• Selected
• Entire List

Abstract:
Over the years numerous portfolio performance measures have been proposed. In general they are designed to capture some particular enhancement that might result from active management. However, if a principal uses a measure to judge an agent, then the agent has an incentive to game the measure. Our paper shows that such gaming can have a substantial impact on a number of popular measures even in the presence of extremely high transactions costs. The question then arises as to whether or not there exists a measure that cannot be gamed? As this paper shows there are conditions under which such a measure exists and fully characterizes it. This manipulation-proof measure looks like the average of a power utility function, calculated over the return history. The case for using our alternative ranking metric is particularly compelling in the hedge fund industry, in which the use of derivatives is unconstrained and manager compensation itself induces a non-linear payoff and thus encourages gaming.

Abstract:
This article develops a Kalman filter model to track dynamic mutual fund factor loadings. It then uses the estimates to analyze whether managers with market-timing ability can be identified ex ante. The primary findings are as follows: (i) Ordinary least squares (OLS) timing models produce false positives (nonzero alphas) at too high a rate with either daily or monthly data. In contrast, the Kalman filter model produces them at approximately the correct rate with monthly data; (ii) In monthly data, though the OLS models fail to detect any timing among fund managers, the Kalman filter does; (iii) The alpha and beta forecasts from the Kalman model are more accurate than those from the OLS timing models; (iv) The Kalman filter model tracks most fund alphas and betas better than OLS models that employ macroeconomic variables in addition to fund returns.

Abstract:
Consider an economy in which the underlying security returns follow a linear factor model with constant coeffcients. While portfolios that invest in these securities willin general, have a linear factor structure, it will be one with time-varying coeffcients. However, under certain assumptions regarding the portfolio's investment strategy, it is possible to estimate these time-varying alphas and betas. Importantly, this can be done without direct knowledge of either the portfolio manager's exact investment strategy or of the alphas and betas of the individual securities in which the portfolio invests. This paper develops and estimates a Kalman filter statistical model to track time-varying fund alphas and betas. Several tests indicate that relative to a rolling OLS model the Kalman filter model produces more accurate fund factor loadings both in and out of sample. This appears to be in large part due to the attempts of fund managers to time the market by varying their fund's risk exposure from period to period. Another advantage of the Kalman filter model is that the dynamic parameter estimates can be used to classify funds by their trading strategies and to determine the source of a fund's profits or losses. The tests in this paper indicate that the superior and inferior returns produced by some funds arise almost entirely from attempts at market timing rather than managerial selectionability. However, as other research in the area of mutual fund performance measurement have found, overall there appears to be little evidence that, inaggregate, fund investors earn superior returns.

Abstract:
This paper proposes a simple back testing procedure that is shown to dramatically improve a panel data model's ability to produce out of sample forecasts. Here the procedure is used to forecast mutual fund alphas. Using monthly data with an OLS model it has been difficult to consistently predict which portfolio managers will produce above market returns for their investors. This paper provides empirical evidence that sorting on the estimated alphas populates the top and bottom deciles not with the best and worst funds, but with those having the greatest estimation error. This problem can be attenuated by back testing the statistical model fund by fund. The back test used here requires a statistical model to exhibit some past predictive success for a particular fund before it is allowed to make predictions about that fund in the current period. Another estimation problem concerns the use of a single statistical model for all available mutual funds. Since no one statistical model is likely to fit every fund, the result is a great deal of misspecification error. This paper shows that the combined use of an OLS and Kalman filter model increases the number of funds with predictable out of sample alphas by about 60%. Overall, a strategy that uses very modest ex-ante filters to eliminate funds whose parameters likely derive primarily from estimation error produces an out of sample risk-adjusted return of over 4% per annum.

G12, G13

Abstract:
This paper proposes a simple back testing procedure that is shown to dramatically improve a panel data model's ability to produce out of sample forecasts. Here the procedure is used to forecast mutual fund alphas. Using monthly data with an OLS model it has been difficult to consistently predict which portfolio managers will produce above market returns for their investors. This paper provides empirical evidence that sorting on the estimated alphas populates the top and bottom deciles not with the best and worst funds, but with those having the greatest estimation error. This problem can be attenuated by back testing the statistical model fund by fund. The back test used here requires a statistical model to exhibit some past predictive success for a particular fund before it is allowed to make predictions about that fund in the current period. Another estimation problem concerns the use of a single statistical model for all available mutual funds. Since mutual funds often, but not always, employ dynamic trading strategies their betas move over time in a ways that differ from fund to fund. Since no one statistical model is likely to fit every fund, the result is a great deal of misspecification error. This paper shows that the combined use of an OLS and Kalman filter model increases the number of funds with predictable out of sample alphas by about 60%. Overall, a strategy that uses very modest ex-ante filters to eliminate funds whose parameters likely derive primarily from estimation errors produces an out of sample risk adjusted return of over 4% per annum.

