regards to transaction data, and increase resiliency, which gives us a strategic advantage. All of that is benefiting the clients as an overall risk and cost reduction scheme. 

GC: What have you learnt from using two custodians for the middle- and back-office, and what data have they produced?

EM: Having an independent verification of our client assets is really critical for us. Northern Trust and BNY Mellon review and reconcile our positions and cash throughout the day, and as a result, our error rates are substantially less than industry averages. Additionally, the increased resiliency we get from this model is really important. Having two independent technology and operation platforms that are geographically distributed is the best state-of-the-art resiliency. The strategic advantage is not just about ensuring assets are secure, but rather it’s about being supported by two major banks that are continuously improving their platforms, are continuously navigating an ever-evolving regulatory environment and are important global financial institutions, because as such they are subject to controls and oversight, which provides significant comfort.

GC: Have there been any complications?

EM: There have been some processes that moved quicker than we anticipated and some that took longer than initially thought. However, the partnership has been great; both banks and the clients are happy. By and large, I feel pretty good about it, and feel proud to be part of the Bridgewater team that came up with this method for improving the way models work in our industry. There were a lot of long days of dealing with the two banks in getting this ready, but they are done and I believe our strategic partners will continue to see business grown in this arena.

GC: Can the dual outsourcing model be replicated for other business lines?

EM: I can see this model being replicated in other segments of the business. For new areas of regulatory reporting, which requires interpretations and the filling out of new forms with tons of data that needs to be aggregated in a smart way, having this model where two banks help you go through this process is incredibly useful. Also, people may want to look at different types of risk models.

What’s important, however, is determining how important getting those processes right is and where that is in the priority stack because it takes time to get these things right, so we want to do this for the areas that make it better for our clients, firm and employees. For us, safety and independent verification of our clients’ assets is one of our top priorities. If I were a regulator, I would love to see this model across the industry.

GC: What other operational challenges are you facing, and what solutions are you looking for?

EM: We’re going to see more opportunities around real-time data, and artificial intelligence will be applied to these processes, which could make them more efficient and drive costs down. However, there is a lot of work needed to achieve that as you need to make sure that these are very well tested before implementing them. Real-time data in the clearance and settlement space and custody space would be something great to move along.

My view is that down the road this will continue to be a dual independent verification and it will continue to come down in price as technologies continue to enable greater efficiencies. When I look at the size of the markets, the amount of transactions that are done, the amount of data exchanged and the number of counterparties involved, having this double check is well worth it in terms of the high level of comfort you get from the increased level of security and resiliency. In my opinion those are two big things that will continue to drive this model.

In terms of clearing and settlement, custody and fund administration, I believe this model will continue to be adopted by others, because the total cost of ownership is actually cheaper when you consider regulatory filings and the like. As mentioned I would like to see more real-time data tools as well as the development of algorithms that pop not just data but information about counterparts or transactions that might be of interest. If you look at the lifecycle of a trade, it would be great if an algorithm could pick out and flag what to look at. I think you’ll see more and more of that intelligence come into the industry and particularly this part of the industry.