پول، ذخایر و انتقال سیاست پولی: آیا پول چند برابرشده وجود دارد؟

With the use of non-traditional policy tools, the level of reserve balances has risen in the US from roughly $20 billion before the financial crisis to well past $1 trillion. The effect of reserve balances in macroeconomic models often comes through the money multiplier, affecting the money supply and the bank lending. In this paper, we document that the mechanism does not work through the standard multiplier model or the bank lending channel. If the level of reserve balances is expected to have an impact on the economy, it seems unlikely that a standard multiplier story will explain the effect.

A second issue involves the effect of the large volume of reserves created as we buy assets. […] The huge quantity of bank reserves that were created has been seen largely as a byproduct of the purchases that would be unlikely to have a significant independent effect on financial markets and the economy. This view is not consistent with the simple models in many textbooks or the monetarist tradition in monetary policy, which emphasizes a line of causation from reserves to the money supply to economic activity and inflation… [W]e will need to watch and study this channel carefully. Donald L. Kohn, Vice Chairman of the Federal Reserve Board, March 24, 2010 The Federal Reserve’s implementation of a range of non-traditional monetary policy measures to combat a severe financial crisis and a deep economic recession resulted in a very large increase in the level of reserve balances in the US banking system. As a result, there has been renewed interest in the transmission of monetary policy from reserves to the rest of the economy. Since the 1980s, two broad transmission mechanisms have been discussed: an “interest-rate” or “money channel,” in which interest rates adjust to clear markets and influence borrowing and lending behavior; and a “credit channel,” in which the quantity as much as the price of loanable funds transmit monetary policy to the economy. Within the credit channel literature, a narrow “bank lending channel” view of the world follows the textbook money multiplier taught in undergraduate textbooks and suggests that changes in open market operations and the quantity of reserves directly affect the amount of lending that banks can do. A textbook money multiplier and the bank lending channel imply an important role for money in the transmission mechanism. In the past couple of decades however, New Keynesian models used for macroeconomic policy analysis have excluded money. The exceptions in this class of models are those where a money demand equation is appended, and the quantity of money is entirely endogenously determined with no feedback to real variables. This extreme marginalization of money is not universal, however. Some researchers for example, Hafer et al., 2007, Leeper and Roush, 2003, Ireland, 2004 and Meltzer, 2001 and most notably, the European Central Bank, put serious weight on the role of money in the macroeconomy and policy analysis. At the other extreme, many economics textbooks and some academic research, such as Freeman and Kydland (2000) or Diamond and Rajan (2006) continue to refer to the very narrow money multiplier and accord it a principle role in the transmission of monetary policy. Indeed, M1 multiplier is published as a regular statistic by the Federal Reserve Bank of St. Louis.1 The recent rise in reserve balances suggests a need to reassess the link from reserves to money and to bank lending. We argue that the institutional structure in the United States and empirical evidence based on data since 1990 both strongly suggest that the transmission mechanism is inconsistent with the conceptual framework of the standard money multiplier model from reserves to money and bank loans. In the absence of a multiplier, open market operations, which simply change reserve balances, do not directly affect lending behavior at the aggregate level.2 Put differently, if the quantity of reserves is relevant for the transmission of monetary policy, a different mechanism must be found. The argument against the textbook money multiplier is not new. Theoretically, the bank lending channel is not operative if banks do not face a perfectly elastic demand for their open-market liabilities (see e.g. Bernanke and Gertler, 1995). The issue, then, is to determine whether or not this channel is operative from an empirical perspective. We supplement our conceptual critique with a quantitative analysis. In their influential work, Kashyap and Stein (2000) found empirical evidence in favor of the narrow bank lending channel over the 1976–1993 period. We argue that the existence of a narrow bank lending channel contradicts the institutional developments over the last two decades that improved the banks’ access to external funding drastically. This paper provides institutional and empirical evidence that the money multiplier and the associated narrow bank lending channel are not relevant for analyzing the United States anymore. We find that bank loans are primarily demand driven. Combined with the rising share of loan commitments, these findings imply that bank loans increase following a monetary tightening, which is not consistent with the simple multiplier framework or the narrow bank lending channel. This main argument and the supportive empirical evidence are strongly in line with the new and alternative interpretation of the bank lending channel modeled in Disyatat (2010).