Mutual fund performance, back test

Abstract:
The roles played by idiosyncratic risk and liquidity in determining stock returns have recently received a great deal of attention. However, recent empirical tests have not examined the interaction between these two factors. As others have shown (and this paper confirms) stocks idiosyncratic risk and liquidity are negatively correlated. To what extent then is each variable responsible for the observed cross sectional patterns in stock returns? Overall, using monthly data, the paper finds that stock returns are increasing with the level of idiosyncratic risk and decreasing in a stock's liquidity. However, while both liquidity and idiosyncratic risk play a role in determining returns, the impact of idiosyncratic risk is much stronger and often eliminates liquidity's explanatory power. The point estimates indicate that a one standard deviation change in idiosyncratic risk has between 2.5 and 8 times the impact of a corresponding change in liquidity on cross sectional expected returns.

idiosyncratic risk, liquidity, stock returns

Abstract:
This paper examines governance explanations for the discount of preferred shares to common shares in the Russian market. Conflicts between shareholder classes may help explain the discount. However, for this to be the sole explanation the estimated models suggest that the magnitude of future adverse shareholder events would have to be very high. Nevertheless, evidence of a common factor potentially related to governance seems evident in the data, implying that corporate control issues may at least be partially responsible for the observed preferred share discount.

Abstract:
Home equity is the single largest component of household wealth for the majority of American households. Yet, there is virtually no way for the average family to insure itself against drops in home value and the ensuing destructive financial loss. Much of U.S. housing policy has focused on helping them against the risk that home ownership entails.

In this paper, we document the development and implementation of a home equity insurance program launched in 2002 in Syracuse, New York. The range of issues arising from the practical implementation of a home equity insurance program, as well as the institutional challenges offer useful data for further extensions of the program.

Highlights of the outcome, to date, of the pilot program include the finding that implementation of the program was feasible on the local level, that customers understand and wanted to take part, and that clean data on housing transactions is a vital component of the future success and expansion of the project.

Abstract:
This paper presents a model in which investors cannot remain in the market to trade at all times. As a result, they have an
incentive to set up trading firms or financial market
intermediaries (FMI's) to take over their portfolio while they engage in other activities. Previous research has assumed that such firms act like individuals endowed with a utility function. Here, as in reality, they are firms that simply take orders from their investors. From this setting emerges a theory of mutual funds and other FMI's (such as investment houses, banks, and insurance companies) with implications for their trading styles, as well as for their effects on asset prices. The model provides theoretical support for past empirical findings, and provides new empirical predictions which are tested in this paper.

Abstract:
Nearly any standard financial model concludes that two assets with identical cash flows must sell for the same price. Alas, closed-end mutual fund company share prices seem to violate this
fundamental tenant. Even when one considers several standard frictions, such as taxes and agency costs, classical financial models cannot explain the large persistent discounts found
within the data. While the standard financial markets model may not explain the existence of large closed-end fund discounts, this paper shows that a rather close version of it does. In an
otherwise frictionless market, if asset supplies vary randomly over time and agents posses finite lives a closed-end mutual fund's stock price may not track its net asset value. Furthermore, the analysis provides a number of conditions under which these discrepancies will lead to the existence of systematic discounts for the mutual fund's shares. In addition, the model provides predictions regarding the correlation between current closed-end fund discounts and current changes in stock prices and future changes in corporate productivity. As the analysis shows the same parameter values that lead to systematic discounts also lead to other fund price characteristics that resemble many of the results found within empirical studies.

