Quantitative easing is when the central bank creates new money electronically, and uses it to buy long-term bonds from the government. It is an unconventional monetary policy that increases the value of long-term bonds and thereby decreases their interest rates. The government’s aim is to lower the reward for saving and in turn increase investment and stimulate aggregate demand. Quantitative easing is often used when central banks have reduced their interest rates to their lower bound but still need to stimulate the economy. The Bank of England (UK) and Federal Reserve (US) used quantitative easing after the recession in 2008/09 to try and to boost GDP (Weale and Wieladek, 2016). Furthermore, Japan used quantitative easing in March 2001 to help eliminate persistent deflation (Bowman et al., 2015). Using economic theory and the examples above, I will evaluate whether quantitative easing has worked.
Conventional monetary policy transmission, as described above, is when the central bank purchases long-term government bonds to try help stimulate GDP. Central banks use their leverage over nominal, short-term interest rates which decreases the rewards to saving and the cost of capital. The price level is sticky, meaning it doesn’t change in the short-run, and therefore, the change of the short-term interest rate affects the real interest rate. This encourages firms to invest and households to increase consumption on durable goods meaning aggregate demand increases which positively affects GDP. In response to the 2008/09 financial crisis both the Bank of England (UK) and the Federal Reserve (US) purchased a large amount of government assets to provide monetary stimulus and increase GDP (Weale and Wieladek, 2016). The conventional monetary policy transmission was deemed successful after the UK and US both showed improvement with increasing GDP. However, there is an identification problem. Quantitative easing in the UK and US would have only been effective if monetary policy had a causal effect on GDP. There is a problem of reverse causality where monetary policy can cause GDP to change but authorities choose their monetary policy in response to current GDP. This can be explained with equation (1) where we see reverse causality between y (output/GDP) and m (monetary policy).
(1)
Equation (1) also highlights an assumption made about time lags. We assume that causation for both monetary policy and GDP take time. This is because monetary policy affects the economy with a time-lag and GDP figures are announced by the government with a lag. Furthermore, there is an issue of endogeneity where:
(2)
This means that the wanted exogenous change in monetary policy is correlated with the error term to GDP and therefore is not entirely exogenous. To overcome this issue, we need to decompose the error term into endogenous and exogenous parts. We can do this by using a Structural Vector Auto-Regression (SVAR). A SVAR uses time-series data to estimate the relationship between macroeconomic (i.e. GDP) and monetary policy variables. Weale and Wieladek (2016) use a SVAR model to analyse the impact of quantitative easing on real GDP and the CPI in the UK and US after the 08/09 recession. Their study explores three mechanisms, other than the conventional monetary policy transmission, through which quantitative easing could influence GDP and CPI. The first mechanism they explore is the portfolio balance channel. This mechanism relies on investors that prefer a given maturity in the government bond market and therefore we would see a reduction in yields on the debts of the maturities purchased (Vayanos and Vila, 2009). The second mechanism that Weale and Wieladek (2016) explore is the signalling channel. This is the idea that increased quantitative easing signal that the government will keep the policy interest rate at its lower bound for longer. The third possible mechanism is that quantitative easing manages expectations about future economic outcomes and reduce economic uncertainty. All three mechanisms could lead to wealth effects where we would see an increase in aggregate demand through higher investment and consumption. Weale and Wieladek (2016) identify variables that would be expected to be influenced by asset purchases if each of the mechanisms played a role and therefore, estimated which, if any, mechanisms were most effective. The study showed that the maximum impact on GDP from an asset purchase announcement shock worth 1% was 0.58% in the US and 0.25% in the UK (Weale and Wieladek, 2016). Further results showed that the GDP of US was mostly affected through the impact on long-term government bond yields and that short-term swap rates were unaffected. This shows that quantitative easing in the US worked most effectively through portfolio balance channel and that it is unlikely signalling was important. In contrast, neither portfolio balance nor signalling played a significant role in increasing GDP in the UK. However, there was evidence that quantitative easing reduced household uncertainty in both the UK and the US. These results are important when thinking about whether quantitative easing works. Firstly, we can see from the SVAR that monetary policy did in fact have a causal relationship with GDP. Furthermore, the results show that quantitative easing in both the UK and US was effective following the 08/09 recession but that the increase in GDP was not necessarily due to the conventional monetary policy transmission and that instead other mechanisms were important.
