Strate's Big Data Analyst, Dr Merrill van Der Walt
Strate’s Big Data Scientist, Dr Merrill van Der Walt

History tends to repeat itself, if the age-old axiom is to be believed.


The true test of character, for an individual or an organisation, is whether they can harness the intelligence of these events from the past, and apply this intelligence to find opportunities in the challenges.


Today we talk to Strate’s Big Data Scientist, Dr Merrill van der Walt, who has the formidable task of evaluating historic trends to be able to better predict future reoccurrences.


After all, the data pool exists, but it takes human intervention to derive any form of insights out of it.





Q: Dr. van der Walt, what is Big Data?

Consider Big Data to be large, continuous and granular data sets that may be analysed computationally to reveal patterns, trends, and associations, especially relating to human behaviour and interactions, or in our case, how those behaviours or interactions impacts the financial services industry, and the market as a whole.


Q: What are the characteristics that define Big Data?

The three Vs

  1. Volume
  2. Velocity
  3. Variety


If you want to analyse a Big Data set, you can only really refer to it as a Big Data set if it has these characteristics. Interestingly, Big Data analyses allows for the consolidation of both structured and unstructured data sources. Including such a richness and diversity of data, brings about deeper levels of analytics.


Q: How do you gather data?

There is a range of analytical tools that can pull and process Big Data sets. The software we use has advanced analytical capabilities. It’s not a one size fits all approach – if you want to evaluate Big Data, you have to be as agile as Big Data.


Q: Once gathered, what do you do with the data? What is your objective?

Our goal is to build predictive financial models. Something that will warn us if it detects recurring behaviours that may potentially lead to market “crashes”. The more we experiment with the information, the more opportunities come to the fore. Having strong analysis means we can add more value to our industry partners.


Q: How do you account for human emotion when building predictive financial models?

The Financial Services industry has been built upon human emotion. Therein lies the charm of a robo-adviser, as it can give you unadulterated advice, devoid of emotion. Machine learning is a huge part of this, and it wouldn’t surprise me if we start seeing the rise of artificial intelligence designed specifically for the financial services industry.


Q: What are the short-, medium-, and long-term benefits of evaluating and interrogating this data?

Who wouldn’t want an accurate predictive financial model? Increased accuracy means good news for the industry, market, and consumer alike. Personally, I hope that this will provide some long-term stability, where current events don’t impact the market that dramatically. Well, dependent on the event, of course. Consider #Brexit – if the world had an accurate predictive financial model, we may not have seen losses as high as $2-trillion.


Q: What kind of data do you evaluate? What are your sources?

As I mentioned earlier, we have to focus on volume, variety and velocity. Which is why we focus on as many data sets as we can get our hands on. Firstly, our own book, evaluating the nigh on 300 000 trades we process every day for the last few years – that’s a great (and statistically rich) starting point. But we also have to consider unstructured data, because that’s where we’ll find the exciting intelligence, such as the nuances of language. For that purpose, we look at video and email, and maybe even social media data sets.


Q: Are you aware of any organisations that are incorporating Big Data Analytics into their operating models, for instance, to improve customer relations, or refining the supply chain?

I know of a few insurers abroad that are successfully using Big Data Analytics to offer customers personalised premiums. In our industry, however, the majority of organisations playing in the big data pool are in the middle of experimentation, and trying to figure out exactly what it is that they want to achieve with this intelligence. A few are doing it, without much ado, but my guess is that we will see more and more organisations entering the fray in the coming months and years.




Recommended Posts