Final Report Summary - EMF-FEIM (Empirical Macro-Finance and the Financial Economics of Insurance Markets)
The project has two components: The economics of insurance markets and asset pricing. I discuss each component in turn.
1) The economics of insurance markets: The traditional literature on insurance markets emphasizes demand-side frictions due to adverse selection or moral hazard. Such frictions are used to explain insurance prices across different groups of consumers and over time, and to explain both contract characteristics and missing insurance markets. In this project, we explored the role of supply-side frictions in life insurance markets, and in particular financial/regulatory frictions and imperfect competition. Such frictions also affect the equilibrium in insurance markets.
One example is our work on “shadow insurance,” in which we show that insurance companies can relax regulatory requirements by moving life insurance policies from operating companies, which sell directly to consumers, to reinsurance companies. As the reinsurance company is owned by the holding company of the operating company, there is no risk transfer (the traditional motive for reinsurance). We show that the possibility of using shadow reinsurance lowers prices in insurance markets and increases the size of insurance markets, which may be beneficial to consumers. However, as shadow insurance raises the insurer’s overall leverage, this benefit has to be compared to the potential cost of increased financial instability.
Some of this work may be relevant in understanding the potential for risk mismatch in the insurance sector and risk transmission to other parts of the financial sector (and hence systemic risk). An important concern in this context is that the failure of an insurer that causes significant losses for households may affect the trust that consumers have in insurance companies. As a result, households may scale back on their holdings of insurance products, which is costly as we have shown as part of the project.
2) Macro-finance and asset pricing: The first part of this project studies risk premia across maturities and across asset classes. In our project, we have shown that Sharpe ratios decline with maturity for equities, government bonds, corporate bonds, and volatility. This consistent feature across asset classes is at odds with various leading, representative-agent asset pricing models. In other parts of this project we explored risk premia associated with a security’s “carry” (i.e. the return an investor earns if market conditions do not change).
Traditional general equilibrium asset pricing models have limited success in explaining these facts. As the second part of this project, we proposed a new equilibrium asset pricing framework that starts from the demand curves of households and intermediaries. While any asset pricing model implies an asset demand curve, the underlying demand curves are often quite simple with limited heterogeneity, and easily rejected using actual holdings data. Consequently, asset pricing models are generally not tested using holdings data. We directly model the demand curves and estimate them using actual holdings data. We show how to compute counterfactuals in this framework to analyze the role of institutions in financial markets. This framework can also be used to analyze the impact of the ongoing asset purchase programmes of central banks as we have done in one of our papers.
1) The economics of insurance markets: The traditional literature on insurance markets emphasizes demand-side frictions due to adverse selection or moral hazard. Such frictions are used to explain insurance prices across different groups of consumers and over time, and to explain both contract characteristics and missing insurance markets. In this project, we explored the role of supply-side frictions in life insurance markets, and in particular financial/regulatory frictions and imperfect competition. Such frictions also affect the equilibrium in insurance markets.
One example is our work on “shadow insurance,” in which we show that insurance companies can relax regulatory requirements by moving life insurance policies from operating companies, which sell directly to consumers, to reinsurance companies. As the reinsurance company is owned by the holding company of the operating company, there is no risk transfer (the traditional motive for reinsurance). We show that the possibility of using shadow reinsurance lowers prices in insurance markets and increases the size of insurance markets, which may be beneficial to consumers. However, as shadow insurance raises the insurer’s overall leverage, this benefit has to be compared to the potential cost of increased financial instability.
Some of this work may be relevant in understanding the potential for risk mismatch in the insurance sector and risk transmission to other parts of the financial sector (and hence systemic risk). An important concern in this context is that the failure of an insurer that causes significant losses for households may affect the trust that consumers have in insurance companies. As a result, households may scale back on their holdings of insurance products, which is costly as we have shown as part of the project.
2) Macro-finance and asset pricing: The first part of this project studies risk premia across maturities and across asset classes. In our project, we have shown that Sharpe ratios decline with maturity for equities, government bonds, corporate bonds, and volatility. This consistent feature across asset classes is at odds with various leading, representative-agent asset pricing models. In other parts of this project we explored risk premia associated with a security’s “carry” (i.e. the return an investor earns if market conditions do not change).
Traditional general equilibrium asset pricing models have limited success in explaining these facts. As the second part of this project, we proposed a new equilibrium asset pricing framework that starts from the demand curves of households and intermediaries. While any asset pricing model implies an asset demand curve, the underlying demand curves are often quite simple with limited heterogeneity, and easily rejected using actual holdings data. Consequently, asset pricing models are generally not tested using holdings data. We directly model the demand curves and estimate them using actual holdings data. We show how to compute counterfactuals in this framework to analyze the role of institutions in financial markets. This framework can also be used to analyze the impact of the ongoing asset purchase programmes of central banks as we have done in one of our papers.