Quantitative easing vs credit easing: A comparative study of the impact on interest rates and effectiveness of transmission channels as a result of Japan’s QEP (2001-06) and USA’s CE (2008-13)

Abstract

In the past 20 years, time and again Central banks have faced constraints from conventional monetary policy tools as policy rates approached zero and have resorted to using unconventional monetary policy tools like Quantitative Easing (QE) and Credit Easing (CE). In the wake of the fallout from Japan’s asset price bubble in the 1990s that resulted in a prolonged period of declining economic growth and deflation, the Bank of Japan (BOJ) lowered short term interest rates to 0 in Feb 1999 (also known as the ZIRP – Zero Interest Rate Policy). Two years later, owing to the ZIRP’s failure to boost inflation, the BOJ launched the Quantitative Easing Policy (QEP) in March 2001. The three main pillars of QEP were: (i) the change in BOJ’s main operating target from money market operations to the outstanding current account balances (CABs) maintained at the BOJ. (ii) the BOJ’s commitment to continue with the new framework as outlined in (i) above until CPI (excluding perishables, on a nationwide basis) registered stably a 0% or an increase year on year. (iii) the BOJ aimed to increase the outright purchases of long term Japanese Government bonds (JGBs) as necessary to meet the targeted CABs. Ben Bernanke used the term ‘Credit easing’ at his speech in Londonii to describe and differentiate the policy response of the US Federal Reserve in 2008. Though he agrees that both CE and QEP involve an expansion in central bank balance sheets, CE aims to improve the credit conditions in the economy by changing the composition of assets held by the central bank, unlike Japan’s QEP where the CABs were an important target for the BOJ. We use the methodology used by Krishnamurthy and Vissing-Jorgensen (2011) that mentions that QE works in different channels that affect particular assets differently and uses an event study method to test for the various channels. We aim to evaluate how the channels have worked under Japan’s QEP and USA’s CE to see if either of the two policy responses is superior. To check for the signaling channel, we will look at the spread between 10-year and 5-year JGB futures. To check for liquidity channel, we look at the spread between JGBs yields and Japanese Agency Bond yields. Since there is no prior Japanese Agency Bond Index, we will use individual agency bonds to create custom indices that we will compare with JGB yields across different maturities. To check for safety channel, we will compare the movement of spreads of investment grade and non-investment grade bonds with JGB yields across different maturities. Similar to agency bonds, we will use individual corporate bonds to create custom bond indices. We have identified 7 dates when major policy announcements under Japan’s QEP had taken place. We performed a regression analysis to check whether the yield changes in different maturities of Japanese Government Bonds were significant on these dates. We regress the two-day changes (from the day before to the day after the announcement of QE) against 14 dummy variables, 7 indicating a QE announcement and another 7 indicating a QE announcement on the previous day on any particular date. We use daily data from January 2001 to December 2006 for this analysis and to account for heteroskedasticity in the data, we employ ordinary least squares regression with robust standard errors. For statistical significance of the two-day changes, we use F test to test whether the sum of coefficients of the variables representing QE announcement on a day and QE announcement on the previous day for a particular QE announcement is zero.