Abstract:
Prior to a takeover bidders have the option to purchase a toehold in the target firm at market prices. Yet, despite the availability of this option empirical studies show that in many cases firms purchase very small or no toeholds. This is somewhat puzzling since current theoretical models indicates that this is suboptimal. Why do bidders routinely ignore what appears to be a readily available profit opportunity? This paper proposes an answer via a detailed analysis of the legal environment in which tender offers take place. Essentially, toeholds drive up the preoffer price which in turn drives up the legally required "clean up" price, that is the price paid for shares that are not taken up in the initial tender offer. As a result, we show that firms will purchase toeholds only if rival bidders are expected. Comparative statics demonstrate how optimal toeholds change in response to changes in the legal and economic environment. The paper also shows that toeholds provide the initial bidder with both low cost shares and insurance in the event that a rival bidder enters and purchases the target. However, the while toeholds do provide some insurance to losing bidders it is incomplete. Initial bidders always do better when a rival does not appear. Further analysis of the model also shows that "fair price" provisions always allow the most efficient rival to win a takeover battle, despite any toeholds the first bidder may have. Finally, using simulations, we show how some of the more puzzling results in the empirical literature can be explained by the model. For example, there appears to exist a negative correlation between toeholds and stock prices. The simulations show that this may be due to correlations among the model's deeper parameters that have not been controlled for in the current empirical literature.

Abstract:
Prior to the announcement of a tender offer, the bidding firm is legally allowed to acquire shares in the open market, subject to some limitations. These pre-announcement purchases are known as toeholds. This paper presents a simple model that describes the bidder's optimal toehold acquisition strategy, within an environment that closely parallels the present legal institutions. The model shows that toeholds and bids interact in a complex manner even without the presence of asymmetric information. By examining a simple environment the paper provides a useful alternative hypothesis for tests of other, presumably more complex, models. One of the main implications of our model is that if no competing bidders are expected, no toeholds should be purchased. Indeed, under a wide variety of conditions small toeholds are optimal. The paper also demonstrates that the correct specification of an empirical model can be critical. For example, under some parameter values toehold purchases may exhibit a negative cross-sectional correlation with the pre-announcement run up in the stock price. This occurs even thought prices are strictly increasing the size of the toehold. Several implications concerning various aspects of merger legislation are considered. For example, we demonstrate that a rule similar to a "fair price" provision has the desirable property that a second bidder arrives and wins if and only if he places a higher value on the target that the initial bidder.

Abstract:
This paper examines governance explanations for the discount of preferred shares to common shares in the Russian market. Conflicts between shareholder classes may help explain the discount. However, for this to be the sole explanation the estimated models suggest that the magnitude of future adverse shareholder events would have to be very high. Nevertheless, evidence of a common factor potentially related to governance seems evident in the data, implying that corporate control issues may at least be partially responsible for the observed preferred share discount.

Abstract:
A cross-sectional analysis of all trading suspensions that occurred during the period 1974-1988 in the New York Stock Exchange reveals that though the desire to maintain price continuity remains an important motivation to suspend trade, inventory imbalance fears are pronounced for large firms. Adverse selection concerns afflict all news related suspensions irrespective of firm size. Further, we find substitutability amongst the various dimensions of liquidity: while large cap stocks have lower bid-ask spreads, they halt more often. A time-series analysis shows that the resiliency of the exchange -- its ability to absorb severe volatility shocks -- has improved in this period.

Abstract:
A cross-sectional analysis of all trading suspensions that occurred during the period 1974-1988 in the New York Stock Exchange reveals that though the desire to maintain price continuity remains an important motivation to suspend trade, inventory imbalance fears are pronounced for large firms. Adverse selection concerns afflict all news related suspensions irrespective of firm size. Further, we find substitutability amongst the various dimensions of liquidity: while large cap stocks have lower bid-ask spreads, they halt more often. A time-series analysis shows that the resiliency of the exchange -- its ability to absorb severe volatility shocks -- has improved in this period.

Trading halts; Market resiliency; Liquidity; Inventory control; Adverse selection

Abstract:
Home ownership increases the incentive to maintain property and neighborhood, as well as decreasing the outflow of rents from low-income zones. However, these benefits are not costless to homeowners. With a mortgage comes the possibility of default, the financial demands of maintenance, a reduction in alternate investment opportunities, an increased exposure to fluctuations in local economic conditions, and a drastic reduction in the liquidity of personal wealth.

Recently, policy makers have sought to increase mortgage lending in traditionally underserved markets. In this paper we consider the effects of this policy in light of the risk and return of housing and the current tax treatment of the home mortgage deduction. We find housing to be a relatively poor asset class in which to invest the bulk of family wealth. Trends in housing suggest that a large percentage of homeowners who bought and sold within a five year horizon in the United States over the last twenty years lost money on the investment. Lowering the equity required to purchase a home does little to alleviate the problem. We show that the current tax code - if anything - encourages renting over buying and gentrification of low income housing markets.

