Knowledge that Transforms

This paper examines segregation in American cities from 1890 to 1990. We divide the century into three time periods. From 1890 to 1940, ghettos were born as blacks migrated to urban areas and cities developed vast expanses filled with nearly exclusively black housing. From 1940 to 1970, black migration continued and ghettos expanded. Since 1970, there has been a decline in segregation as blacks have moved to suburban areas and central cities have become less segregated. Across all of these time periods there is a strong positive relation between urban population or density and segregation. We then examine why segregation has varied so much over time. We find evidence that the mechanism sustaining segregation has changed. In the mid-20th century taken by whites to exclude blacks from their neighborhoods. By 1990, these legal barriers enforcing segregation had been replaced by decentralized racism, where whites pay more than blacks to live in predominantly white areas.

Up to six months ahead of actual production, U.S. automakers announce plans for their monthly domestic production of cars. A leading industry trade journal publishes the initial plans and then a series of revisions leading up to the month in question. We analyze a panel data set spanning the years 1965–95, matching the production forecasts with data for actual monthly production. We show that a firm's plan announcement affects competitors' later revisions of their own plans and eventual production. The interaction appears to be complementary: large plans or upward revisions cause competitors to revise plans upward and increase production. The results are consistent with theoretical models in which firms share information about common demand parameters.

The objective of this paper is to compare the effectiveness of financial markets and financial intermediaries in financing new industries and technologies in the presence of diversity of opinion. In markets, investors become informed about the details of the new industry or technology and make their own investment decisions. In intermediaries, the investment decision is delegated to a manager, who is the only one who needs to become informed, which saves on information costs, but investors may anticipate disagreement with the manager and be unwilling to provide funds. Financial markets tend to be superior when there is significant diversity of opinion and information is inexpensive. Journal of Economic Literature Classification Numbers: G1, G2.

Households and businesses in the U.S. prefer to use checks over less costly means of payment. Earlier studies have focused on check “float” as an explanation for the continued popularity of this seemingly inefficient technology. We construct a general equilibrium model of check payment and show that the presence of float does not necessarily lead to inefficiency. However, we also identify two potential sources of inefficiency associated with check float: (1) if float is not always priced, then it acts as a distorting tax, and (2) inefficiencies can result if people engage in costly activities designed to accelerate check presentment. Journal of Economic Literature Classification Numbers: E58, G21, G28.

We compare institutional execution costs across the major U.S. exchanges using a sample of institutional equity orders in firms that switch exchanges. Execution costs including commissions are essentially indistinguishable across these exchanges. We also find the fraction of trading volume from momentum traders is greater on the NYSE than either the Nasdaq or AMEX and that orders are more likely to be worked by an institution's trading desk on the NYSE than on the Nasdaq. These results suggest that institutions actively manage execution strategies, taking into account characteristics of the markets in which they trade.

In 1982, Dimensional Fund Advisors launched a mutual fund intended to capture the returns of small-cap stocks. The ‘9–10 Fund’ is based on the CRSP 9–10 Index, an index of small-cap stocks constituting the ninth and tenth deciles of NYSE market capitalization, although the 9–10 Fund incorporates investment rules and a trading strategy that are aimed at minimizing the potentially excessive trade costs associated with such illiquid stocks. As a result, the 9–10 Fund provided a 2.2% annual premium over the 9–10 Index for the 1982–1995 period. We show that both the investment rules and the trade strategy components of the Fund’s design contribute significantly to this return difference.

This paper tests models of mutual fund market timing that allow the manager's payoff function to depend on returns in excess of a benchmark, and distinguish timing based on publicly available information from timing based on finer information. We simultaneously estimate parameters which describe the public information environment, the manager's risk aversion, and the precision of the fund's market-timing signal. Using a sample of more than 400 U.S. mutual funds for 1976–94, our findings suggest that mutual funds behave as highly risk averse, benchmark investors. Conditioning on public information improves the model specification. After controlling for the public information, we find no evidence that funds have significant market-timing ability.

We simulate standard tests of performance persistence using alternative return-generating processes, survival criteria, and test methodologies. When survival depends on performance over several periods, survivorship bias induces spurious reversals, despite the presence of cross-sectional heteroskedasticity in performance. Look-ahead biased methodologies and missing final returns typical of U.S. mutual fund datasets can also materially affect persistence measures. Our results reinforce previous findings that U.S. mutual fund performance is truly persistent. When fund performance is truly persistent, fund attrition affects persistence measures, even when the sample includes all nonsurvivor returns. We also examine the specification and power of the various persistence tests.

When a rate of return is regressed on a lagged stochastic regressor, such as a dividend yield, the regression disturbance is correlated with the regressor's innovation. The OLS estimator's finite-sample properties, derived here, can depart substantially from the standard regression setting. Bayesian posterior distributions for the regression parameters are obtained under specifications that differ with respect to (i) prior beliefs about the autocorrelation of the regressor and (ii) whether the initial observation of the regressor is specified as fixed or stochastic. The posteriors differ across such specifications, and asset allocations in the presence of estimation risk exhibit sensitivity to those differences.

We use a genetic algorithm to learn technical trading rules for the S&P 500 index using daily prices from 1928 to 1995. After transaction costs, the rules do not earn consistent excess returns over a simple buy-and-hold strategy in the out-of-sample test periods. The rules are able to identify periods to be in the index when daily returns are positive and volatility is low and out when the reverse is true. These latter results can largely be explained by low-order serial correlation in stock index returns.