Macro Analysis /

QE for all! EMs learn to navigate the risks of unconventional monetary policy

  • At least 17 EM central banks have initiated QE since Covid, but to a smaller degree than AEs and not yet at all in FMs

  • QE has so far reduced yields without undermining FX stability, but loss of credibility and fiscal dominance are risks

  • Risks most acute in India, Turkey & S. Africa, manageable in S. Korea, Thailand, Chile, Philippines, Mexico & Indonesia

QE for all! EMs learn to navigate the risks of unconventional monetary policy
Tellimer Research
16 July 2020
Published byTellimer Research

Once reserved for developed country central banks, unconventional monetary policy (UMP), and more specifically quantitative easing (i.e. direct central bank purchases of government bonds or other financial assets), has made its way to emerging markets. Struggling with the ongoing Covid crisis and with little room for fiscal stimulus, EM central banks have taken advantage of ample global liquidity to adopt QE into their monetary policy toolkits.

Advanced economies (AEs) have myriad tools to deal with the current crisis, rolling out fiscal stimulus measures averaging 19.8% of GDP in 2020 to combat the Covid crisis. However, debt sustainability concerns have constrained the ability of EMs to respond, with fiscal measures averaging only 5.1% of GDP in large mainstream EMs and 1.1% of GDP in low-income developing countries (see here).

The large reliance on fiscal stimulus in AEs stems in part from a lack of traditional monetary policy ammunition, with an average policy rate of 0.705% heading into the current crisis across the largest five central banks (Fed, ECB, BoJ, BoE and BoC) vs 4.875% pre-GFC. However, major AE central banks have rolled out QE on an unprecedented scale so far this year, proving that they still have ample “unconventional ammunition” when operating at the zero-lower bound.   

According to Fitch, global QE is expected to reach US$6trn by the end of this year (driven mainly by the Fed), equalling half of the 2009-18 total in just one year. Since the beginning of the year, the balance sheets of the four largest central banks (Fed, ECB, BoJ and BoE) have already expanded by c12.7% of GDP on average, and the BIS expects this figure to reach 15-23% of GDP across these four banks plus Canada by year-end.

Despite a weak economic outlook, QE has successfully supported asset prices around the globe. After an initial c34% drop, the S&P 500 is nearly flat ytd and is only c5% from its all-time high. Likewise, after record-shattering outflows exceeding US$100bn from EM assets in the three months through mid-May, AE monetary stimulus has pushed investors back into EM assets in a renewed global search for yield.

The emergence of QE in emerging markets

Taking advantage of the global backdrop, EM central banks have responded by rolling out QE programmes en masse for the first time ever. Following the GFC, just two EM central banks carried out asset purchases (South Korea and Israel, which many would argue are not even EMs), but this time around that number has risen to at least 17. Except for South Africa, Turkey, Costa Rica and Croatia, all these countries are investment grade.

Compared to advanced economy QE, programmes have been much smaller in EMs. Across the five EM central banks which have announced an explicit size for asset purchases, the average is only 0.9% of GDP. That said, since the beginning of the year, central bank balance sheets have expanded on average by 3.7% of GDP (unweighted) across a sample of 13 EMs with QE programmes (compared to 12.7% across the biggest four AE banks), showing a conservative approach to this largely untested new policy tool.    

Positive impact amid lower rates and stable currencies

Thus far, much of the evidence points to a positive impact of QE in EMs. An event study by the BIS shows that 10-year domestic yields fell by 10 basis points on average on the day of announcement and 50bps over the following five days (although this drops to 25bps when controlling for other confounding factors). At the same time, exchange rates appreciated by 1ppt on average on the day of announcement and were broadly unchanged after five days.

Likewise, a CEPR event study of 17 global central banks finds that developed market QE announcements had a statistically significant -0.14% 1-day impact (slightly smaller than past interventions post-GFC), but the impact increases to -0.37% and -0.63% over 1-day and 3-day windows, respectively, for EMs. This is perhaps unsurprising given existing literature that implies diminishing marginal returns to QE, and speaks to its relative efficacy as a policy tool in EMs without previous QE experience.

Updating bond and currency data for the 13 countries in the BIS study shows that 10-year yields have fallen by c120bps on average following the initial QE announcement, though this does not control for other confounding factors and much of this constitutes a clawback of the c80bps selloff across the sample from end-2020 to the imposition of QE.

In addition, policy rates have been reduced by just over 120bps on average since QE started being rolled out in mid-March, potentially accounting for much if not all of the reduction.

Even if QE was not particularly effective, it does not appear to have been harmful either. Since the imposition of QE across the 13-country BIS sample, exchange rates have outperformed the J.P. Morgan EM Currency Index (which is itself up by c3.5% since mid-March) by c2% on average, with Turkey and India as the only exceptions.

