Galton's fallacy and economic convergence

Friedman (1992) argues that regressing cross-country incomechanges on their final levels can be informative about -convergence(the tendency for the dispersion of income levels to narrow)whereas a similar regression on initial levels of income cannotbe. In this note we show that Bliss's (1999) dismissal of thisargument is in error.     

Education and the natural rate of unemployment

We show how changes in the educational composition of the labourforce affect both the level and the behaviour over time of aggregateunemployment series. We also demonstrate that if it had notbeen for such changes, the US unemployment series would look'European' since the within-group unemployment series all havethat same appearance. We derive a natural-rate model of unemploymentfor two education groups, providing microfoundations for inter-groupdifferences in wages and unemployment, and evaluate its plausibilityin light of microeconomic evidence.     

Sticky Price Models,Durable Goods,and Real Wage Rigidities

The standard two‐sector New Keynesian model with durable goods is at odds with conventional wisdom and vector autoregression (VAR) evidence: Following a monetary shock, the model generates (i) either negative or no comovement across sectoral outputs and (ii) aggregate neutrality of money when durable goods' prices are flexible. We reconcile theory with evidence by incorporating real wage rigidities into the standard model: As long as durable goods' prices are more flexible than nondurable goods' prices, we obtain positive sectoral comovement and, thus, aggregate nonneutrality of money.     

Valuation of bankrupt firms

This study compares the market value of firms that reorganizein bankruptcy with estimates of value based on management'spublished cash flow projections. We estimate firm values usingmodels that have been shown in other contexts to generate relativelyprecise estimates of value. We find that these methods generallyyield unbiased estimates of value, but the dispersion of valuationerrors is very wide - the sample ratio of estimated value tomarket value varies from less than 20% to greater than 250%.Cross-sectional analysis indicates that the variation in theseerrors is related to empirical proxies for claimholders' incentivesto overstate or understate the firm's value.     

Valuing American options by simulation: a simple least-squares approach

This article presents a simple yet powerful new approach forapproximating the value of American options by simulation. Thekey to this approach is the use of least squares to estimatethe conditional expected payoff to the optionholder from continuation.This makes this approach readily applicable in path-dependentand multifactor situations where traditional finite differencetechniques cannot be used. We illustrate this technique withseveral realistic examples including valuing an option whenthe underlying asset follows a jump-diffusion process and valuingan American swaption in a 20-factor string model of the termstructure.