Adam Wielowieyski is the Head of Collaborative Intelligence & Analytics at Hong Kong Exchanges and Clearing Limited (HKEX). Adam leads HKEX’s data strategy and is responsible for data adoption and analytics, as well as managing the infrastructure, governance and data science teams.
Can you tell us about your career path to date?
I studied computer science for my degree, and while studying worked with a travel agency to build one of the first air travel booking websites in Europe. After graduation, I established a consulting company that specialised in transactional database systems, and entered the banking industry, first with Morgan Stanley, then with Goldman Sachs - where I was a quantitative strategist on the Credit desk.
I moved to Hong Kong in 2007 and since then have been involved in a number of ventures – I ran a Portuguese real estate fund, and saw an opportunity to adopt Blockchain technology to improve relationships among fund investors. I advised HKEX on a Blockchain initiative, and soon ended up staying full-time in HKEX’s Group Strategy team, looking at strategic initiatives and M&A.
Data has now become a key driver of the Group’s strategic objectives, and I established the Collaborative Intelligence and Analytics team last year to drive and execute the data strategy for the firm. This team spans across the business and technology functions and drives the infrastructure, management and adoption of data and statistical analysis for the Group.
What does HKEX do?
HKEX is Hong Kong's core market operator, running cash equity and derivatives exchanges and associated clearing houses.
We are regularly the No.1 venue for IPOs globally, and have more than 2,400 listed companies with a combined market capitalisation of HK$34 trillion.
We also created the world’s most unique cross-border mutual market scheme – the Stock Connect in 2014. It enables international investors and Mainland Chinese investors to trade securities in each other's markets through the trading and clearing facilities of their home exchange. The Connect scheme then expanded to bond markets with the launch of Bond Connect in 2017.
Apart from our operations in Asia, we also own the London Metal Exchange in the UK, the global centre for industrial metals trading.
How is your company utilising data and machine learning and what kind of problems are you solving?
In many ways, I think all companies are technology companies these days, and we're all data-driven. Ultimately, data is information, and information gives you competitive advantage.
We use data in lots of different ways: we use it internally to provide colour to our business initiatives, for instance enhancements to our market microstructure help us decide what we should do and how it will affect our clients. We save costs by using data to drive digitalisation, and we’re using artificial intelligence to help validate and categorise unstructured document submissions into machine-usable data.
We also use data as a value generator - we sell our historical and real-time market data and other datasets that can provide insight into the market. But we have to be very careful, and take privacy and security issues as our top considerations.
What is the hardest challenge you have faced in terms of data within Financial Services?
Ensuring privacy. Broad availability of data spurs market innovation, but we must balance our clients' privacy requirements and ensure we are fully comfortable that data about our market is being used in a legitimate way.
Data can reveal a lot without you even realising it. Even if you have data that doesn't contain any personally identifiable information, you could glean trading behaviour or understand someone's trading algorithm. One must be very careful about accidental or involuntary disclosure.
One of the hardest challenges is balancing that need for data openness with the requirement for privacy and security. It’s key to ensure that you understand and know where your data is coming from; who owns it exactly; what's in that data; and that the quality of the data is controlled. That gives you the confidence not just to be able to open the data for use internally, but also to trust it.
That level of data governance can't be easy.
It's difficult. One thing that most companies have historically had is data silos, and we were no different. However, now that we operate quite an integrated business and create so many different forms of data, good decision making is critical to ensuring that we can share and access data efficiently across silos.
As you expand data availability, you do face the challenge of getting staff to really think about their roles and responsibilities for their data, because otherwise, any value you generate from your data is quickly going to be overwhelmed by the risk you’ve added to your business.
Do you see technology being more accepted in Financial Services and Hong Kong in particular?
In the last couple of years, it's been heartening to see how much things have exploded in financial technology and innovation here. There was a concern that innovation in Hong Kong was falling behind. But that's certainly not true anymore. Technology is starting to become much more visible, particularly in the retail segment, where banks and brokerages are competing on their technology offerings. But where technology often has the most impact is where it’s invisible – the automation of the likes of settlement and allocation work that we do, for example.
At HKEX, we've also embraced technological innovation in a big way. We set up the Innovation Lab in 2018 to experiment with new and emerging technologies, and we’ve taken some big steps in adopting many of the ideas generated inside there, not just for us as a business, but to benefit our markets broadly.
Is data access in the cloud a big topic for financial services in Hong Kong in 2020?
The cloud has been one of the biggest topics for the last couple of years, and there's no question that for data it's going to be more important as time goes on.
The main concern is the balancing of your data security with the convenience of cloud infrastructure. In the last few years, many of the cloud services providers have started to look at data privacy, security and encryption key management as being critical business value drivers. This has made the use of public cloud storage not only possible but also cost-effective, and we have been adopting cloud for our data storage and analytics in a big way.
This comes with challenges – it takes time to build expertise in cloud systems and learn how to best leverage the multiple providers that are now available; but nowadays it’s no longer a question of whether to adopt cloud technology, but rather to what degree you are doing so.
Are you seeing a shift from businesses towards the cloud?
It depends on the company and what they do. To benefit, you really need to think about where utilising cloud technologies will best solve your business needs.
The services available evolve very quickly and platforms differ in many ways. One must build knowledge across cloud platforms, and that can hold many companies back from adopting cloud services.
Across finance, there are some companies that have adopted cloud for the bulk of their systems, but that is still the minority. However, I would say that most are now using the cloud in some way. Anyone with a dependency on data is certainly very likely to be using cloud for their research and development and model training arms, though I think we’re still a way from seeing cloud adoption for trading and execution.
What successes have you had with AI?
A lot of the quick wins that come from effective data usage actually bring a surprising amount of tangible benefit, like the elimination of manual processes.
For example, it used to require us huge amounts of effort to read through, check everything and log information of the regulatory announcement of the listed issuers. We experimented in our Innovation Lab and subsequently deployed a couple of AI machine learning tools to read those documents, saving significant manual effort with 95% accuracy. Both have made our ability to monitor and respond to the market much more efficiently.
It doesn't make sense to hire more and more people as we grow our markets; we have to be able to scale effectively as we grow. But if you keep having unstructured data coming in, you've got to have some way of making sense of it and it’s here that machine learning/AI is really helping us.
What advice would you give to companies looking to invest in AI and data?
The biggest challenge that companies face is to demonstrate success and value from the use of data. You need users from across your business to understand the value of data as a corporate asset for them to maximise its utilisation and hence extract that value.
You will need to have a data governance policy to comply with the data privacy and security related issues. If you want to build a data business plan or strategy around your data, you will need to be confident in your data quality and that you are not going to fall foul of any compliance problems internally. A data governance foundation is also critical because that helps with the adoption within the business.
After the infrastructure and policies, it's about building a data-driven culture. You need to work on the talent, find the right people, and build a small centre of excellence who can help extract quick wins. You also need to train some of the key stakeholders across the business to learn how to utilise data effectively.
Have you struggled to find the right kind of talent for your team?
Yes. It's a minority of people who have that unique skill set spanning technology and business drivers and it's ever more valuable as time goes on. Good talent really is the key.
My view is that you don't necessarily need to have a background in financial services. The business of buying and selling stocks is intuitive. What is difficult is finding people who think logically and can explain quantitative aspects in layman terms. As we use computers to discover patterns more, it becomes more critical for people to provide logical thinking skills to understand why patterns are being identified.
At HKEX, we often have to think about, and understand the businesses and sectors our issuers are in, not just the smaller world of financial services. So, we look for people with not only great logical thinking skills, but who are generalists and interested in a range of topics.Posted about 2 years ago