Refine
Year of publication
- 1999 (14) (remove)
Document Type
- Report (14) (remove)
Language
- English (14) (remove)
Has Fulltext
- yes (14)
Is part of the Bibliography
- no (14)
Keywords
Institute
- Informatik (3)
- Center for Financial Studies (CFS) (1)
- Extern (1)
- Mathematik (1)
- Physik (1)
- Wirtschaftswissenschaften (1)
Collateral, default risk, and relationship lending : an empirical study on financial contracting
(1999)
This paper provides further insights into the nature of relationship lending by analyzing the link between relationship lending, borrower quality and collateral as a key variable in loan contract design. We used a unique data set based on the examination of credit files of five leading German banks, thus relying on information actually used in the process of bank credit decision-making and contract design. In particular, bank internal borrower ratings serve to evaluate borrower quality, and the bank's own assessment of its housebank status serves to identify information-intensive relationships. Additionally, we used data on workout activities for borrowers facing financial distress. We found no significant correlation between ex ante borrower quality and the incidence or degree of collateralization. Our results indicate that the use of collateral in loan contract design is mainly driven by aspects of relationship lending and renegotiations. We found that relationship lenders or housebanks do require more collateral from their debtors, thereby increasing the borrower's lock-in and strengthening the banks' bargaining power in future renegotiation situations. This result is strongly supported by our analysis of the correlation between ex post risk, collateral and relationship lending since housebanks do more frequently engage in workout activities for distressed borrowers, and collateralization increases workout probability.
Following on the ADEA/APNET study on inter-African Book trade that was commissioned in 1999, ADEA tasked APNET to facilitate the production of national book industry updates in each country. The updates are aimed at encouraging commercial development of inter-African book trade and to make available to the public, total systematic and current situations on the book trade in each country.
Languages vary in whether or not primary grammatical relations (PGRs) are sensitive to information from clause-level case or phrase structures. This variation correlates with a difference between verb agreement systems based on feature unification and systems based on feature composition. The choice between different PGR and agreement principles is found to be highly stable genetically and to characterize Indo-European as systematically different from Sino-Tibetan. Although the choice is partially similar to the Configurationality Parameter, it is shown that Indo-European languages of South Asia are nonconfigurational due to areal pressure but follow their European relatives in PGR and agreement principles.
Ever since Wilhelm von Humboldt’s (1836) pioneering study of Nahuatl, linguists have recurrently recognized that languages differ fundamentally in the syntactic weight they attribute to noun-phrases as the arguments of a verb. Currently, the most prominent attempts to turn this intuition into a precise hypothesis revolve around the notion of ‘configurationality’.
In the following I will discuss grammatical structures of Inuktitut, an Eskimo language spoken in the Canadian Eastern Arctic. Inuktitut is a polysynthetic language exhibiting an exceedingly elaborate verbal inflectional system including polypersonal marking. Furthermore, Inuktitut features free word order and optionality of noun phrases crossreferenced with the predicate. But Inuktitut also exhibits a number of features which seem to contradict the possibility of its being a "pronominal argument language" -- or as I would prefer to express it, a morphological argument language.
For the efficient management of large image databases, the automated characterization of images and the usage of that characterization for searching and ordering tasks is highly desirable. The purpose of the project SEMACODE is to combine the still unsolved problem of content-oriented characterization of images with scale-invariant object recognition and modelbased compression methods. To achieve this goal, existing techniques as well as new concepts related to pattern matching, image encoding, and image compression are examined. The resulting methods are integrated in a common framework with the aid of a content-oriented conception. For the application, an image database at the library of the university of Frankfurt/Main (StUB; about 60000 images), the required operations are developed. The search and query interfaces are defined in close cooperation with the StUB project “Digitized Colonial Picture Library”. This report describes the fundamentals and first results of the image encoding and object recognition algorithms developed within the scope of the project.
The prevention of credit card fraud is an important application for prediction techniques. One major obstacle for using neural network training techniques is the high necessary diagnostic quality: Since only one financial transaction of a thousand is invalid no prediction success less than 99.9% is acceptable. Due to these credit card transaction proportions complete new concepts had to be developed and tested on real credit card data. This paper shows how advanced data mining techniques and neural network algorithm can be combined successfully to obtain a high fraud coverage combined with a low false alarm rate.