Research topics,
fundings and grants

Main Funding

Chaire innovation de l’UQAM

Co-operators – (2018-2023, 500 000$)

In 2018, the insurance company Co-operators offered UQAM $500,000 to create the Co-operators Chair in Actuarial Risk Analysis (CARA).

The first scientific objective of the Chair is the use of recent statistical and probabilistic methods allowing optimal segmentation of risks between the various stakeholders involved. To this end, historical loss experience will be applied in combination with advanced methods that take advantage of all available data (loss experience analysis via telemetry data, megadata/Big Data, machine learning algorithms, etc.). Thus, advanced statistical estimation methods will be used. More specifically, the focus will be on pricing and provisioning contexts.

The second scientific objective of the Chair, which is more conceptual, is to mathematically analyze and develop new tools useful for P&C actuarial work. A fundamental understanding of the statistical techniques used, and the development of new inference methods better adapted to the context of P&C actuarial work (machine learning techniques, simulation techniques, Bayesian methods, filtering techniques, etc.) are of crucial importance here.

Other Fundings

Underwriting, actuarial valuation and profitability analysis of a new global insurance product

Collaborative Research and Development (CRD) Grants NSERC – (2019-2024, 653 500$)

The project partner, The Co-operators Insurance Company, is involved in all insurance markets in Canada. This Canadian insurance cooperative strives to meet the insurance needs of all Canadians. We believe there is an insurance deficit in Canada: the insurance products currently available in the Canadian insurance market are offered on a piecemeal basis. Consequently, to obtain full coverage of their property, insureds must purchase separate insurance products for each piece of property they wish to protect, and must remain in constant communication with their insurer to report any changes in their coverage needs.

The main objective of this collaborative research project, based on the actuarial expertise of UQAM professors, is to develop an innovative insurance product in which clients would be offered an insurance product that meets all of their current and future insurance needs. Such a product would ensure that insureds would no longer have to worry about having to notify their insurer about any changes to their insurance contract: all their insurance needs would automatically be covered by a single annual insurance premium.

The research project poses real actuarial challenges, and is divided into three distinct areas, which include sub-steps. The pricing research axis (T) aims to set a premium for various insurance profiles for the overall insurance product; the provisioning and diversification axis (P) aims to quantify and manage the insurance liabilities generated by this insurance product; and the customer value research axis (CV) assesses the short- and long-term profitability of this product.

Premium liabilities and dependency between coverages

Education and Good Governance Fund (EGGF) of the Autorité des marchés financiers AMF – (2017-2019, 47 446$)

Claim liabilities are the reserve that the insurance company must record in its liabilities to allow for the full settlement of claims already incurred by its insureds, but for which the amount of claims is not fully known or paid. The actuarial scientific literature is largely devoted to the modelling, valuation and risk management of these liabilities. Yet very little has been done about the risk caused by premium liabilities. A major component of this risk is unearned premiums (UNP), which corresponds to the insurance coverage related to the portion of the premium written that remains to be paid under an insurance contract.

The objective of this project is to propose a model of the risk associated with UNPs that would take into account the dependence between the different lines of business present within a company, and to analyze the diversification possibilities that this dependence allows. It is possible to define several partial objectives to achieve this overall objective:
(i) critically review the few models available in the actuarial literature dealing with NLPs, (ii) develop one or more classes of models that will allow the inclusion of dependence between lines of business, and (iii) analyze potential diversification benefits.

Adequate modelling of this risk is essential for the Autorité des marchés financiers (AMF), or any other regulatory body, to ensure adequate oversight. Better understanding UNP risk will also be useful for general insurance companies, which will benefit, in particular, from diversification benefits. This will allow for more rigorous control of the company’s solvency, more effective risk management and better capital allocation.

Hierarchical pricing models: statistical and bonus-malus systems approaches

Academic Research Grant Program Canadian Institute of Actuaries – CIA (2017-2019, 18 500$)

P&C company actuaries have several longitudinal data sets at their disposal to develop their pricing and reserve valuation models. This means that in each of these databases, the same insured vehicle is observed for several years. Several forms of dependency can therefore be observed, at multiple levels: between coverages of the same annual insurance contract, between insureds or vehicles of the same annual contract, between payments made for a claim, and between annual contracts of the same insured (time dependence). It is important for insurers to analyze and understand the dependencies that may exist in their portfolios. They can then better estimate both their risk and the insurance premiums, together with the discounts they could grant to insureds depending on the coverage chosen and the insureds’ claims experience.

The main objective of the project is to develop hierarchical models that will effectively capture these different levels of dependence in a company’s portfolio. To do this, we consider two approaches: 1) a theoretical approach based on advanced probabilistic models and recent statistical inference techniques; and 2) a more practical approach that bypasses some theoretical and technical problems, inspired by bonus-malus systems (SMB).

Discovery Grants Program

Natural Sciences and Engineering Research Council of Canada (NSERC)

The Discovery Grants Program supports ongoing research programs with long-term research objectives rather than a single project or series of short-term projects. It recognizes that creativity and innovation are at the heart of research breakthroughs. Discovery Grants (DGs) are a “research support” since they provide long-term operating funding and can facilitate access to funding from other programs, but they are not intended to cover all the costs of a research program.

Skip to content