Willing to challenge conventional views, Marvin expressed concern that “central banks [would] be tempted to rely even more heavily on balance sheet policy in lieu of interest rate policy, in effect exerting stimulus by fiscal policy means via distortionary credit allocation, the assumption of credit risk and maturity transformation, all taking risks on behalf of taxpayers and all moving central banks ever closer to destructive inflationary finance.” He understood the associated heartburn of this view, acknowledging that negative interest rates was an idea that would likely require “some getting used to.” He noted, however, that the public also was initially resistant to leaving the gold standard, and later to floating the exchange rate, but gradually accepted these changes.
While the use of negative interest rates gives me pause as a way to address a future encounter with the effective lower bound, I share Marvin’s concerns about the potential side effects of balance sheet policies that pose risks to financial stability and threaten the central bank’s policy independence.
As a voting member of the Federal Open Market Committee (FOMC) in 2013, I expressed my concerns about the continuation of the asset purchase program known popularly as QE3. By then, financial markets were stable and the economy was growing. These concerns about the expansion of the Fed’s balance sheet under those conditions echoed many of Marvin’s concerns. In my view, the possible unintended side effects of the ongoing asset purchases posed risks to economic and financial stability and served to unnecessarily further complicate future monetary policy (George 2017). Today, however, it is conventional wisdom that the benefits of asset purchases have been clearly established and that their potential costs have proven negligible. History and further research may ultimately affirm that wisdom, but it remains less than clear to me that the longer‐run costs of balance sheet policies have been fully taken into account.
As a consequence of large‐scale asset purchases, for example, the Federal Open Market Committee (FOMC) had to evaluate and reconsider its longstanding operating framework. Given the abundant reserves associated with its balance sheet policies, the FOMC had to consider whether the federal funds rate target could be achieved administratively by setting the interest rate on excess reserves. Indeed, to ensure effective interest rate control and establish a firm floor on overnight rates, an overnight reverse repo facility was created. In addition, as the Fed began to shrink its balance sheet, it proved challenging to gauge the minimum reserve balances needed for achieving the federal funds rate target without intervention by the open market desk at the New York Fed. As a result, the desk resumed regularly conducting repo operations and outright purchases of Treasuries to build a bigger buffer and ensure an ample supply of reserves. These operations have caused some confusion in markets as some participants have seen them—incorrectly, in my view—as a type of quantitative easing.
More generally, to the extent that large‐scale asset purchases succeeded in their aim of creating a wealth effect, they also played some role in contributing to elevated asset valuations. These effects, together with the perception that interest rates will remain at historically low levels for a prolonged period, can lead to a buildup of financial imbalances that ultimately pose risks to the real economy. Experience has shown that these imbalances can develop in sectors outside the lens of regulators and, as we witnessed a decade ago, can unwind with little warning.
Another concern I share with Marvin is the risk that income from the Fed’s large balance sheet combined with our capital surplus could tempt fiscal authorities to view the Fed as a source of funding for government programs. I would argue that we have seen a degree of this risk unfold. The funding of the Consumer Financial Protection Bureau (CFPB), as required under the Dodd‐Frank Act, is a case in point. Each quarter the Reserve Banks transfer to the CFPB, without congressional appropriations, the amount of funds requested by the director of the CFPB to carry out its operations. To date, the Federal Reserve has transferred almost $4.4 billion to the CFPB. In addition, the Dodd‐Frank Act required the Fed to fund the first two years of the Office of Financial Research in support of the Financial Stability Oversight Council (FSOC).
Yet another example of congressional funding of programs outside the regular appropriations process is the “Fixing America’s Surface Transportation (FAST) Act.” In the FAST Act, Congress funded highway construction by reducing the Federal Reserve Bank stock dividend rate for member banks with assets of more than $10 billion. The Act also placed a cap of $10 billion on the aggregate surplus funds of the Federal Reserve and directed that any excess be transferred to the Treasury general fund. The potential policy implications of modifying dividends to member banks, or more generally, the requirement for member banks to purchase stock in a regional Federal Reserve Bank, is a concerning development that risks undermining the Federal Reserve’s long‐standing institutional design of public and private interests serving the American public (George 2016).