Journal of Economics and Business 91 (2017) 1–15 Contents lists available at ScienceDirect Journal of Economics and Business Interest on reserves, regime shifts, and bank behavior Donald H. Dutkowsky a,* , David D. VanHoose b a Maxwell School of Citizenship and Public Affairs, Syracuse University, 110 Eggers Hall, Syracuse, NY 13244-1090, United States b Department of Economics, Baylor University, One Bear Place #98003, Waco, TX 76798, United States a r t i c l e i n f o Article history: Received 1 December 2016 Accepted 31 March 2017 Available online 8 April 2017 JEL classification: E58 G21 Keywords: Interest on reserves Federal funds rate Bank lending Monetary policy a b s t r a c t This paper demonstrates that in the post-2008 environment with interest on reserves, monetary policy actions can generate regime shifts that yield quantitatively and even qual- itatively different responses of bank balance-sheet configurations and loan and deposit market outcomes to exogenous changes. In contrast to the view that a one-time structural change occurred in 2008, switching between several different regimes plausibly can arise depending upon settings of the reserve ratio, federal funds rate, and the interest rate on reserves. Our results explain stylized facts regarding excess reserves and interbank lending. Analysis with calibrated values indicates that such regime switching has occurred. © 2017 Elsevier Inc. All rights reserved. 1. Introduction As discussed by Taylor (2016a), the high excess reserves holdings in the post-2008 financial crisis period stands as one of the most notable and important policy issues in U.S. banking. Depicted in Fig. 1A, prior to 2008 aggregate excess reserves of U.S. depository institutions as a percentage of their total assets were near-zero. But from October 2008 onward, excess reserves increased sharply and continued to rise well after the end of the financial crisis. Many believe that this post-2008 increase in excess reserves arose simply from large-scale Federal Reserve bond purchases, associated with quantitative easing, that also occurred during this time. Indeed, the monetary base nearly quadrupled from October 2008 until December 2015, from approximately $1 trillion to nearly $4 trillion. One might conclude that the banking system just absorbed these substantial injections of reserves by holding excess reserves. Closer examination, though, indicates that high excess reserves reflects a fundamental change in bank behavior between the pre-crisis and post-crisis periods. Before October 2008, the Fed also conducted bond purchases and injected new reserves steadily over time, albeit not to the same magnitude. Still, from January 1997 to September 2008 the monetary base nearly doubled, from $463 billion to $910 billion. Yet the share of bank assets held as excess reserves remained near zero throughout. The explanation for this behavior in the pre-crisis period is standard. During this time, the banking system responded to injections of new reserves by lending in retail markets, with bank deposits and M2 increasing commensurately. Individual banks with surplus reserves loaned them to borrowing banks in the wholesale markets. As a result, these reserves were distributed across the fractional-reserve banking system, until this injection ultimately became an addition to aggregate required reserves. This logic raises a question: Why did banks change their behavior after 2008 to hold high excess reserves? * Corresponding author. E-mail addresses: dondutk@maxwell.syr.edu (D.H. Dutkowsky), David VanHoose@baylor.edu (D.D. VanHoose). http://dx.doi.org/10.1016/j.jeconbus.2017.03.003 0148-6195/© 2017 Elsevier Inc. All rights reserved.