Vedat Akgiray

The Stable-Law Model of Stock Returns

This study investigates the tail shapes of empirical distributions of returns on an extensive group of common stocks. The tails of the return distributions are found to be thinner than those of infinite variance stable distributions. Therefore, although homogeneity is evident in general, economic and statistical inferences drawn from stable-law parameters estimated from samples of stock returns may be misleading. This is in spite of the apparent overall similarity (in shape) between empirical and stable distributions.

Mixed Diffusion-Jump Process Modeling of Exchange Rate Movements

This study demonstrates that the mixed diffusion-jump process is superior to the stable laws or a mixture of normals as a model of exchange rate changes for the British pound, French franc, and the We st German mark relative to the United States dollar. The parameter value s for the mixed diffusion-jump process are dependent on the monetary policy regime in force in the United States, with the estimates for the franc and mark being intertemporally similar but different from the pound. Copyright 1988 by MIT Press.

Estimation of Stable-Law Parameters: A Comparative Study

The stable distribution has many desirable properties and is applicable in many areas of scientific pursuit (e.g., the study of stock-return behavior). Despite this, little is known about the properties of the various extant estimation techniques for the parameters of the stable laws. This article compares the iterative regression technique with the latest version of the fractile technique, using both simulated and actual data.

Distribution properties of Latin American black market exchange rates

Abstract This paper conducts a statistical analysis of the distribution properties of black market exchange rate changes for 12 Latin American currencies. The infinite variance stable law hypothesis is tested using a regression- type technique to estimate the parameters of the stable Paretian distribution. Strong evidence is found supporting this hypothesis. A separate analysis of extreme observations using a robust statistical procedure further supports these findings.

A causal analysis of black and official exchange rates: The turkish case

ZusammenfassungEine Kausalanalyse von Schwarzmarktkursen und offiziellen Wechselkursen: Der Fall der Türkei. - Diese Arbeit untersucht statistisch den Kausalzusammenhang zwischen Schwarzmarkt- und offiziellen Wechselkursen für das türkische Pfund, und zwar im VerhÄltnis zum US-Dollar und zur D-Mark in der Zwei-Jahres-Periode ab September {dy1985}, wobei Granger-Techniken benutzt werden. TÄgliche Änderungen des Schwarzmarktkurses für die D-Mark gehen den tÄglichen Änderungen des offiziellen Kurses voraus. Bei wöchentlichen und monatlichen Änderungen liegt eher eine Gleichzeitigkeit vor. Die Ergebnisse für den Schwarzmarktkurs und den offiziellen Dollar-Kurs sind Ähnlich, obwohl es Hinweise für einen schwachen Feedback gibt, wenn Tagesdaten benutzt werden. Diese Befunde stützen die These, da\ SchwarzmÄrkte Informationen effizient verwerten.RésuméUne analyse de causalité pour des taux de change noirs et officiels: le cas de la Turquie. - Cet article examine statistiquement la relation causale entre les taux de change noirs et officiels pour la lira turque vis-à-vis le dollar américain et la mark allemande pour une période de deux ans qui commence en septembre 1985. L’auteur applique des techniques de type Granger. Les changements quotidiens sur le marché noir pour la mark allemande précèdent ceux sur le marché officiel. Une relation simultanée existe pour les changements hebdomadaires et mensuels. Les résultats pour le dollar noir et officiel sont similaires bienqu’il y ait quelque évidence pour une répercussion faible si les données quotidiennes sont utilisées. Les résultats supportent la notion que les marchés noirs sont des processeurs efficients des informations.ResumenUn análisis de causalidad de tipos de cambio oficial y paralelo: el caso turco. - En este trabajo se examina estadísticamente la relación causal entre los tipos de cambio paralelo y oficial de la lira turca con respecto al dólar EE.UU. y el marco alemán para un período de dos años, comenzando en setiembre de {dy1985} y utilizando técnicas de tipo Granger. Los movimientos diarios en el mercado paralelo del marco dan lugar a movimientos diarios en el mercado oficial. Existe una relación contemporánea en los movimientos semanales y mensuales. Los resultados para el dólar paralelo y oficial son similares, a pesar de la evidencia de una respuesta débil en el caso de utilizarse datos diarios. Los resultados apoyan la noción de que los mercados paralelos son eficientes procesadores de información.

Statistical models of German stock returns

examples. Similar findings regarding German stock returns are provided by Conrad and Jiittner (1973), Ronning (1974), Muhlbradt (1978), and Moller (1986), among others. For American stocks several suggestions have been offered to explain the empirical distributional shape. For instance, Mandelbrot (1963) and Fama (1965) suggest that the explanation may be found in the stable laws. Others indicate that mixture models may be more appropriate. The two most common candidates of this type of model are a finite mixture of normal distribu

Knowledge representation for investment strategy selection

A hybrid knowledge representation for the problem of investment-strategy selection is described. This representation scheme is efficient in knowledge organization and inference performance. The types of knowledge required to support the selection process are identified. The interrelationships among these different types of knowledge are represented in a frame abstraction, and a frame-slot structure in clausal form is described. Knowledge acquisition, consistency, and completeness checks are shown in terms of sample Prolog rules.<<ETX>>