Big Data and Marine insurance: what is the status? – IUMI ARTICLE 09/22


Let us see what is shared on the net about Innovation: “disrupt the future of post pandemic
underwriting including blockchain,… AI, and machine learning: learn how to leverage data to support
augmented underwriting”, and so on…


Is that so easy to deal with innovation and traditional insurance business? And what is the
balance between concrete added value and challenges?


Surely, some return on experience about business cases will help us…


1/ Business case:


Who? Actuaries of Marine Insurer


Pain point: lack of accurate insured values impacting badly the insurance pricing and the amount
of indemnification


Solution: proceed to clean internal data of vessel references, hunt external data (values) and
aggregate internal and external data


Added value: a better pricing and indemnification process, as well as increased productivity and
better interactions between actuaries and underwriting


The next step is to correlate vessel characteristics with loss data to discriminate risk profiles.

2 / Business Case:


Who? A marine underwriter


Pain point: Lack of visibility of actual risk per vessel and lack of KPI regarding portfolio risk
profile.


Solution: an insurtech platform automatically connected with global data sources, supported by
data science and providing: a Risk Rating and an Emission Estimation of vessel, fleet and
portfolio, with detailed risk radar explaining the rating,

Added value: a quick and much more accurate risk analysis with productivity. Moreover, Emission
Estimation, as an underwriting criterion, is supporting the operational declination of ESG, aligned
with the Poseidon Principles
To conclude, here are key transversal learnings we are happy to share as an insurtech:

  • The status of historical data is a challenge but not a blocking point,
  • The leverage of data science is huge regarding the whole combined ratio:
    productivity, losses (right amount/avoidance), premium (better matching with each
    risk profile),
  • At the end, innovation should serve every stake holder with the democratization of
    the data science: insurers, clients and brokers, with more transparency about risks.