There is little agreement about how exactly monetary policy impacts the economy (Bernanke and Gertler, 1999). Therefore, it is important to examine both the direct and indirect effects of quantitative easing. One example of an indirect amplification mechanism that works alongside the conventional monetary policy transmission is the credit channel. The credit channel affects the economy through endogenous changes in the external finance premium (Bernanke and Gertler, 1995). The external finance premium ‘reflects the deadweight costs associated with the principal-agent problem that typically exists between lenders and borrowers’ (Bernanke and Gertler, 1995, p. 35). Expansionary monetary policy, including quantitative easing, is said to reduce the size of the external finance premium and increase credit availability in the economy. However, the external finance premium takes both adverse selection and moral hazard into consideration. The problem of adverse selection arises from the ‘lemons’ premium whereby borrowers have more information about their liquidity and ability of repaying a loan and moral hazard is about whether the borrower will pay back the loan. Bernanke and Gertler (1995) break down the credit channel further by discussing two mechanisms that link monetary policy and the external finance premium: the balance sheet channel and the bank lending channel. The balance sheet channel is based on the fact that the finance premium that a borrower faces should depend on the borrower’s liquid assets and collateral. If a borrower has a strong financial position, they should face a lower external finance premium. This is because the greater the borrower’s net worth the less risky the loan. The balance sheet channel changes when quantitative easing arises as an increase in long-term government assets would not only affect the market’s interest rates but also the financial positions of borrowers. One example of this is that expansionary monetary policy would reduce interest rates and lower the burden on borrowers outstanding debts, strengthening their financial position. Furthermore, a decline in interest rates would increase asset prices, meaning the value of the borrower’s collateral increases. Bernanke and Gertler (1995) plot the federal fund rate (US) against the coverage ratio which represents a borrower’s health. They find a strong relationship between the two variables showing that a change in the fund rate translates into a change in the coverage ratio and in turn the financial position of borrowers. We can relate this information to quantitative easing by imagining an improvement in the financial position of borrowers, encouraging them to take out more loans and use the money for investment. However, this is where the issue of adverse selection arises. In some cases, without this sudden financial stability, the borrowers would be credit constrained and unable to obtain a loan. The banks may trust these borrowers and agree to the loans which in the short run will increase investment and in turn improve GDP. However, in the long run these borrowers may not be able to pay back their loans. Furthermore, if quantitative easing is reversed, and interest rates increase in the long run, these loans may be unsustainable for those borrowers and they may have to default on their loan. This highlights an issue with quantitative easing that its effect in the long run may lead to financial instability. Therefore, quantitative easing will only work if the effect is large enough for borrowers to grow their collateral to a level that will be financially stable if quantitative easing were to be relaxed and interest rates were to increase.
The second mechanism that Bernanke and Gertler (1995) discuss is the bank lending channel. The bank lending channel explains how the change in the external finance premium also affects the credit of banks and the number of loans they provide. Quantitative easing should inject liquidity into banks and increase the supply of banks loans. Bowman et al. (2015) explore the bank lending channel of Japan during their use of quantitative easing in March 2001. In 1990, Japan’s bubble economy burst, leading to low economic activity and price deflation (Bowman et al., 2015). By 1999, the Japanese Government had failed to boost the economy with the use of expansionary monetary policy whereby they reduced the short-term policy rate to 0. Therefore, the Bank of Japan decided to use quantitative easing to try reverse price deflation. The Bank of Japan increased the current account balances held by financial institutions and used this to buy long-term Japanese Government bonds. The aim was that the injected liquidity and reduced interest rates would encourage investment, spending and reduce deflation. The Bank of Japan announced it would maintain the quantitative easing programme until the core consumer price index stopped declining. However, the Bank of Japan formally ended the quantitative easing programme in March 2006 when it had failed to reverse deflation. This does not necessarily mean quantitative easing does not work. Bowman et al. (2015) used panel data regressions to measure the effect of the policy. They wanted to see whether quantitative easing had any effect and whether a positive effect was just overwhelmed by other negative forces in Japans post bubble economy. Bowman et al. (2015) used data from 137 Japanese banks from March 2000 to March 2009 to form panel data regressions. The regressions related to the bank lending channel and explained each banks change in loans using the lagged liquidity position of the bank as well as an array of control variables. For quantitative easing to have worked, the results should show an increase in the number of loans when there has been an injection of liquidity. The results showed that on average a 1% increase of liquidity ratio led to a 0.11% increase in loan growth in the next six months. This shows a small, positive and statistically significant effect of bank liquidity on bank lending. This suggests that although the effect is small, and was evidentially not big enough to reverse deflation, it is not the case that quantitative easing did not work at all. Bowman et al. (2015) discuss how this positive effect could have been offset by the reduction in loans between central and inter-banks during this time. Furthermore, the results showed that the effect of quantitative easing varied depending on the original liquidity of banks. Weak banks benefited more from the quantitative easing programme than stronger banks as it reduced their liquidity constraints. Therefore, this study shows some evidence that quantitative easing worked when affecting the supply of credit. However, the overall effect was quite small and was more effective on weaker banks. Overall, the Bank of Japans aim of reversing deflation with quantitative easing could not have been achieved without huge amounts of liquidity.
To conclude, we can assess as to whether quantitative easing has worked by looking at both the direct and indirect effect of it. However, we can draw from the examples above that the success of quantitative easing is dependent on the country’s economic state. Quantitative easing would usually occur when an economy is experiencing a recession but the effectiveness of the monetary policy is reliant on many variables including consumer and business confidence, banks liquidity and borrower’s collateral. Regardless of this, quantitative easing has had a positive effect on the examples above, even if the impact has only been small, therefore, showing that quantitative easing has worked.