Sainsbury: What are your priorities for the next 12-18 months?

Yuen: My priorities for the next 12-18 months are to implement Basel III compliance requirements in accordance with the timeline specified by Hong Kong Monetary Authority. On the already implemented AIRB, further enhance and extend model use to other functions as practised in other best practice banks. Continue to monitor the results of using quantitative risk management tools developed under Basel II to assess risk in the Bank’s business in the current economic environment where quantitative easing becomes a norm for the major trading blocs of US, European Union and Japan.

Sainsbury: How do you see the role of data changing in banking?

Yuen: Data used to be seen as a vehicle to achieve certain business objectives in a bank through system implementation in specific functional areas. Data were only managed when they were used for certain business purpose. But increasingly more and more banks realize that data are precious intangible assets to a bank in developing its business, particularly in the acquisition of new or existing customers for its new products and/ or services in its expansion plan. With the popularity of data architects and data managers being hired and data management departments being set up to manage enterprise-wide data, banks recognize the important role of  Bank’s internal structured data to enable its business expansion plan.

The next big waves for banks to pay attention to is the unstructured big data gathered in the social media and mobile devices in the retail business. In the next big challenges for banks, quantum leaps in technological advancement on data mining and data security protection would be necessary for banks to reap benefits on big data for their future banking business. 

Sainsbury: What impact does regulation have on a bank’s use of data?

Yuen: The privacy legislation in some advanced economies severely restricts banks in their use of customer data. Not only that different legal entities within a bank group would not be allowed to access freely on their customer data for cross selling purpose, even sharing data between departments within the same legal entity has to be justified on the “need to know” basis.

Sainsbury: What challenges is leveraging customer data placing on your systems, and how are those being resolved?

Yuen: The major challenges to banks in using the data extracted from computer systems are the silo approach in the development of legacy systems in which data were defined for a specific functional purpose, rather than using a bank-wide enterprise definition. So moving from one system to another requires extensive and tedious data mapping exercise. That tends to limit the efficiency of extracting data from current systems. But the latest technological advancement on data marts and data bases help to alleviate this constraints. More advanced banks also appreciate the importance of data analysis and data architecture, which results in the more recently developed computer systems to adopt the enterprise-wide data tables that can be accessed by multiple functional systems. 

Sainsbury: What opportunities and challenges are there to leverage mainland China infrastructure to provide competitive advantages in Hong Kong?

Yuen: With the increasing economic influence of China after the global financial crisis in 2008/09, our Bank benefits from the stated policy by the Chinese government in expanding China’s overseas investments in foreign assets. The Hong Kong banking industry, being the conduit for such overseas investments, benefits from these investment activities from mainland China. The geographic proximity of Hong Kong to China and the use of the Chinese language as one of the two official languages in Hong Kong do have its advantages in competing for the investment business from China when compared to other Asian locations or other financial centres. Although Hong Kong appears to understand China most as compared with other places in the world, we still experience challenges on whether we do know enough about China for all its intentions in the pursuit of these investments.

Sainsbury: What opportunities do you see to leverage a customer’s consumer electronics devices to provide them with a positive experience in engaging with the bank?

Yuen: I still believe it is still too early to talk about doing serious banking business over customer mobile devices. With so much unstructured data gathered on the behaviour of retail customers over the social media in mobile devices, the potential for future business through mobile devices cannot be ignored. At this juncture, banks should be concentrating in providing a positive image of the bank and its product offerings to generate brand awareness to mobile device users, hence gaining recognition from mobile device users that the bank is willing to progress with them on their preferential communication channels for banking business. However, major breakthroughs can only be achieved if the issues on the speed and security of these transactions would be resolved by advances in mobile technology.  

Sainsbury: Banks are looking outside of financial services to industries such as retail for examples of effective analytics applications. What are some of the most impressive examples that you have seen, and why?

Yuen: One should make a clear distinction between risk analytics practised in banking for consumer credit approval against marketing analytics practised in the solicitation of customer purchase in the retail industry. The latter is trying to guess customer preference in shopping consumer items based on past purchase behaviour or patterns in web browsing in the internet or social media. If the guess work on the customer purchase preference is not correct, there is little harm done to the retailer other than losing a potential selling opportunity. But the former involves a bank granting credit to a potential customer based his profile on facts such as customer delinquency records, income or employment for the purpose of approving a residential mortgage loan application or a credit card limit increase. The financial impact on a statistical error to an internet retailer is a small potential loss of business, while to that of a financial institution, is the actual loss of principal and interest.

As such, the data used for “guessing” a customer’s purchase preference can be quite lose because cost-benefit analysis would not justify spending too much resources by a retailer in gathering such data. However, the customer data collected in driving the risk score for making a credit approval decision has to be validated very carefully by the financial institution. The model developed to profile a customer’s purchase preference should therefore not subject to the same rigorous statistical validation test as in a risk model developed for credit approval purpose.

As a quantitative risk professional in analytics, I don’t see how banks would be looking to the retail industry for statistical techniques or tools to develop more effective risk models for credit approval.

Sainsbury: How do you encourage a culture of innovation within your team?

Yuen: Be intellectually curious and not take things for granted – that is my constant reminder to my team all the time. Always question whether there is other better ways to get things done. Doing it this way, our team forms a habit of critically analysing issues arising from work, and is able to generate innovative solutions to resolving these issues.

Sainsbury: If you weren’t working in financial services and technology, what would you be doing?

Yuen: Had I not worked for the financial services industry, I would probably be continuing my career in the IT industry. I started my career in the IT industry as an analyst programmer because I believe I had the logical mindset and discipline to be a good system professional. After a stint of ten years in IT system development as analyst programmer, project manager, and project director, my first exposure towards the financial services industry was in the management role in computer operations for a major bank in Singapore. From a computer system development and computer operations start, I moved on to become a risk manager in the financial industry.

Sainsbury: What advice would you offer to a young executive looking to step into a CAO role?

Yuen: Analytics is not just about risk modelling, market segmentation modelling or data modelling, it is about building usable models for business in functions of risk, market analytics or data marts. Although statistics are very often used as a tool to develop models, the success in the application of models in business lies in the good understanding of business, for which the model is built.  For young executive looking to step into a CAO role, I think building a solid foundation through career development on risks, data or marketing would certainly be beneficial.