Data Aggregation Issues for Crop Yield Risk Analysis

With increased emphasis on risk management in agriculture and a lack of disaggregated or farm‐level yield time series, decision makers are often faced with having to make adjustments to temporal yield risk measures obtained from readily available but aggregated yield data. This paper provides some empirical evidence on what type of aggregation bias to expect when measuring temporal yield risk using yield observations averaged across a region relative to yield risk estimated from quarter‐section yield time series in wheat. This study highlights some of the challenges faced when estimating aggregation distortions in measuring yield risk defined by temporal variance, especially given the nature of the empirical data set used. Cluster analysis, visual examination of relative frequency distributions and mapping of yield risk clusters suggest that using a readily available, aggregate temporal yield risk measure has the tendency to underestimate yield risk observed at the quarter‐section level and that clear, geographic yield risk boundaries do not exist in municipalities or across larger areas in this study. Further research on crops more risky than wheat appears promising. Avec un plus grand intéret sur la gestion du risk dans l'agriculture et un manque de données détaillees ou bien de collections de séries temporelles sur les rendements, les décideurs sont souvent tenus d'apporter des correctifs aux measures du risk obtenues a partir des données de rendements qui sont disponibles. Cet artcle apporte une preuve empirique du type de biais lie a l'agrégation qui peut être présent dans le calcul du risk de rendement temporel obtenu a partir de rendements moyens de blé observés au niveau régional en comparaison du risk de rendement qui est estimé a partir de données basées sur des quart‐de‐sections. Cette étude met en exergue quelques uns des obstacles qui se présentent dans l'estimation de distosions liées a l'aggrégation dans le calcul du risk de rendement défini par la variance temporelle, speciallement étant donne la charactère empirique des données utilisées. L'analyse de groupe, l'examen visual de la distribution des fréquences relatives, et la cartographie de classes de risk de rendement suggèrent que l'utilisation de la measure du risk de rendement basée sur des données disponibles de risk aggrège temporel a tendence a sousestimer le risk de rendement observe au niveau des quart‐de‐sections et qu'il n'y a pas de frontières de risk de rendement certaines, géographiques qui existent entre les municipalités ou bien a travers les zones plus larges examinées dans cette etude.     

Intermarket efficiency: An application of interbattery APT to mortgage-backed securities

Increasing popularity of investments in mortgage-backed securities has led to closer integration of the mortgage market into traditional capital markets. Using monthly returns during 1982–1988 for common stocks, Treasury bonds and GNMA and FHLMC mortgage-backed securities, the interbattery factor analytic Arbitrage Pricing Theory of (Cho, 1984) is used to test five hypotheses for intramarket and intermarket integration. Results indicate that three to five common factors are found within the same security market, while only one to three factors are found common between different markets.The APT could not be rejected within the same security market, but was rejected in most intermarket comparisons. While risk-free rates are found to differ between markets, the risk premium tests are conclusive indicators of integration. Our results support claims that the stock, bond, and the mortgage-backed securities markets are integrated.     

Monetising the Classical Equations: a theory of circulation

The Classical Equations describe output and income in real terms.To use them to analyse aggregate demand, the transactions theydescribe must be ‘monetised’. A sum of money equalto the wage bill of the capital goods sector can be shown tobe necessary and sufficient to carry out all transactions, ina process of circulation which also defines an expression forvelocity. When money has intrinsic value, the quantity approachmay hold in the short run but, in the long run, money will beendogenous. In these conditions, the rate of interest will bedetermined by the supply and demand for reserves, but when moneyis purely nominal, only a minimum rate will be fixed, and therate of interest will have to be pegged. The Appendix developsthe Classical Equations and shows that they define an invariableunit of account.     

A divisia version of the country-product-dummy method

A Divisia version of the country-product-dummy method of Kravis et al. is obtained and extended to include income as an explanatory variable. Income elasticities of relative prices for eight goods across 15 countries are obtained and compared with the income elasticities of the demand for these goods.