AdobeStock_303117429.jpeg
VoxEU Column COVID-19 Monetary Policy

International spillovers of quantitative easing

Several major central banks announced new rounds of massive asset purchases following the outbreak of the Covid-19 pandemic. This policy instrument seems to have performed well for economies that have been implementing it since the Global Crisis, but its spillover impact on external countries has remained a bone of contention within the policy debate. Using previous episodes of quantitative easing as a guideline, this column analyses its international spillovers, showing that they are qualitatively and quantitatively different from the impact of changing short-term rates by the major central banks.

Quantitative easing (QE) is a the tool that was once was referred to as ‘unconventional’,  but that now seems to have become a standard monetary policy instrument in advanced economies that have hit the ‘zero lower bound’. The primary goal of these operations was to reduce long-term interest rates and overcome the ongoing slump in economic activity. In this respect, QE has been found to have performed  well in the countries that implemented it in recent years (Chen et al. 2012, Baumeister and Benati 2013, Kiley 2014).

However, given the importance of the major central banks for global financial cycles (Rey 2013), the scale of asset purchases provoked a heated debate among policymakers on their net benefits to other economies. Some critics (most notably Raghuram Rajan and Per Jansson) expressed concerns about excessive currency appreciation in small open economies (SOEs) and possible imbalances in emerging markets. However, other prominent economists such as Olivier Blanchard, Ben Bernanke, and Mario Draghi stressed a favourable impact of quantitative easing on global demand.

Given the large scale of recent asset purchases by the Federal Reserve, the ECB and the Bank of England, the unsolved question of QE policy spillovers becomes even more urgent. In our recent paper (Kolasa and Wesołowski 2020), we address this issue by investigating the past experience of long-term government bond purchases within a two-country dynamic stochastic general equilibrium model. It matches a number of stylised facts associated with QE spillovers that we document in the paper and provides some insights into spillovers of different monetary policy tools.

Stylised facts about QE in 2009-2017

First, we document the unprecedented scale of asset purchases by showing that they substantially lowered the share of long-term government bonds in the total supply of consolidated public sector liabilities in major advanced economies. By the end of 2017, this share had decreased by over 14 percentage points, as compared with 2009 (Figure 1).

Figure 1 QE impact on the share of long-term government bonds (excluding central bank holdings) in total public sector liabilities in the US, UK and EA

There is strong evidence in the empirical literature suggesting that QE induced capital flows to sovereign bond markets in SOEs. To illustrate their magnitude, we present the share of non-resident investors holdings in the outstanding bonds issued by emerging market governments (in their home currencies, as shown in Figure 2). Since 2009, this share has increased by approximately 15 percentage points. It is important to note that this capital inflow into emerging economies' sovereign bond markets was not matched by offsetting capital outflows associated with other debt securities.

Figure 2 Share of foreign investors in sovereign bond markets of emerging economies

Note: Based on Credit Suisse monthly note "Emerging Markets: Non-residents' holdings in local currency government bonds".

Naturally, flows of this type and scale have affected asset prices. First, they have impacted prices of long-term bonds issued by the recipient countries. As Figure 3 strikingly reveals, the ‘co-movement’ between the term premium on 10-year US treasuries, and 10-year bonds issued by the governments of SOEs, has significantly increased since 2009. Second, they affected exchange rates. According to our panel estimation for 17 small open economies, the relationship between exchange rates and short-term interest rate differentials moved significantly further away from the level implied by the uncovered interest parity condition during the QE period.

Figure 3 Term premium on 10-year bonds in the US and small open economies

Quantitative easing versus short-term rate spillovers

In the paper we show that our model replicates the key empirical facts described above once we simulate the quantitative easing path (as described in Figure 1). First, it accurately mimics the inflow of foreign capital to emerging economies' sovereign bond markets. Second, it accounts for the very strong cross-country co-movement of the ‘term premia’ during the period of quantitative easing (but not necessarily during ‘normal’ times). Third, it is able to produce a downward shift in the parameter of exchange rate projections on the interest rate differential during the period of QE.

Importantly, our model allows us to analyse the impact of QE by major central banks on other variables in small open economies (Figure 4). We find that QE abroad boosts their domestic demand, but strongly undermines their international competitiveness and depresses economic activity as measured with GDP, at least in the short run. This is in contrast to the effects of conventional monetary easing abroad, which positively affects output in other economies. Our framework is hence consistent with the empirical findings from the literature on conventional monetary policy spillovers (Maćkowiak 2007, Banerjee et al. 2016, Dedola et al. 2017).

Figure 4 Impact of quantitative easing abroad on a small open economy

From the model's perspective, the differences between these two forms of monetary accommodation are related to the size of international capital movements and the exchange rate adjustments they induce. For a given magnitude of impact on output in a large economy that engages in QE, asset purchases by its central bank generate a larger inflow of non-residents into sovereign long-term bond markets of other countries, resulting in a much sharper appreciation of their real exchange rates (by approximately three times). In this sense, our model-based predictions support the concerns raised by critics (including Raghuram Rajan) with regard to the impact of quantitative easing in advanced countries on emerging economies.

One way of interpreting our results is that a small economy's central bank using only conventional policy can easily control the short end of the domestic yield curve. However, it is less powerful in affecting its long end. Therefore, following the implementation of QE in a large economy (that strongly depresses foreign long-term rates), the equalisation of ex ante returns on home and foreign bonds is achieved mainly by exchange rate and term premium adjustment, the mirror image of which are massive capital flows.

What can past QE teach us?

Currently, we do not observe inflow of capital nor appreciation of emerging market currencies following the QE in the advanced economies as we did in the recent past. Clearly, amidst the Covid-19 shock ‘fly-to-safety’ motive is a much more important driver of investors’ behaviour than ‘search-for-yield’ which was dominant after the Global Crisis. Consequently, a spike in risk aversion led most currencies to sharply depreciate against the US dollar or the euro.

Nevertheless, we believe that emerging markets’ experience from more tranquil times is still a good guideline for the international spillovers of the recent QE round. If this is the case, new rounds of asset purchases by major central banks may be limiting the scale of capital outflows from small open economies and their currencies’ depreciation caused by the Covid-19 shock. In this way they will be conducive to stabilising not only the large economies that implemented these unconventional policy measures, but also emerging markets.

References

Banerjee, R, M B Devereux and G Lombardo (2016), “Self-oriented monetary policy, global financial markets and excess volatility of international capital flows”, Journal of International Money and Finance 68: 275-297.

Baumeister, C and L Benati (2013), “Unconventional Monetary Policy and the Great Recession: Estimating the Macroeconomic Effects of a Spread Compression at the Zero Lower Bound”, International Journal of Central Banking 9: 165-212.

Chen, H, V Curdia and A Ferrero (2012), “The Macroeconomic effects of Large-scale Asset Purchase Programmes”, Economic Journal 122(564): F289-F315.

Dedola, L, G Rivolta and L Stracca (2017), “If the Fed sneezes, who catches a cold?”, Journal of International Economics 108: 23-41.

Kiley, M T (2014), “The Aggregate Demand Effects of Short- and Long-Term Interest Rates”, International Journal of Central Banking 10: 69-104.

Kolasa, M and G Wesołowski (2020), “International spillovers of quantitative easing”, Journal of International Economics

Maćkowiak, B (2007), “External shocks, U.S. monetary policy and macroeconomic fluctuationsin emerging markets”, Journal of Monetary Economics 54(8): 2512-2520.

Rey, H (2013), “Dilemma not trilemma: the global cycle and monetary policy independence”, Proceedings - Economic Policy Symposium - Jackson Hole.

1,995 Reads