Marvels and Misconceptions of Insurance Production

AbstractHow does insurance work? What is the relative advantage of the insurance firm in bearing risks? In the literature of business, mathematics and economics these questions are answered by ‘risk pooling'and the law of large numbers (LLN). The popular argument is: by selling many contracts LLN helps predicting fairly accurately the loss. Because the risk of insolvency is real and troubling, enhanced confidence in its underwriting results affords the competing larger firm an advantage that ought to be reflected in lower prices. In other words, with regard to solvency there are returns to scale in insurance production. I dispute this specious argument on empirical grounds. Concentrating on firm's survival, without invoking utilities, using a simple model of insurance with binomial risk. LLN indeed affords returns to scale in survival but only in numbers too large to be practical. In the actual demand for insurance there are negative returns to scale in risk bearing as each additional customer exacerbates the hazard of financial ruin of a firm. There is simply not enough demand for insurance in the world to enjoy LLN.According to Samuelson's  conjecture (1963), which was never followed, the relative advantage of the stock insurance firm in bearing risk lies in the multitude of owners, not of clients, through the spreading of the risks among the many shareholders.  I show that insurance production manifests returns to scale in its fundamental and unique inputs: equity and shareholders and that reinsurance is a perfect substitute for those two inputs.  

Elath Hall, 2nd floor, Feldman Building, Edmond J. Safra Campus
Friday, May 4, 2007 - 10:00 to 12:00
Old Lecturers: 
Yuval Shilony
Old Lecturers University: 
Bar-Ilan University