Search

Using a large data set with 80,214 repeat sales, we find that the real return on a diversified portfolio of modern prints sold at auctions worldwide averaged a modest 1.51% during the period 1977–2004. We address several issues regarding the performance of modern prints as investments: the selection bias arising from the self-interest of auction houses; the impact of an ever-expanding universe of auction houses on investment returns; the “masterpiece” effect, or whether more expensive works of art outperform the market as a whole; and the differences in returns that arise due to random fluctuations in collector tastes. ( JEL Z11, G11, G14) [ABSTRACT FROM AUTHOR]

This article explores the properties of rational expectations in the context of autoregressive forecasting, the expectations-generating mechanism implicit in the use of distributed lag expectations proxies. The article then tests the rational expectations hypothesis on a specific set of directly observed expectations data. [ABSTRACT FROM AUTHOR]

In their analysis of corporate pension plans, financial economists typically invoke the simplest and the most tractable model of the labor market. This is the spot model, where the worker's cash wage plus accruing pension benefit equals the value of his marginal product in each and every period. This paper provides evidence against the empirical validity of the spot model, and uses provisions common to most pension plans to do so. These findings are relevant to the appropriate measurement of pension liabilities and thus to the present controversy surrounding the termination of overfunded plans and the recapture of surplus assets by the employer. The existence of incentive effects, ruled out by the spot model, may help to unravel certain well known "puzzles," such as the failure of employers to fully fund their plans in spite of the tax advantages of doing so. [ABSTRACT FROM AUTHOR]

The dependent variable in a distributed lag model frequently exhibits a distinct seasonal pattern which cannot be attributed to the seasonal behavior of the independent variable. This article illustrates the advantages of using the variable lag model to incorporate these seasonal influences directly into the distributed lag relationship. [ABSTRACT FROM AUTHOR]