How can the actuarial profession contribute to data science and risk management in a way which is valued beyond insurance and finance?
Following the Casaulty Acturial Society's lead, I start with the Drew Conway data science diagram – which surely has an anaology for risk management.
I suggest skills gaps, but also that there is a way forward which focuses on that most central of acturial skills: model building.
Data science skillshow actuaries measure up
What skills do data scientists need? What have actuaries got?
The US Casualty Actuarial Society's Data and Technology Working party 'seeks to research and identify the knowledge and skills actuaries must possess to participate in the changes brought about by a rapidly evolving technology supporting data and analytics'.
Conway suggests that good data science needs:
- maths and statistics knowledge
- substantive expertise (e.g. domain knowledge of the industry / firm you're working in
- hacking skills (data wrangling, using software and machine learning techniques)
The initial conclusion from the diagram above is that actuaries have what Conway calls the 'traditional research' skills; maths/statistics and domain knowledge, especially for insurance.
Compete or collaborate?
Actuaries and Data Scientists – Match Made In Heaven or Hell? suggests that actuaries and data scientists have overlapping skills and should collaborate. That also seems to be Cherry Chan's conclusion. I want to collaborate, but fear that actuaries may face an uphill challenge. I'll explain why, and what we might do.
What the actuarial profession saysthe 2016 strategy
- Actuarial science is the application of mathematical and statistical methods to assessing financial risk.
- It is a science that can be applied in a variety of business contexts where understanding the likely financial impact of possible future outcomes is beneficial.
- Our members are regulated professionals, bound by an ethical code of conduct...
- ... who are qualified and remain relevant through a rigorous system of education, examination and continuing professional development.
- As quantitative risk professionals, actuaries take the long view to inform sustainable business decisions that balance risk and reward.
- Actuaries understand how to model uncertainty of financial outcomes and how to communicate these implications in a business and policy context.
- Their trusted expert analysis and independent judgement ...
- ... helps them deliver innovative integrated solutions and advice to help make financial sense of an uncertain future.
Applying the actuarial skillset more widelybeyond financial services
Actuaries seeking to deploy their skills beyond insurance and financial services will face two questions:
- Do you have a specific technical skillset?
- For data science this will be the Conway diagram and, specifically, data wrangling and machine learning skills.
- For risk management this may be different risk/uncertainty assessment techniques – there are more than 40 – and control expertise.
Without these an actuary may fall back on domain-driven statements, with a technical and professional wrapper e.g. 'actuaries are expert in the modelling and management of financial and other risks, serving the public interest with a professional code which provides a moral compass'.
- Do you have specific domain expertise?
- When the insurance / financial service specialism is noted this may prove a barrier.
- But other professions have succeeded – think of the mobility of finance directors.
- The question comes back to: do actuaries have that cross-industry expertise?
The best approach is to note that some skills are, by their very nature, applicable across industries: risk management, data science and model building are three (related) examples. My judgement; actuaries' current and core competitive advantage lies in model building, not in risk management or analytics.
The good news for the profession (if I'm right) is that most for profit organisations – and perhaps others – could benefit from actuaries' model building skills. And actuaries could kick on from there, most naturally into model-led risk management. Model building might feel like a step backwards. It's not.
Less positively, I see little possibility of actuaries' skills as a profession being acknowledged in either data science or risk management, without adopting the model building approach. In data science actuaries will generally be outgunned technically, while in risk management they will be undercut on price.
A cashflow model is a central tool – but not the only tool – of risk management. By contrast, data science modelling involve extending the statistical modelling concept, where actuaries have the potential but have less well-developed day-to-day skills. There is scope for an actuary with a suitable appetite to excel.