During the financial crisis of 2008, several major central banks took drastic unconventional policy measures. As short-term interest rates were close to zero, the limit of conventional monetary stimulus had been reached. Quantitative easing (QE) – the large-scale purchasing of assets such as long-term government debt and mortgage-backed securities – was perhaps the most far-reaching of these unconventional policies. These purchases allowed central banks to inject large amounts of liquidity into the economy, in the hope of stimulating aggregate demand and avoiding economic collapse.
Ten years later, the effects of QE are still controversial and poorly understood. In part this is because our workhorse models of monetary policy predicted that QE would be completely ineffective (Wallace 1981 or Woodford 2012, for example). But central bankers went ahead with it, despite the lack of guidance from economic theory, and there is a widespread belief that QE helped to prevent a deeper recession. Nominal interest rates are still at historically low levels, so in the next recession it is likely that central banks will once again use QE.
The power of QE in a world with liquidity frictions and inequality
Research in monetary economics has made important strides since the crisis. New models have been developed in which households are unequal and hold assets with different degrees of liquidity (Kaplan et al. 2018 is an example). We recently compared the efficacy of QE to conventional monetary policy in this new class of models (Cui and Sterk 2018) and found that QE can provide a powerful stimulus to the macroeconomy, in sharp contrast to Wallace’s result that QE is ineffective.
The stimulus that we found comes from a transformation of the liquidity of household’s asset portfolios. When the central bank conducts QE, it purchases long-term government debt from 'mutual funds'. These purchases are financed by the issuance of reserves that trigger intermediating banks to create deposits. The mutual funds do not hold these deposits because they offer poor returns. Therefore, the deposits end up in the hands of households, who value liquidity as it helps them to smooth consumption during spells of unemployment.
The net effect of QE is that households hold a larger fraction of their wealth in highly liquid deposits. In the US, QE interventions led to a large increase in reserves. This triggered a surge in deposits, much of which ended up being held by households (Figure 1).
Figure 1 Reserves and deposits in the US
Source: Federal Reserve Board, Flow of Funds accounts.
Note: Grey areas denote rounds of QE purchases by the Federal Reserve.
This liquidity transformation stimulates aggregate demand if households have a higher propensity to consume out of deposits than out of other, less liquid forms of wealth such as mutual fund shares. Empirical research suggests this is the case. For example, Fagereng et al. (2018) estimate that households spend more than 60% of a windfall gain in deposits in the first year, while Di Maggio et al. (2018) estimate that households spend only 5-13% of a similar increase in mutual fund wealth.
Traditional models of monetary policy do not reflect this empirical evidence. In these models, households spend only a small fraction of windfall gains, irrespective of the type of gain. This implies that QE has little or no effect. By contrast, our model matches the empirical evidence, and QE emerges as a powerful instrument to manage aggregate demand.
Quantitative easing versus conventional monetary policy
We use the model to evaluate QE as a substitute for conventional policy in achieving the core objective of central banks – that is, to minimise a combination of volatility of inflation and output. When the nominal interest rate is stuck at zero, QE can anchor inflation expectations and avoid belief-driven equilibria. A comparison of optimal QE and interest rate rules reveals that QE tends to be relatively effective in stabilising fluctuations in output and inflation. But even a simple QE rule, under which the central bank targets a constant level of real reserves, turns out to be quite successful in stabilising the economy.
Macroeconomic effects of quantitative easing during the Great Recession
Our heterogeneous agents model can be solved quickly using standard methods employed in central banks and in academia. This allowed us to estimate the model using US data from 2008Q3 to 2015Q4. At this time conventional monetary policy was constrained by the zero lower bound on the nominal interest rate.
We include several shocks in the model, including productivity shocks, cost push shocks, and government expenditure shocks. The estimation exercise gives consideration to both the systematic and the discretionary component of QE policy. In a counterfactual policy experiment we shut down both systematic and discretionary components of QE (Figure 2). The counterfactual suggests that QE during the Great Recession was extremely effective, avoiding a large decline in output and inflation during 2009.
Figure 2 The effects of QE in the US during the Great Recession
Source: Cui and Sterk (2018).
Note: Grey areas denote rounds of QE purchases by the Federal Reserve.
The side-effects of quantitative easing
So should QE replace conventional policy permanently? This is not obvious, because in heterogeneous-agent economies welfare depends not only on the volatility of output and inflation, but also inequality (Gornemann et al. 2017, Bilbiie and Ragot 2017, Bhandari et al. 2017).
Active QE has strong side-effects on inequality. QE interventions create swings in the aggregate supply of deposits that induce time variation in the extent to which households are able to cushion income fluctuations. Fluctuations in firm profits due to QE may also redistribute wealth. So households experience larger swings in consumption under QE policy, reducing social welfare.
Therefore in our model social welfare tends to be lower under optimal QE than under optimal conventional policy. These side-effects mean that QE is second best to conventional monetary policy, despite its effectiveness in stabilising macroeconomic fluctuations.
Bhandari, A, D Evans, M Golosov, and T Sargent (2017), "Inequality, business cycles and monetary-fiscal policy", NBER working paper 24710.
Bilbiie, F O, and X Ragot (2016), "Optimal monetary policy and liquidity with heterogeneous households", CEPR Discussion Paper 11814.
Cui, W, and V Sterk (2018), "Quantitative easing", CEPR Discussion Paper 13322.
Del Negro, M, M Giannoni, and C Patterson (2012), "The forward guidance puzzle", Federal Reserve Bank of New York staff report 574.
Del Negro, M, G Egertsson, A Ferrero, and N Kiyotaki (2017), "The great escape? a quantitative evaluation of the fed's liquidity facilities", American Economic Review 107(3): 824-857.
Di Maggio, M, A Kermani, and K Majlesi (2018), "Stock market returns and consumption", NBER working paper 24262.
Fagereng, A, M B Holm, and G J Natvik (2018), "MPC heterogeneity and household balance sheets," CESifo working paper 7134.
Kaplan, G, B Moll, and G L Violante (2017), "Monetary policy according to hank", American Economic Review 108(3): 697-743.
Gornemann, N, K Kuester, and M Nakajima (2016), "Doves for the rich, hawks for the poor? distributional consequences of monetary policy", Board of Governors of the Federal Reserve System International Finance discussion paper 1167.
McKay, A, E Nakamura, and J Steinsson (2016), "The power of forward guidance revisited", American Economic Review 106(10): 3133-3158.
Wallace, N (1981), "A Modigliani-Miller theorem for open-market operations," American Economic Review 71(3): 267-274.
Woodford, M (2012), "Methods of policy accommodation at the interest-rate lower bound", Jackson Hole Conference: The Changing Policy Landscape, Federal Reserve Bank of Kansas City.
 Central banks also used forward guidance at this time (Del Negro et al. 2012, McKay et al. 2016) and emergency liquidity provisions to firms and banks (Del Negro et al. 2017).