Technology and Cost Efficiency in Universal Banking A “Thick Frontier”-Analysis of the German Banking Industry

Using 1992 data of 1490 banks covering about 40% of German banking, we specify a multi-product translog cost function and follow the thick frontier-approach to control for cost inefficiency when evaluating the technology of banking. Scale economies are found to exist up to a size of about 5 billion DM of total assets, with diseconomies being caused by non-operating costs. There is hardly any evidence of economies of scope. Compared to cost inefficiency external factors play a surprisingly strong role in explaining cost differences between high-cost and low-cost banks. Smaller banks turn out to be more responsive to input prices.     

An Efficiency Comparison of European Banking Systems Operating under Different Environmental Conditions

The paper investigates the operating efficiency differences of a sample of commercial banks across 10 European countries. First, the paper analyzes the technical efficiency of each country sample following the basic Data Envelopment Analysis (DEA) model incorporating only banking variables. Then, a complete DEA model is introduced, incorporating environmental factors together with the banking variables of the basic model. The comparison between the two models shows that country-specific environmental conditions exercise a strong influence over the behavior of each country's banking industry.     

Index numbers and demand functions

The article proposes a method of approximating the Konüs true cost of living index, given a system of integrable demand functions. It is shown that the approximation is contained within the Staehle bounds so long as the Engle curves satisfy a weak monotonicity condition. The result is likely to be most useful where the demand functions are known-and known (or constrained) to be integrable-but the solution for the form of the direct or indirect utility function is unknown.     

Die asymptotische Verteilung des Prognosefehlers bei Prognosen mit Hilfe eines dynamischen ökonometrischen Modells höherer als erster Ordnung

Zusammenfassung In dieser Arbeit wird die asymptotische Verteilung des Prognosefehlers, wie er sich im Rahmen einer dynamischen Simulation eines allgemeinen autoregressiven ökonometrischen Modells der Ordnungp (einschließlich verzögerter exogener Variabler der Ordnungq) ergibt, abgeleitet. Daran anschließend werden einige Fragestellungen, die damit in unmittelbarem Zusammenhang stehen, diskutiert: Die Frage der relativen Effizienz der Prognoseschätzung, basierend auf der unrestricted- bzw. der derived reduced form, die Verwendung der asymptotischen Verteilung des Prognosefehlers für einen predictive test des Modells. Außerdem werden asymptotische simultane Prognoseintervalle abgeleitet.

---

Summary The asymptotic distribution of the forecast error in the dynamic simulation of a higher than first order linear dynamic econometric model is derived and related topics are discussed.

     

Una luce nuova su una vecchia storia: La “scindibilitá” di Cantelli-insolera e la struttura a termine dei tassi d'interesse

Si mostra come la moderna teoria della struttura a termine dei tassi d'interesse consenta di definire diversi esempi (più o meno naturali) di leggi di capitalizzazione a tre variabili, e dunque sostanzialmente del tipo di quelle considerate da Insolera. La relativa condizione di scindibilità, già oggetto di polemica con Cantelli, appare formulabile in più di una versione, una delle quali equivale alla odierna ipotesi di equilibrio del mercato finanziario.

---

Summary The modern theory of term structure of interest rates leads one, in a natural way, to consider several models of accumulation factors which depend on three variables, and can therefore be counted among those considered in this time by the Italian finance mathematician F. Insolera. The question of how to formulate the principle of consistency for such factors has been the subject of an animated debate between Insolera and F.P. Cantelli. The present paper shows that it admits more than one solution; one of the possible versions rendering the principle equivalent to the market equilibrium condition.

     

Multiple patterns in the dynamics of a stock market model

We study a non-linear model of the interactions between stock market prices and the level of assets owned by investment funds. The model dynamics is described, in continuous time, by a smooth vector field in the plane, which presents, under suitable hypotheses, a unique equilibrium point.Our analisis of the system flow is qualitative and focuses on detecting endogenous fluctuations of the state variables, i.e. on checking existence and number of limit cycles.We prove that several, and quite different, dynamical patterns can occur, even in cases where the system isoclines assume that most simple geometrical forms.It is shown, in particular, that the equilibrium point can undergo either a sub-critical or a super-critical Hopf-bifurcation whenever two economically meaningful exogenous parameters are made to cross a given set of critical values. Hence, in the subcritical case, as a trapping region exists, at least two limit cycles appear.Next, we give analytical examples of model-consistent vector fields which present a multiplicity of fluctuating trends, and prove the apparently surprising result that the number of limit cycles can be as large as one wants, provided a specific isocline assumes a cubic shape.Both authors are members of the Gruppo Nazionale per l'Analisi Funzionale e le sue Applicazioni-G.N.A.F.A. of the Italian Council of Researches-CN.R.The present paper refers to the activities of the National M.U.R.S.T. Group Dinamiche Non Lineari ed Applicazioni alle Scienze Economiche e Sociali.     