If the government wishes to encourage home ownership among low income families despite the risks, then we argue that government agencies should share information about the risk and return of home ownership with its citizens. In addition, a direct subsidy through a tax credit may be both warranted and necessary to achieve the desired result.

Abstract:
This paper presents a model of a mature city that depends upon the rehabilitation of old home sites for new housing. Within the model housing returns, housing construction, mortgage loan terms, and household maintenance behavior are all endogenous. The model?s premise is that the value of a home, unlike the value of many other financial assets, depends upon the care its owner exerts on upkeep. Banks respond to this moral hazard problem by restricting the size of the loans they are willing to issue. As a result people bid what the can for housing, rather than what they may wish to. This in turn ties housing prices to changes in the endowment process which are both predictable and time varying. When endowments are growing quickly (a city with a rapidly growing economy) housing prices exhibit above market expected returns. Because banks within the model act rationally, they set mortgage terms based upon their beliefs regarding future housing prices. This leads to the empirically verified prediction that current mortgage loan to value ratios can be used to forecast future housing returns. Developers are also fully cognizant of how housing prices are set and react accordingly. When housing prices are expected to increase faster than the rate of interest developers acquire land for construction. Then once the developers believe housing returns will stop increasing they develop and sell their holdings leading to a construction boom. Thus, the model predicts that developer holdings can be used to forecast housing returns.

Abstract:
This paper presents the results from a statistical analysis of the first Florida recount. The findings indicate that it is highly unlikely that the relative increase in Gore's vote total can be explained by mechanical reading errors. Rather it appears partisan biases influenced the outcome. Estimates indicate that on average if a ballot's status changed from no vote to a vote, the chance that it went to Gore was about 15% higher than one would expect given his fraction of that county's vote. Overall then, controlling for each candidate's vote in a county and the type of ballot used, this paper estimates that Gore picked up 903 "too many" votes in the recount relative to what would have been expected by chance machine read errors. If humans influenced the results how did they do it? During the recount the ballots were put through the tabulating machines several times. However, machine readings tend to vary from run to run and this means humans, partisan humans, have to decide which of several tallies to report. Potentially, this discretion allowed the preferences of those conducting the recount to impact the reported totals.

Abstract:
This article addresses the issue of how closely the fortunes of suburbs are tied to the fortunes of the central city. We develop housing price indices for most of the zip codes in California and use them in a clustering procedure to determine whether city and suburban housing markets naturally aggregate or move separately. We find that central cities tend to group with their suburbs, suggesting that the housing markets of cities and suburbs are closely linked.

Abstract:
This paper presents a model in which asymmetric information and extreme uncertainty lead to the exclusive use of equity and riskless debt for small business financing. The paper derives these results without any restrictions on the available contract space, the distribution function governing a project?s payoff, or the risk aversion of most potential entrepreneurs. Linear securities derive from the assumption that small business financing involves more uncertainty than is captured in most financial models. Instead of assuming that business people are faced with a given menu of projects, the model allows entrepreneurs to create (over time) an unlimited number of non-positive net present value projects with any payoff distribution they desire. Also, outside investors cannot observe project choice but only terminal cash flows. As a result, suppliers of funds must design contracts so that in equilibrium entrepreneurs do not wish to undertake undesirable investments. Further analysis of the model shows that in equilibrium entrepreneurs must contribute some of their own capital. The paper also finds that when a firm does not have any collateralizable assets, the equilibrium funding agreements have the property that the investors split the project?s proceeds in proportion to their initial investments. We further demonstrate that the existence of an infinite number of non-positive NPV projects can lead in equilibrium to positive abnormal returns earned by outside financiers. The model also produces a pecking order theory of financing. It is shown that highly profitable firms, and/or those companies run by relatively risk neutral individuals will first try to raise funds with riskless debt, and then turn to equity only when the supply of riskless debt has been exhausted. Finally, the paper shows that the linearity of the equilibrium contracts derives from the entrepreneur?s discretion with regard to the payoff distribution, rather than the flexibility associated with non-positive NPV projects. As part of the same analysis, we also demonstrate that under some conditions our underpricing results can be derived solely from the infinite supply of non-positive NPV projects.

