Theoretically, Fed asset purchases, known as quantitative easing, are supposed to stimulate economic activity through bank lending. While the Fed can grease the skids of money supply, it's really fractional reserve banking that drives money creation (see "The Fed, the Banks, Money Creation, Inflation, and the Bond Market," February 21, 2011). That is because for every dollar the Fed creates, the banking system can create multiple dollars. In short, if central bank monetary easing is to stimulate economic activity, banks must increase lending.
Banking is a spread business, i.e. banks borrow money and lend it back out at higher rates, earning the difference, or the spread, between their borrowing and lending rates. Generally speaking, the maturity of bank borrowing (checking and savings accounts, CDs, commercial paper, etc.) is shorter than the maturities of their loans. As such, the steeper the yield curve, i.e. the wider the spread between short and long-term interest rates, the more attractive it is for banks to lend money.
The current spread remains well below the maximum level, which occurred in January, 2011, but it is also more than twice the median level of the last 40+ years and nearly twice the average (Chart 1).
In fact, spreads have been at historically attractive levels since the beginning of 2008 (Chart 2). From the beginning of 2009 through the middle of 2011, the spread was consistently more than 200bps, averaging 244bps from the beginning of 2009 through August of last year.
Spreads over the last four and a half years have been far wider than those experienced during far better economic climates. Using the same month-end data, from 1983 through 1999, the 2-10-year spread averaged 83bps, nearly identical to the average of the last 40 years. From 1994-1999, it averaged a scant 47bps. Yet in both those periods banks were far more willing to lend.
But banks aren't lending much of the money the Fed has been creating. How do we know that? The Fed provides monthly data on monetary aggregates and bank reserves as a percentage of those aggregates. From 1959 through August of this year, banks kept an average of 9.77% of M1 on reserve with the Fed with a median level of 5.3%. From January 1959 through August 2008, bank reserves as a percentage of M1 never exceeded 8.03% and declined steadily from the inception of the data in 1959 through August 2008 (Chart 3).
The average has been skewed higher by the dramatic increase in reserves since the fall of Lehman Brothers in September 2008. From September 2008 though August of this year reserves as a percentage of M1 has averaged nearly 63%! Moreover, reserves as a percentage of M1 grew enormously during each of the Fed's two previous quantitative easing programs (for these purposes, we do not include Operation Twist activities under which the Fed has sold short-term debt while simultaneously buying longer term debt as it does not result in an increase in the Fed's balance sheet). While remaining enormous relative to historical norms, they declined significantly in the aftermath of each of the last two quantitative easing programs (Chart 4). Reserves first peaked at just over 70% in February 2010, just before the end of QE1. They then declined to just below 60% at the start of QE2 and rose to new highs at just more than 85% as the program ended in June 2011.
While the economy dug itself out of a deep contraction during the implementation of QE1, its important to note that there was massive fiscal stimulus being conducted at the same time. The Treasury Department rolled out TARP (the Troubled Asset Relief Program) while Congress passed the American Recovery and Reinvestment Act. As such, it is virtually impossible to delineate the impact of the Fed's quantitative easing on the recovery. Yet we do know that the Fed purchased $600 billion worth of Treasuries from November 2010 through June 2011 with no discernible impact on the U.S. economy. Outside of a 4.1% GDP print in the 4th quarter of 2011, the economy has not grown faster than 2.5% in any quarter since QE2 was initiated and the economy has experienced four quarters in which growth was 2% or less. Furthermore, the strength in the 4th quarter of last year can largely be written off as an anomaly as it was largely the result of a lack of inflation. Nominal growth, which was 4.2%, was very much in line with the trend of that has been in place for the last three years.
The Fed's attempts at money creation have been stifled by financial institutions, which have placed the vast majority of the proceeds received from the sale of assets back on deposit with the Fed. We also know that this is occurring despite historically attractive spreads between short and long-term interest rates. Why might that be the case? There are likely two culprits: the deleveraging process and the low overall term structure of interest rates. Financial institutions significantly increased their leverage in the 15-20 years prior to the crisis, allowing consumers to do the same. With both continuing the process of de-levering, the demand for money is far lower than the supply. Additionally, banks are likely fearful that a rapid rise in rates could push their borrowing costs above the yield on their loan books given the exceptionally low level of yields across the curve.
The Fed's QE programs and Operation Twist have largely been failures. Despite the stated goal of lowering longer-term interest rates, the yield on the 10-year Treasury rose during both QE1 and QE2 only to fall once the programs ended. Meanwhile, equity markets rallied during both QE1 and QE2, but struggled in the aftermath of each. We suggest that the Fed's impact on the "risk-on, risk-off" has been greater than its affect on the U.S. economy. While longer-term yields have declined since the inception of Twist a year ago, this has not done much to stimulate bank lending, which is no surprise. Lower rates on the long-end have resulted in tighter spreads. As a result, banks, which were not eager to lend at historically wide spreads, have become even more reluctant to do so.
Not only is the Fed fighting a battle its unlikely to win, but it may well be sowing the seeds of future wars. Clearly its actions increase the potential for a significant and sustained rise in inflation, although that is unlikely to be an imminent threat while banks continue to hoard the money the Fed creates. Another risk is to the Fed's own massively over-levered balance sheet, which is levered more than 50:1. Moreover, the Fed's policy initiatives are in unchartered waters and may well result in 2nd or 3rd order effects that are as yet unknown.