The Value of Data 

In December 2022, Global Investor brought together the industry’s leading transition managers and consultants to discuss some of the key issues affecting clients including costs, the impact of interim solutions, the evolution of data analysis, and how business models are adapting to market conditions.

A portion of the 2023 Transition Management Roundtable is available in the video above. See below for a transcript of the highlights. 

Amelie Labbe: Is the transition management industry fully realising the value of data?

James Woodward: Transaction cost analysis and the use and application of data has advanced significantly over the last three to five years. Historically, TCA has been strong in the equity space with a good set of vendors and that has extended to fixed income, foreign exchange, and exchange traded derivatives.

It is important to understand the quantum of data that we are talking about for each transition event. In the equity space, hundreds of individual orders can actually result in thousands of orders and tens of thousands of fills. It is those fills that are a result of people using algorithms and smart order routers to acquire and distribute orders to where liquidity is strongest and analysing that data is key.

For the client it is used to provide transparency, a third-party lens into the transition data that enables the client to see items such as performance and attribution. It can also be very useful to drill down into a very granular level of detail where there might be a single stock in a particular transition that drove the overall outcome of events.

The client may be able to get a detailed understanding of participation, stock performance before and after trading, reversion, particular algorithmic strategies that were used by the transition manager and the counterparty used, even down to gauging the level of signalling risk and whether the algorithm was deemed to be having too much market impact.

Moving away from a specific client event, transaction level data becomes more meaningful when it is examined over a longer period of time and spans a number of events. That particular data set can be used by a transition provider to analyse whether an algorithm or a particular counterparty is more or less effective than alternatives. It can be useful in assessing participation levels across different markets and to help improve future outcomes or future strategies that have been generated.

Transaction cost analysis can also help assess trading team performance, and individual trader performance.

Ashish Patel: Data has very quickly become the most valuable commodity in our industry aside from human capital. In a transition we get data from various sources - whether it be the investment manager or the custodian – and also generate much of it ourselves, whether it be for the purpose of transaction cost analysis or defining trading strategies at the outset.

It is our job to know the client and how much data we present that makes sense whilst being transparent. It is about converting potentially hundreds of thousands of lines of data into an easily digestible format.

It is hard to talk about data and analytics without thinking about the new wave of machine learning and evolution of algorithms. This can be divided into four pillars, the first of which is client intelligence - how we can better understand or use data to understand what our clients want from us and deliver that service in an ever-evolving complex world.

The next pillar is market intelligence, how we can use that data to create more efficient trading algorithms. The third pillar is automation and we have talked about how we can create operational efficiencies and mitigate operational and project management risk, which is within our control, as opposed to execution risk or market risk. The fourth pillar is supervisory intelligence - how can we use that data to protect ourselves and our clients to ensure that we uphold the highest level of fiduciary obligations during the transition?

David Edgar: Transaction cost analysis in itself has limited value for each individual transition because as James pointed out, the value lies is looking at lots of trades where you are doing the same thing over and over again. But for all the people working at trading desks, who are monitoring their daily trades and using it to constantly improve their systems, processes and algos it is invaluable.

Can you look at the minutiae of the orders in a particular transition and draw too many conclusions from that as it relates to that transition? Possibly not, because you are often required to trade in a particular way due to external factors. You have to start with the macro strategy and work down and providing you have got your strategy right at the top level, everything else should follow. However, transaction cost analysis is key for trading desks to improve their processes.

James Woodward: About 80% of a good transition post-trade report is attribution at the parent level - high level attributions, starting at sectors and countries and then moving down to the stock level. It is only probably a subset of that post-trade report that warrants driving down into the individual transaction cost analysis for a particular stock, although if there are those outliers it is obviously important to be able to provide that level of detail should the customer be interested.

It also very much depends on the sophistication of the customer. You can’t go and dump 20,000 securities on a customer and expect them to know what to do with it – they need the right views and lenses that some of the transaction cost analysis vendors can provide.

Chris Adolph: One of the key evolutions of data is that we are able to drill down much more into the component parts of risk in a portfolio. An example would be looking at very stock specific risks - so instead of just looking at whether you are overweight Russia vs Germany, for example, you can drill down to whether 5% of the total risk in the whole transaction is this particular name.

Andy Gilbert : Data for the sake of data does not deliver much to a client – what they really want are insights into how the markets and execution came together to give us the outcome we got? Data can help, but are they going to be able to derive conclusions from it? Maybe not. That is where we help them understand what the data means and what insights they can draw from it.

Cyril Vidal: I think I was the first to mention around this table a few years ago the electronification of fixed income. We see far more diversified, granular portfolios which brings another challenge around data – how do you synthesise all the information about liquidity of particular bonds into a single number that you can present to the client and evaluate the transaction cost accordingly?