Future Floods Workshop Flood models in the insurance industry Marie Delalay 27 February 2015 Workshop co-sponsored by NUS and AONBenfield.

Future Floods Workshop

Flood models in the insurance industry

Marie Delalay27 February 2015

Workshop co-sponsored by NUS and AONBenfield

© 2015 Deloitte AG. All rights reserved.

Agenda

• Introduction

• Purpose of Nat Cat models

• Flood models: how do they work?

• Challenges of flood models in South East Asia

• Conclusion

2

© 2015 Deloitte AG. All rights reserved.3

Introduction


Natural disasters in 2014Introduction

• 258 natural disasters

• 8’000 life losses among which:

− Floods in Afghanistan killed 2’600 people

− Floods in India and Pakistan caused 650 deaths

• In total: USD 132 billion of economic losses and USD 39 billion of insured losses (about 30%)

• Floods in India and Pakistan: USD 18 billion of economic losses and USD 700 million insured losses (about 4%)

4


IntroductionUneven insurance penetration

Source: Munich Re

Natural catastrophes worldwide 1980 – 2012. Overall losses and insured losses by income group

Source: Munich Re


Purpose of Nat Cat models


Purpose of Nat Cat modelsPricing the losses from natural disasters

• Gap between economic and insured losses gets smaller with rapid ongoing economic development, fast population growth and high urbanisation rate.

• Control measures and relief measures are improving. They reduce the humanitarian and financial consequences of natural catastrophes.

• But some of the impact cannot be mitigated and demand for insurance is rising in developing countries.

• Insurance companies have to be able to predict the losses in order to price them!


Purpose of Nat Cat modelsWhy does the insurance industry need Nat Cat models?

• Nat Cat models generate a loss distribution for the risk exposure of an insurance or reinsurance company.

• Need to determine the losses from Nat Cat insurance portfolio for:

− Underwriting and pricing for traditional insurances and reinsurances

− Pricing for Nat Cat bonds

− Risk management

− Capital requirement


Flood models: how do they work?


Flood models: how do they work?The 3 modules of a Nat Cat model

Monte Carlo Simulations

1. Hazard Module

2. Vulnerability

Module

3. Financial Module

Event Loss Table, EP

curve


Flood models: how do they work?1. Hazard module

• Generates a catalogue of flood events. This is a set of probabilistic and historical flood events with an associated return period and height of water at every grid cell.

• Input: Data on historic flood events

Historical flood events catalogue

Sampling of probabilistic

events

Complete floods

catalogue


Flood models: how do they work?2. Vulnerability module

• Given the intensity of an event at each location and characteristics of the insured item(s), what will the Mean Damage Ratio resp. the insured loss be?

• Input: historical insurance loss data

Source: Eqecat


Flood models: how do they work?3. Financial module

• Consideration of the insurance conditions: limits, deductibles, proportional or not proportional insurance, policy coverage

• Input: exposure of the portfolio in scope


Flood models: how do they work?Aggregation via Monte Carlo simulations

• Event Loss Table is built via Monte Carlo simulations

• Input: the flood catalogue, the vulnerability curves and the insurance portfolio


Challenges in flood models in SE Asia


Challenges in flood models in SE AsiaDifference between modelling Nat Cat and conventional risks

• Traditional risks are diversified by risk pooling

• Risk pooling for Nat Cat risk is less foreseeable

− Historical data are sparse

− Accumulation of risk

− Long-tail (either no loss or huge loss)

• Modelling losses from Nat Cat events is a challenge!


Challenges in flood models in SE AsiaLimitations in the data

• Sparse historical data

• Historical data are not representative of the future because:

• Increasing insurance penetration

• More concentrated exposure

• Increasing human intervention, e.g. improved defence structures


Challenges in flood models in SEModel uncertainty

• Intensity of floods depends on many factors and require a very high resolution.

• Gap in geographical coverage

• Failure to reflect effect of floods on supply chain

• Climate change only implicitly captured


Challenges in flood models in SE AsiaClimate change is implicitly modelled

Rationale behind the complex exercise of modelling climate change explicitly:

• There is a market demand towards understanding how climate change is captured in model.

• Historical data is not representative of the future.

• The long-term effects have to be considered and averaged out to obtain a year-on-year impact.


Conclusion


Conclusion

• Steep increase in the demand for flood (re-)insurance triggers need for strong flood models.

• Modelling flood events for the insurance industry is a complex task

• Challenges for flood models in Asia are the limited event and loss data, the need for a granular and wide geographical coverage, the current failure to model effects on the supply chain and the lack of explicit modelling of climate change.

• Because of the nature of the models both the knowledge of geographers and insurance practitioners are needed to overcome these challenges.


Q&A and Discussion

Deloitte refers to one or more of Deloitte Touche Tohmatsu Limited (“DTTL”), a UK private company limited by guarantee, and its network of member firms, each of which is a legally separate and independent entity. Please see www.deloitte.com/ch/about for a detailed description of the legal structure of DTTL and its member firms.

Deloitte AG is a subsidiary of Deloitte LLP, the United Kingdom member firm of DTTL.

Deloitte AG is an audit firm recognised and supervised by the Federal Audit Oversight Authority (FAOA) and the Swiss Financial Market Supervisory Authority (FINMA).

This publication has been written in general terms and therefore cannot be relied on to cover specific situations; application of the principles set out will depend upon the particular circumstances involved and we recommend that you obtain professional advice before acting or refraining from acting on any of the contents of this publication. Deloitte AG would be pleased to advise readers on how to apply the principles set out in this publication to their specific circumstances. Deloitte AG accepts no duty of care or liability for any loss occasioned to any person acting or refraining from action as a result of any material in this publication.


Future Floods Workshop Flood models in the insurance industry Marie Delalay 27 February 2015 Workshop co-sponsored by NUS and AONBenfield.

Documents

deloitte ag

challenges of flood

insured losses

economic losses

aonbenfield slide

eqecat slide

nat cat insurance portfolio

catalogue of flood events