The role of reserves and money in macroeconomics has a long history. Simple textbook treatments of the money multiplier give the quantity of bank reserves a causal role in determining the quantity of money and bank lending and thus the transmission mechanism of monetary policy. This role results from the assumptions that reserve requirements generate a direct and tight linkage between money and reserves and that the central bank controls the money supply by adjusting the quantity of reserves through open market operations. Using data from recent decades, we have demonstrated that this simple textbook link is implausible in the United States for a number of reasons. First, when money is measured as M2, only a small portion of it is reservable and thus only a small portion is linked to the level of reserve balances the Fed provides through open market operations. Second, except for a brief period in the early 1980s, the Fed has traditionally aimed to control the federal funds rate rather than the quantity of reserves. Third, reserve balances are not identical to required reserves, and the federal funds rate is the interest rate in the market for all reserve balances, not just required reserves. Reserve balances are supplied elastically at the target funds rate. Finally, reservable liabilities fund only a small fraction of bank lending and the evidence suggests that they are not the marginal source of funding, either. All of these points are a reflection of the institutional structure of the US banking system and suggest that the textbook role of money is not operative. While the institutional facts alone provide compelling support for our view, we also demonstrate empirically that the relationships implied by the money multiplier do not exist in the data for the most liquid and well-capitalized banks. Changes in reserves are unrelated to changes in lending, and open market operations do not have a direct impact on lending. We conclude that the textbook treatment of money in the transmission mechanism can be rejected. Specifically, our results indicate that bank loan supply does not respond to changes in monetary policy through a bank lending channel, no matter how we group the banks. Our evidence against the bank lending channel at the aggregate level is consistent with other recent studies such as Black et al. (2007), who reach a similar conclusion about the limited scope of the bank lending channel in the United States, and Cetorelli and Goldberg (2008), who point out the importance of globalization as a way to insulate the banks from domestic monetary policy shocks. Our main finding that the contraction or expansion of bank loans are driven by the demand side is in harmony with the theoretical framework of the alternative bank lending channel that is developed in a recent paper by Disyatat (2010).21 Our findings are also consistent with the predictions of Bernanke and Gertler (1995) from over a decade ago that the importance of the traditional bank lending channel would likely diminish over time as depository institutions gained easier access to external funding. Our evidence against the bank lending channel at the microlevel is consistent with Oliner and Rudebusch (1995), but it contrasts previous findings of a lending channel for small, illiquid, or undercapitalized banks (see Kashyap and Stein, 2000 and Kishan and Opiela, 2000 and Jayarante and Morgan (2000)). What is common in all these studies is that their sample periods cover the period prior to 1995, when reservable deposits constituted the largest source of funding. As we have shown in Table 1, this is no longer a feature that characterizes bank balance sheets in the post-1994 period. Furthermore, Kashyap and Stein (2000) and Kishan and Opiela (2000) interpret a change in the sensitivity of bank lending to monetary policy as evidence of a bank lending channel. We argue that changes in the sensitivity of bank loans may stem from the demand side, and that a better test for the lending channel is to check whether bank loans are financed by reservable deposits. Our findings suggest that this is not the case. In general, our results echo Romer and Romer (1990)’s version of the Modigliani–Miller theorem for banking firms. They argue that banks are indifferent between reservable deposits and non-reservable deposits. Hence, shocks to reservable deposits do not affect their lending decisions, and changes to reserves only serve to alter the mix of reservable and non-reservable deposits. Our findings in this paper support the argument that shocks to reservable deposits do not change banks’ lending decisions. Since 2008, the Federal Reserve has supplied an enormous quantity of reserve balances relative to historical levels as a result of a set of non-traditional policy actions. These actions were taken to stabilize short-term funding markets and to provide additional monetary policy stimulus at a time when the federal funds rate was at its effective lower bound. The question arises whether or not this unprecedented rise in reserve balances ought to lead to a sharp rise in money and lending. The results in this paper suggest that the quantity of reserve balances itself is not likely to trigger a rapid increase in lending. To be sure, the low level of interest rates could stimulate demand for loans and lead to increased lending, but the narrow, textbook money multiplier does not appear to be a useful means of assessing the implications of monetary policy for future money growth or bank lending.