Abstract:
This paper presents a model in which asymmetric information and extreme uncertainty lead to the exclusive use of equity and riskless debt for small business financing. The paper derives these results without any restrictions on the available contract space, the distribution function governing a project?s payoff, or the risk aversion of most potential entrepreneurs. Linear securities derive from the assumption that small business financing involves more uncertainty than is captured in most financial models. Instead of assuming that business people are faced with a given menu of projects, the model allows entrepreneurs to create (over time) an unlimited number of non-positive net present value projects with any payoff distribution they desire. Also, outside investors cannot observe project choice but only terminal cash flows. As a result, suppliers of funds must design contracts so that in equilibrium entrepreneurs do not wish to undertake undesirable investments. Further analysis of the model shows that in equilibrium entrepreneurs must contribute some of their own capital. The paper also finds that when a firm does not have any collateralizable assets, the equilibrium funding agreements have the property that the investors split the project?s proceeds in proportion to their initial investments. We further demonstrate that the existence of an infinite number of non-positive NPV projects can lead in equilibrium to positive abnormal returns earned by outside financiers. The model also produces a pecking order theory of financing. It is shown that highly profitable firms, and/or those companies run by relatively risk neutral individuals will first try to raise funds with riskless debt, and then turn to equity only when the supply of riskless debt has been exhausted. Finally, the paper shows that the linearity of the equilibrium contracts derives from the entrepreneur?s discretion with regard to the payoff distribution, rather than the flexibility associated with non-positive NPV projects. As part of the same analysis, we also demonstrate that under some conditions our underpricing results can be derived solely from the infinite supply of non-positive NPV projects.

Abstract:
This paper models the interactions among product market innovation, product market competition, and corporate financing decisions in the context of a dynamic duopoly. One competitor faces an opportunity to adopt a new technology. If adopted, the firm must also determine whether it will obtain public or private financing. Our results allow us to relate current firm and industry characteristics to these decision variables. In particular, larger, more profitable firms with small rivals have the greatest incentive to innovate. The private versus public financing decision depends mainly on the magnitude of the technological improvement and length of the period during which private financing extends the innovator's product market advantage. Due to the model's formulation it is both tractable and amenable to empirical estimation. We estimate the model and provide estimates of the value of innovation and private financing for a sample of industries and firms.

stragetic financing, finance, market competition, corporate financing

Abstract:
It is now well known that the Sharpe ratio and other related reward-to-risk measures may be manipulated with option-like strategies. In this paper we derive the general conditions for achieving the maximum expected Sharpe ratio. We derive static rules for achieving the maximum Sharpe ratio with two or more options, as well as a continuum of derivative contracts. The optimal strategy rules for increasing the Sharpe ratio. Our results have implications for performance measurement in any setting in which managers may use derivative contracts. In a performance measurement setting, we suggest that the distribution of high Sharpe ratio managers should be compared with that of the optimal Sharpe ratio strategy. This has particular application in the hedge fund industry where use of derivatives is unconstrained and manager compensation itself induces a non-linear payoff. The shape of the optimal Sharpe ratio leads to further conjectures. Expected returns being held constant, high Sharpe ratio strategies are, by definition, strategies that generate regular modest profits punctunated by occasional crashes. Our evidence suggests that the 'peso problem' may be ubiquitous in any investment management industry that rewards high Sharpe ratio managers.

Abstract:
We investigate a general multiple security equilibrium model in which firms adjust their capital stock in response to economic shocks. Asset values are determined by competitive risk-averse investors. When corporate capital increases in value, firms react by creating more of it. This leads to additional risk that must be borne by investors. Overall, the model generates a VAR(1) structure for the state variables determining the cross-section of expected returns, and is broadly consistent with stylized facts (e.g., the value premium, size premium, earnings momentum, and investment premium). In addition, the paper tests a new prediction of the model and finds support for it in the data.

capital investment, profitability of capital, equity issuance, expected return, overlapping generations model, supply shocks

Abstract:
This paper examines governance explanations for the discount of preferred shares to common shares in the Russian market. conflicts between shareholder classes may help explain the discount. However, for this to be the sole explanation the estimated models suggest that the magnitude of future adverse shareholder events would have to be very high. Nevertheless, evidence of a common factor potentially related to governance seems evident in the date, implying that corporate control issues may at least be partially responsible for the observed preferred share discount

Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

Abstract:

