Creativity needed for big data to drive revenue, says risk actuary Stefan Swanepoel

“Taking advantage of opportunities requires more creative thinking than simple analysis of data,” says risk actuary Stefan Swanepoel, the principal consultant for Directrix Risk Services, an actuarial and corporate risk management consultancy specialising in the pricing, financing and control of risks. Swanepoel has more than 13 years of experience in this field and currently consults to many of the largest South African companies, as well as many large international corporates. A fellow of the Institute of Risk Management and the Actuarial Society of South Africa, Swanepoel followed his father, a state actuary, into the profession after serving time in the military. He spoke to us about the impact of Big Data, risk management, the changing climate in South Africa and how it compares to foreign nations.

Do you find that South African firms have problems defining problem specifications and simplifying models?
"If the model tries to satisfy too many requirements in advance, there can be too many distractions and you will end up polluting the problem specification. If a model is built to solve a problem and this has been done conceptually correctly, it shouldn't be too complicated. Sure, they can be big and have complex elements, but they don't necessarily need to be convoluted. Model parity means not adding a variable if it does not make a difference to the quality of the answer. We work on the basis of maximum output and quality with the minimum number of assumptions and variables. It is inherent in what we do to keep things as simple as possible, because we start with complex problems anyway. If you further complicate things, you will end up chasing your tail very quickly."

What are some of the challenges firms face concerning Big Data?
"Computing speed and capability are considerations when building a model, but the biggest concern about Big Data is the quality of the source and capturing, which can be mitigated to some extent by automated systems. We still sometimes find outliers, due to faulty systems. These won't all lead to incorrect conclusions, but you need to be aware of issues prior to starting modelling. Manipulating data is the easy part - it is the data clean-up and validation that can be challenging."

Can Big Data help firms improve their performance?
"I believe the use of data goes back to the first Model T in the production line. We need to ensure the benefit of analysing this data is worth the effort. Of course, you won't always know this in advance, but understanding your environment and your impact on it is definitely a benefit of intelligent data analysis."

How are local firms adapting to the influx of Big Data?
"
In terms of the internet-based generation, there is a fair amount of sophistication. Massive amounts of operational data are being created and analysed extremely well from an operational engineering perspective by the large mining houses and similar operations. However, there is sometimes a disconnect when supplying data at board level. This is a necessary consequence, because summarisation of data by definition means you lose detail in the process."

You've worked in Bermuda and Taiwan. How do their systems compare to those of South African firms?
"Bermuda is an extremely focused offshore captive domicile and has fairly well standardised systems. Some of the larger organisations I worked for over there have very good operational reporting systems. It is a global phenomenon that captive can be tools to manage your business and central repositories for data. The data management and operational efficiency in Taiwan was an eye-opener for me. Every input is clearly understood and utilised in the planning process. In one of the big companies I worked for, there were hardly any variances from budget, because they plan down to the last cent. At the same time, while they are extremely well run and adapt quickly, they can lose out on opportunities by focusing n a single element like risk management. Taking advantage of opportunities requires more creative thinking than simple analysis of data."

You've mentioned that level of education in Taiwan is high. Does South Africa have the human resources to take full advantage of Big Data?
"Definitely - the level of education in Taiwan is a consequence of that country's socio-economic structure. I don't think it is a pre-requisite for success. An example of this is the MCSE frenzy in the late nineties. Everyone was desperate for graduates at the time but now you can barely get a job with one. You do, however, need technically minded people. The market naturally adapts to provide what is required. There are quite a lot of challenges in this country, but they are not insurmountable. There are a fair amount of resources allocated to Big Data but there is a fair amount of scope for growth in that area, too."

Are companies using Big Data to drive revenue or purely to streamline operations?
"
My experience to date has been mostly on the efficiency side because I tend to focus more managing risks, but yes, there has been revenue growth. I've had several clients ask me to build models to test products and new ideas. It takes creative thinking prior to massive data analysis to drive revenue growth. These novel ideas and concepts may need data be gathered going forward, while firms tend to analyse data that is currently available to improve efficiencies."

How effectively are South African companies managing risk?
"
There are pockets of excellence. The well-established, well-financed industries have sophisticated risk management processes and new legislation is forcing banks and insurers to comply with certain requirements."

"The link between operational management and risk finance is often tenuous but even this is improving as we progress a country. Potential growth is dependent on finance and operations coming together. Qualified people understand the natural framework for financial and operational risk."

"There is a culture change occurring in SA towards better ownership of risk. There is some way to go in creating the capacity and resources necessary for effective risk management and to create an exercise structured and measured properly to add value and not become an endeavour in itself."

Is there still scope for actuaries in the public sector?
"Yes and no. Several state-owned enterprises and insurers, including the Development Bank of Southern Africa, the Land Bank and the Road Accident and Unemployment Insurance funds, require actuarial services related to insurance. However, extending this into the risk space from pure actuarial science is a new area. One of the large municipalities I am working with is doing an exercise to integrate risk financing into its insurance. The perception of actuaries as mere number-crunchers for insurance purposes is changing."

You were the main speaker at a recent CFO SA dinner. What were your impressions?
"There is definitely value in this sort of intimate event. When the audience is bigger, you lose the opportunity to interact directly. Maintaining a conversational tone helps clarify context. You can't do this with a speech and the lessons can be difficult to assimilate."

  • Stay connected, up to date and in the loop on what is happening in the world of finance and keep track of newly published expert insights and interviews with CFOs and CEOs. Become an online member and receive our newsletter, follow us on Twitter, like us on Facebook and join us on LinkedIn.