Die Charakterisierung der Normalverteilung nach Gauß

A proof is offered for the best possible version of the following Gauss type characterization of normality: LetF () be a family of distribution functions with translation parameter such thatF (0) has a densityf with certain regularity properties, and letM{1, 2, ...}. If the mean of every sample of any sizemM is a maximum likelihood estimate of , thenF (0) is normal with zero expectation. While in the best prior version of this theorem,f satisfies a continuity assumption andM={2, 3}, here no regularity condition is needed, andM can be any of the sets {3}, {4}, ....     

Cost and technical change: Effects from bank deregulation

Banks were substantially deregulated during the 1980s. This altered the cost-minimizing mix between deposit interest payments and operating expenses (capital and labor for branch convenience and free deposit services). Measured bank output was relatively unaffected by these changes. The proportion of deposits receiving higher interest rates increased more rapidly than factor quantities were reduced, so unit deposit costs rose. Consumers benefited, but the measured net effects for banks were negative, averaging –0.8% to –1.4% a year in net technical change from 1977 to 1988. These influences were measured three different ways and for both equilibrium and disequilibrium factor input specifications. All three approaches—a standard time trend, a time-specific index, and shifts in cross-sectional cost functions—gave consistent results. The results were robust whether measured at the banking firm or branch office level or on the cost frontier.     

Detecting Outliers in Frontier Models: A Simple Approach

In frontier analysis, most of the nonparametric approaches (DEA, FDH) are based on envelopment ideas which suppose that with probability one, all the observed units belong to the attainable set. In these deterministic frontier models, statistical theory is now mostly available (Simar and Wilson, 2000a). In the presence of super-efficient outliers, envelopment estimators could behave dramatically since they are very sensitive to extreme observations. Some recent results from Cazals et al. (2002) on robust nonparametric frontier estimators may be used in order to detect outliers by defining a new DEA/FDH deterministic type estimator which does not envelop all the data points and so is more robust to extreme data points. In this paper, we summarize the main results of Cazals et al. (2002) and we show how this tool can be used for detecting outliers when using the classical DEA/FDH estimators or any parametric techniques. We propose a methodology implementing the tool and we illustrate through some numerical examples with simulated and real data. The method should be used in a first step, as an exploratory data analysis, before using any frontier estimation.     

Global dynamics in models of fluctuating growth Part II: Three dimensional systems

The dynamics of a model of fluctuating growth, where non-constant returns are allowed, is represented, under Goodwin's classical assumptions, by a non-autonomous two-dimensional system, which can be transformed into an autonomous three-dimensional one. We describe the global phase portrait of the latter, in the two cases of increasing and decreasing returns, proving, in particular, the absence of economically meaningful attractors. However the orbits exhibit different features in the two cases: namely they asymptotically converge to a singular point, where the economy dies, if the returns are decreasing, and diverge, spiralling around a certain line, if the returns are increasing.

---

Riassunto Un modello non-lineare di crescita alla Goodwin, in cui i rendimenti possono essere non costanti, è rappresentato da un sistema dinamico piano non autonomo, che può essere trasformato in un sistema autonomo tridimensionale. Viene descritto il comportamento globale di quest'ultimo nei due casi di rendimenti crescenti o decrescenti. Si dimostra, in particolare, l'assenza di attrattori economicamente significativi ed il diverso carattere delle orbite: mentre, se i rendimenti sono crescenti, le traiettorie convergono asintoticamente ad un punto singolare dove l'economia muore, nel caso opposto le traiettorie sono illimitate e spiraleggiano intomo ad una retta che ha la direzione dell'assek (la variabile di stato che rappresenta lo stock di capitale).

---

---

This paper relates to the activities of the M.P.I. Group Dinamiche Non-Lineari nelle Scienze Economiche e Sociali.     

“Good Cause Beliefs” in an “At-Will World”? A Focused Investigation of Psychological Versus Legal Contracts

This study assesses the psychological contracts of a group of at-will employees and compares their relevant psychological contract beliefs with the terms of the controlling legal employment contract. In addition, we test specific hypotheses regarding the relationship between employers formal job security policy (employment at-will vs. good cause protection, and employee personal characteristics; equity sensitivity, organizational tenure) to 2 focal psychological contract beliefs (the employers obligation to have a good reason to discharge the employee, and the employees reciprocal obligation to have a good reason to leave his/her employer). These issues are investigated using survey data sampled from 15 diverse U.S. organizations, and independently collected information regarding organizations job security policies. The results provide evidence of a widely shared psychological contract belief that, despite explicit at-will policies, U.S. employers are highly obligated to have a good reason to discharge employees. Further, personal characteristics are more strongly related to the focal psychological contract beliefs than employer job security policies. The theoretical contribution of the study and its practical implications are discussed.     

Erneuerungsprozesse mit Wartungsstrategien

Summary The following renewal process is considrred: given intervals (kt ₁,(k+1)t ₁],k=0, 1, 2, ..., 0<t ₁<, there will be with probabilityp, 0p1, a renewal in each interval at a time selected by random. The costs for each of this renewals are a units, while the costs of the other renewals areb units each. The renewal function and the cost function are derivided and their asymptotic behavior is discussed.     

