It was an excellent turnout for the Investment Data & Technology Summit in Sydney last week with so much discussion around the role of data as a competitive advantage. During the day there was a mix of expert speakers presenting on many topics including:
- front office
- hedge fund world
- data transformation programs
- academia and investment management specialists
Here is our round up of the best takeouts from the day.
Best-of-breed systems to get up and running quickly
Several large organisations are using best-of-breed systems to fulfil a specific role in business processes. By selecting an agnostic platform, this approach is possible. An off-the-shelf solution can be implemented a lot more quickly in a phased approach and, importantly, an organisation is not tied to a large-scale enterprise system (or vendor) which can be hard to move away from.
Using a best-of-breed solution will enable organisations to get up and running on systems which have been bullet tested already, as opposed to trying to build it yourself when its not a core competence.
Data Quality is a big concern
Data quality was a big topic throughout the day as a core problem and there are issues of identifying anomalies within data sets. If data quality issues exist, then it really doesn’t matter what systems you have in place, because decisions will still be made on data which is lacking trust by investment teams.
Quality and speed of the data are very important and while visualisation is a great feature, it is irrelevant if the data is not trusted. The hedge fund panel mentioned that as data is becoming easy to access, the people accessing the data don’t really understand it and then they are basing investment decisions on this. Another risk.
Having quality data and a single source of truth will help in creating that alpha.
Is the rise of the machine going to make current investment roles redundant? What does the future look like?
Around 20 years ago, 50% of trades were completed by a computer in Australia – now it’s close to 85%. Most trades are now completed via computers with only about two open outcry markets left in the world.
There is no doubt that technology is having an impact. However, computers are dependent on assumptions that are made within models and for machine learning to be successful, machines must properly understand the assumptions to avoid errors.
Quantifiable data is only a small subset of available data. A hybrid approach of humans and machine becomes so much more powerful than any one on its own. The computational power of machines means vast data can be processed very fast, but humans can see around corners and computers can’t.
There is still a reliance for humans to apply critical thinking and it will still be a long time for machines to be able to do this. It is this reason why humans still have a distinct advantage over machines. The concept of machine learning should be incremental to the process and what teams are doing.
Several organisations are going through extensive data transformation programmes
In terms of data transformation programmes, it is vital to have executive sponsorship to help drive change. It is a challenge to change behaviour of the people that are making the investment decisions. It’s a common obstacle for innovation and trying something different, such as implementing new systems. People may have been doing tasks the same way for years, so it is important for investment teams to be brought into the programme early and included in the journey.
Another success factor on data transformations was to isolate the teams from their daily activities and co-location has proven to work the best. This all helps lead to the ideal optimum solution, as opposed to a temporary band-aid solution.
The goal of transformation programmes is around “How do we make investment teams better”.
Data analytics to analyse trends and forecast
The data available, both structured and unstructured, is so vast now. The winners will be those organisations that can process data quickly and then act upon the results in their investment decisions. For example, there was talk of Google being able to predict when the next flu epidemic is going to be via models and when during the year the world is at its most fat and most fit. Amazing really.
Creating Alpha
Successful organisations have the mindset and discipline to be successful. It is important to source the right people first and foremost, and then you need continuous innovation of what you are doing, particularly in model risk. Innovation is about taking your analytics to benefit the workflow.
Success will come from always pushing and evolving and not being static. History tells us that there is a range of different challenges at a specific point in time. Models can help us predict, but they don’t make the decisions… it’s people that do this.
—–
AlphaCert is an investment data management platform which acts as the hub within our customers’ architecture. Data is pulled into the platform and enriched by numerous data sources (both internal and external) and then is integrated to specialist platforms such as Ortec Pearl, BlackRock Aladdin, Tableau and FactSet.
If you would like an AlphaCert demo or to enquire about how we might be able to help through our professional services, please get in touch here.