Bank- und finanzwirtschaftliche Forschungen Band 395 Institut für schweizerisches Bankwesen der Universität Zürich Schweizerisches Institut für Banken und Finanzen an der Universität St. Gallen
Florentina Paraschiv Modeling Client Rate and Volumes of Non-maturing Savings Accounts Haupt Verlag Bern Stuttgart Wien
Dr. Florentina Paraschiv works as Assistant Professor at University of St. Gal - len, School of Finance. Focusing on financial econometrics modeling, her research fields include: liquidity risk, portfolio risk management as well as energy markets modeling. In 2011 she obtained the PhD in Management with Emphasis in Finance from University of St. Gallen, ior/cf, focusing on the Non-maturing Accounts, under the supervision of Prof. Dr. Karl Frauendorfer and Prof. Paul Söderlind PhD. Redaktion und Satzherstellung durch die Autorin Zugleich Dissertation an der Universität St. Gallen 1. Auflage: 2011 Bibliografische Information der Deutschen Nationalbibliothek Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über http://dnb.d-nb.de abrufbar. ISBN 978-3-258-07706-2 Alle Rechte vorbehalten. Copyright 2011 by Haupt Berne Jede Art der Vervielfältigung ohne Genehmigung des Verlages ist unzulässig. Printed in Switzerland www.haupt.ch
Acknowledgements I have written this thesis during the three years as assistant at the Institute for Operations Research and Computational Finance (ior/cf-hsg), University of St. Gallen. I would like to thank Prof. Dr. Karl Frauendorfer for the great chance he offered me three years ago to write my PhD inside his Institute and under his supervision. I am very grateful for his valuable support with the useful discussions in the key moments of my dissertation, for his advice and for his continuous confidence. My gratitude goes also to Prof. Paul Söderlind PhD. from the Swiss Institute of Banking and Finance (s/bf-hsg), who agreed to act as co-referee of my thesis. I particularly thank him for the very useful advise and constructive discussions concerning the econometrical methods applied in the thesis. I thank very much also my colleagues for sharing their valuable experience with me. A special thanks goes to Michael Schürle, who introduced me into the problematic of non-maturing accounts and from whose experience I learned very much. Furthermore, I thank him and Elisabeth Disch for their permanent moral support and confidence. Finally, I thank my family for their understanding and moral support and very special thoughts go to my grandfather... St. Gallen, March 2011 Florentina Paraschiv
Contents Acknowledgements Contents List of Figures List of Tables V VII XI XV Summary 1 Zusammenfassung 3 Motivation 5 1 Literature overview client rate 7 1.1 Client rate characteristics.................... 7 1.2 Client rate modeling approaches................ 9 1.3 Summary of the literature overview.............. 11 2 Base models for the investigations of the client rate adjustment 17
VIII Contents 2.1 Simple and multiple regression models............ 17 2.1.1 Empirical results.................... 19 2.1.2 Discussion about the statistical properties of the model 21 2.1.3 Economical Interpretation............... 24 2.2 Cointegration analysis. VEC model. Johansen cointegration test................................. 24 2.2.1 Johansen s approach.................. 27 2.2.2 Estimation results.................... 28 2.3 Error correction model..................... 30 2.3.1 Empirical Results.................... 31 2.3.2 Economical Interpretation............... 33 3 Modeling the asymmetries of the client rate adjustment 35 3.1 Reasons for asymmetries and drivers of different types of asymmetries in the literature.................. 35 3.2 Our approach versus the existing literature.......... 37 3.3 Methodology........................... 38 3.3.1 Threshold model - general specification........ 38 3.3.1.1 Hansen s grid search to locate the most likely threshold................... 39 3.4 Empirical results - application on SNB average client rate data 40 3.4.1 Interpretation of results................ 51 3.5 Empirical results - robustness test............... 52 3.6 Overall assessment....................... 55 4 Modeling the rigidity of the client rate 59 4.1 Reasons for the rigidity of client rate adjustments...... 59 4.2 Application of the friction model to savings accounts.... 61 4.2.1 Extension of the friction model............ 62 4.2.2 Estimation procedure.................. 63 4.2.3 Empirical Results.................... 65 4.3 Robustness test......................... 65 4.4 Overall assessment....................... 67 5 Scenario generation based on the client rate models 69 6 Literature overview. Similarities and differences in methodology 77 7 Forecasting model for the volumes of Non-Maturing Accounts 81
Contents IX 7.1 Underlying data for an empirical investigation........ 81 7.2 Modeling approach....................... 83 7.3 Empirical results Swiss Market................. 84 7.4 Empirical results - Robustness test.............. 85 7.5 Including macroeconomic variables in the forecasting model 88 7.5.1 Choice of macroeconomic variables.......... 88 7.5.2 Test for omitted variables............... 88 7.6 Model assessment out-of-sample test............ 89 8 Scenario generation based on the volumes model 93 9 Volumes model - VAR approach 101 9.1 VAR model: Theoretical background............. 101 9.1.1 Stable VAR....................... 101 9.1.2 Impulse response function............... 103 9.1.3 Economical interpretation of VAR models...... 104 9.2 Model specification and empirical results........... 105 9.3 Impulse-response function................... 109 9.4 Overall assessment....................... 110 10 Conclusion 117 References 119