Abstract:
Prior to the announcement of a tender offer, the bidding firm is legally allowed to acquire shares in the open market, subject to some limitations. These pre-announcement purchases are known as toeholds. This paper presents a simple model that describes the bidder's optimal toehold acquisition strategy, within an environment that closely parallels the present legal institutions. The model shows that toeholds and bids interact in a complex manner even without the presence of asymmetric information. By examining a simple environment the paper provides a useful alternative hypothesis for tests of other, presumably more complex, models. One of the main implications of our model is that if no competing bidders are expected, no toeholds should be purchased. the paper demonstrates that the correct specification of an empirical model can be critical. For example, under some parameter values toehold purchases may exhibit a negative cross-sectional correlation with the pre-announcement run up in the stock price. This occurs even though prices are strictly increasing the size of the toehold. Several implications concerning various aspects of merger legislation are considered. We show that corporate charters that affect the number of shares necessary to complete a merger will have an impact only if competition among bidders is expected. The paper further shows that a rule similar to a 'fair price' provision has the desirable property that a second bidder arrives and winds if and only if he places a higher value on the target than the initial bidder. Several additional comparative statics are derived as well.

Abstract:
The Review of Financial Studies has among its missions the facilitation and promotion of a vigorous academic debate across unsettled questions in finance. This issue represents a cross section of views regarding one such debate: Can ourempirical models accurately forecast the equity premium any better than the historical mean? Or, is the forecast our empirical models give us any more accurate than what we would get by simply using the historical mean?

Abstract:
Numerous measures have been proposed to gauge the performance of active management. Unfortunately, these measures can be gamed. Our article shows that gaming can have a substantial impact on popular measures even in the presence of high transactions costs. Our article shows there are conditions under which a manipulation-proof measure exists and fully characterizes it. This measure looks like the average of a power utility function, calculated over the return history. The case for using our alternative ranking metric is particularly compelling for hedge funds whose use of derivatives is unconstrained and whose managers' compensation itself induces a nonlinear payoff.

G11, G23, G24

Abstract:
When the imminence of news announcements is not public knowledge, many traders will lack information on both the mean and variance of private information. Our analysis of such a setting in both single and multi-security contexts implies that disclosure of impending information events by firms can bound variance uncertainty and thereby improve investor welfare by mitigating the market breakdown problem. We also find that the equilibrium pricing functions are non-linear; specifically, convex for small trades and concave for larger ones. In addition, we predict that large transactions will be followed by large levels of volatility.

Abstract:
A number of empirical studies have reached the conclusion that stock price volatility cannot be fully explained within the standard dividend discount model. This paper proposes a resolution based upon a model that contains both a random supply of risky assets and finitely lived agents who trade in a multiple security environment. As the analysis shows there exist 2^K equilibria when K securities trade. The low volatility equilibria have properties analogous to those found in the infinitely lived agent models of Campbell and Kyle (1991) and Wang (1993, 1994). In contrast, the high volatility equilibria have very different characteristics. Within the high volatility equilibria very large price variances can be generated with very small supply shocks. Adding securities to the economy further reduces the required supply shocks. Using previously established empirical results the model can reconcile the data with supply shocks that are less than 10% as large as observed return shocks. These results are shown to hold even when the dividend process is mean reverting.

Abstract:
This paper provides a method for estimating housing indicesat the local level. It develops a "distance-weightedrepeat-sale" procedure to exploit the factor structure ofthe error-covariance matrix in the repeat-sales model. Adistance function defined in characteristic andgeographical space provides weights for the generalizedleast-squares model, and allows the use all of the repeated-sales in a metropolitan area to measure returns for thespecific neighborhood of interest. We use distance-weightedrepeat-sales to estimate return indices for all zip codes inthe San Francisco Bay area over the period 1980 through1994. When distance is defined in terms of socio-economiccharacteristics, we find that median household income is thesalient variable explaining covariance of neighborhoodhousing returns. Racial composition and educationalattainment, while significant, are much less influential.Zip-code level indices often deviate dramatically from thecity-wide index, depending upon income levels. This hasimplications for investors and lenders. Our resultsindicate that rates of return may vary considerably within ametropolitan area. Thus, simply using broad metropolitanarea indices as a proxy for capital appreciation within aspecific neighborhood may not be justified.

Abstract:
This paper addresses the issue of how closely the fortunes of suburbs are tied to the fortunes of the central city. We use similarities in residential housing price dynamics as a measure of how closely the economies of cities and suburbs are related. We develop housing price indices for most of the zip codes in California, and use these in a clustering procedure to see whether cities and suburbs naturally aggregate together, or whether they move separately. We find that central cities tend to aggregate with their suburbs suggesting that the fortunes of the cities and suburbs are closely linked. We find evidence of a two-tiered structure to suburbs. While nearby suburbs continue to group with their metropolitan core, extended suburban economies comprised mostly of affluent neighborhoods, may be "drifting away" from the central cities.