DEA and Activity Planning under Asymmetric Information

The productivity analysis literature has traditionally focused on the evaluation of past performances. In this paper, we consider the post productivity analysis problem of deciding which production plans to choose in the future given information from a productivity analysis. In particular, we demonstrate that Data Envelopment Analysis (DEA) has an important role to play in the reallocation game following a normal productivity analysis. DEA estimates reduce the information rents an agent can extract by claiming high costs for the least reduced or most expanded activity. We also examine how to optimally combine DEA estimates with other information in a planning context, including preference information, ex-ante cost reports and ex-post cost data.     

Distinguishing external from internal scale effects: The case of public infrastructure

The recent applied production theory literature focusing on the economic performance of firms has increasingly recognized the importance of scale effects on costs and therefore efficiency. These scale effects may include short run returns due to fixity of privately demanded inputs (i.e., capital, long run internal returns to scale, and external factors affecting costs. Since these different types of scale effects can be thought of as shifts in and movements along cost curves, the different cost effects of such factors can be identified in a framework which explicitly takes them into account in the definition ofscale.In this article we formalize such a framework, and then use it to measure short run, long run (internal) and external scale effects from fixity of private capital, nonconstant returns to scale and public infrastructure. We then use these measures to identify the impacts of these different scale factors on productivity growth. The focus on public infrastructure as an important external scale factor is motivated by the current theoretical and policy interest in this issue; we show how a structural production theory model provides a rich basis for the analysis of the cost effects of infrastructure investment.     

A consistent and asymptotically efficient two-stage procedure to construct fixed width confidence intervals for the mean

For a normal distribution, a two-stage procedure has been proposed for constructing a fixed width confidence interval for the mean when the variance is unknown. It has all the properties ofStein's two-stage procedure, and at the same time it is asymptotically efficient. We have also discussed the non-normal case as inChow/Robbins [1965].     

Interaction of human and physical capital in a model of endogenous growth

A model of development is presented where growth is initially driven by physical capital accumulation, as in the neoclassical model. After a critical level of physical capital is reached, the economy takes off and enters a stage of sustained growth driven by human capital accumulation. The link between these two stages is provided by the assumption that private incentives for human capital accumulation increase with the average levels of human and physical capital. At the early stages of development, these incentives are low so the level of human capital stays stagnant until sufficient physical capital is accumulated. Other results are that some economies may reach a steady state of physical capital before a take-off is possible. This is especially likely for economies in which agents have low savings propensities. Such economies remain stuck in a no-growth equilibrium forever. Economies that do grow may experience endogenous cycles if the return to investment in human capital is sufficiently increasing in the level of physical capital.     

Bootstrapping Confidence Intervals for Linear Programming Efficiency Scores: With an Illustration Using Italian Banking Data

This article suggests a method for introducing a stochastic element into Farrell measures of technical efficiency as calculated via linear programming techniques. Specifically, a bootstrap of the original efficiency scores is performed to derive confidence intervals and a measure of bias for the scores. The bootstrap generates these measures of statistical precision for the nonstochastic efficiency measures by using computational power to derive empirical distributions for the efficiency measures.     

Law without justice: The dilemmas of formalization and fairness in the legalistic organization

There is growing evidence of a litigation mentality rampant in organizations, which has created a legalistic mindset in many managers. Increasingly, managerial decisions are becoming dominated by a concern for what is legally defensible at the expense of broader social considerations such as justice and fairness. The papers in this special issue explore how this legalistic mindset has created new organizational and social dilemmas, and situations of law whithout justice.     

Estimation problems for rectangular distributions (or the taxi problem revisited)

The problem of estimating the unknown upper bound on the basis of a sample of sizen from a uniform or rectangular distribution on [0, ] has considerable interest. This or the analogous discrete version is variously known as the Taxi-problem or the German bomb (or Tank) problem and has a long history. The emphasis here is on estimation of through the lengths of the observed gaps or spacings which seem natural for this problem.     

Output scaling in a cost function setting

A significant result in production theory is that even under a very weak condition on the technology, the cost function is well defined and linearly homogeneous in factor prices. We could term this homogeneity a scaling law, for an equiproportionate scaling of all factor prices results in an identical scaling of the value of the cost function. The purpose of this paper is to determine sets of conditions under which an analogous result may hold for the output variable in the cost function. Our analysis initially focuses on the cost function of a multioutput technology so that the output variable in the cost function is vector-valued. The results we obtain are that some output-scaling laws exist if and only if the underlying technology is ray-homothetic. In the final section we consider the meaning of the results for the single-output case.The refereeing process of this paper was handled through R. Robert Russell. An earlier version of this paper was presented at the 1989 Conference on Current Issues in Productivity, December 6–8, 1989, Rutgers University, Newark, NJ.