Subscribe to DSC Newsletter

We should attempt to creatively synthesize the many pluralistic approaches as well as focus more on synergistic interpretation of findings of these pluralistic researches.


We can recognize that though we cannot precisely predict black swans but forecasting emerging liabilities and their ratemaking can be a professional-character building experience in itself where we train to be better evolvers rather than better predictors alone.[1]


We can highlight that while recognizing facts (in form of quantitative analysis), it seems as if we only tend to scratch its surface as data, on its own, highlights results; whereas there are plenty of processes that culminate in data generation as well as modeling methodology in the first place. There is an incredible depth once when we start looking beyond the facts into fact-making itself; and this is where expert judgment and qualitative profiling can prove invaluable to guide the modeling exercise.


Agile Risk culture is foremost for any modeling exercise because complex systems like financial and insurance sector are not solely run by quantitative numbers, but by the underlying human psychology as well. It is up to the risk culture to not antagonize in binary opposites like complex/simple, good/bad etc, but to reach the middle ground to converge communication and mentalities between different stakeholders.


In the end, it is useful to keep a few sobering meditations in mind:

  • We suffer too profoundly even from small data glitches.
  • Better than many complicated equations are few statements that give clarity to shareholders
  • The experience of all deep datasets is slow. They must wait long until they know what has fallen into their depths. Machine learning can lower that waiting time.
  • Generally, there is either over-reliance on data and models or negligible reliance on them. We have to be familiar with the golden mean that resides between two vices. So here our data and modeling orientation should be in between the extremes of reliance on only opinions and only data and models.
  • Unless one considers intention—philosophy, cognitive system, behavioral bias, etc.—used in building data, models and expert’s analysis, and implications, one can be missing the big picture already.
  • Provide historical data to limit the amount of work required for attaining a context for the data but data should be adjusted to reflect current conditions, not historical circumstances.[2]
  • Focus on developing a ratemaking plan, not numerical premium and projections only.[3]
  • Know your context.[4]
  • Beware of qualitative shifts.[5]
  • Know how the model results will be used.[6]
  • Do not anthropomorphize models. Anthropomorphism is the tendency to characterize animals, objects, and abstract concepts as possessing human-like traits, emotions, and intentions. Models are not reality or real human social behavior. At best models are idols[7]; at worst a distraction and cause for herding.

[1] Mills, Allan: Society of Actuaries (2010): Complexity Science: an introduction and invitation for actuaries.

[2] Philippos Papadopoulos April 2015 The Zen of Modeling

[3] Ibid

[4] Werther; SOA 2013; Recognizing When Black Swans Aren’t: Holistically Training Management to Better Recognize, Assess and Respond to Emerging Extreme Events

[5] Ibid

[6] Ibid

[7] Wilmott, P. & Derman, E, 2009. “The Financial Modelers’ Manifesto”

Views: 379


You need to be a member of Data Science Central to add comments!

Join Data Science Central

© 2021   TechTarget, Inc.   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service