Part of the reason EM central banks have been able to roll out QE is broad-based inflation stability in recent years, with inflation averaging 3.9% yoy over the past decade across the 13 countries in question and currently sitting at 2.8%. This is well within the average inflation band of 2-4%, with every country but Turkey experiencing inflation within or below the target band prior to the imposition of QE (including seven below the midpoint and two below the bottom of the band).

Hard-earned credibility must not be squandered

Increased credibility across EM central banks has created the policy space necessary for many EMs to successfully roll out QE. However, QE-related risks to central bank credibility and independence must be carefully monitored to avoid eroding that hard-earned credibility. 

As Fitch pointed out in a recent report, EMs with an independent central bank, a successful inflation targeting record and no debt sustainability issues are better placed to implement unconventional monetary policies. On the other hand, countries with high funding needs and growing debt sustainability challenges could be tempted to use QE more extensively, increasing the risk of fiscal dominance and undermining efforts to develop local capital markets via price distortion.

Other supportive factors include a deep local capital market with a solid investor base, low dollarisation rates and low external liabilities, with Fitch citing Poland and South Korea as countries that meet those criteria (and Chile as meeting the first two). Encouragingly, across our sample of 13 EM countries, local debt makes up c75% of the total debt stock. However, countries like Mexico, Indonesia and South Africa where foreigners hold over a third of domestic debt must still carefully monitor risks.

Putting all these factors together, we have devised a risk scorecard for our sample of 13 EMs. South Korea, Thailand, Chile, Philippines and Mexico standout as countries with manageable QE-related risks, while India, Turkey, South Africa and Hungary are subject to greater risks and must carefully weigh any decisions to embark upon or expand their QE programs.

Countries that lack independent or credible central banks risk stoking inflation and FX depreciation if they attempt QE. Turkey and India are potent examples of this, and have thus seen their currencies underperform the broader EM index by c7% and c2%, respectively, since the imposition of QE.

On the other hand, countries with independent and credible central banks but concerns over debt sustainability face the risk of fiscal dominance if they are pressured to monetise ballooning deficits. South Africa falls into this camp, with the SARB attempting to reassure investors that recent asset purchases are not QE but rather an attempt to restore market functionality amid faltering liquidity (just like the Fed did last September to support the malfunctioning repo market). If the SARB embarked on a bona fide QE program, fears of fiscal dominance could quickly stoke inflation and ZAR deprecation.

Who is next, and should we be worried?

While QE has so far been deployed across 17 EMs, it has yet to be attempted by any frontier market economies. Looking ahead, will lower income countries be tempted to try their hand at QE given its apparent success so far in mainstream EMs?

They have less fiscal space, so may be tempted into UMP, although most are not yet at the zero-lower bound so they have more scope on traditional MP levers. However, despite a dispersion of risks across countries that have so far embarked on QE, they all (except for South Africa and Turkey) are investment grade. If more sub-IG countries follow suit, they may not enjoy such favourable results.

Recent history of global policymaking is littered with examples of emerging and frontier markets copying developed market policy, dating back to the adoption of the 60% debt-to-GDP ratio from the Maastricht Treaty and extending to the decade of countercyclical fiscal policy following the GFC. Many have found it difficult to reverse course, with countries like Kenya failing to make any headway on consolidation (see here). If other B-rated countries (in addition to Turkey) blindly adopt QE simply because others are doing it, while ignoring their relatively greater vulnerabilities, they may experience swift punishment from markets.

Below we quantify the risk associated with QE across a sample of 45 emerging and frontier economies. In the top left are countries with low inflationary risk but high risk of fiscal dominance, and in the bottom right are countries with high inflationary risk but low risk of fiscal dominance. The top right is populated by countries with high risk for both, who should not attempt QE under any circumstances. Conversely, the bottom left is populated by countries with low risk of both, which can probably safely experiment with QE.

This template can be used to identify countries that might benefit from QE vs those where it should be avoided. In the near horizon, QE has been discussed as an option in Brazil and the central bank has been granted legislative approval to purchase public and private debt instruments. However, a history of high inflation, high and rising debt, and growing concerns over fiscal dominance mean Brazil should tread very carefully before embarking on a program of asset purchases.

Other countries that should avoid QE on our metrics include Suriname, Egypt, Zambia, Ghana, Pakistan, Angola and Ukraine. On the other hand, countries at lower risk that have yet to implement QE include Czech Republic, Cambodia and Slovenia, and to a lesser extent Morocco, China, Kazakhstan, and the Dominican Republic (though admittedly Cambodia could be an anomaly given relatively underdeveloped markets).

In the event of a protracted post-Covid recovery, these countries could consider QE to complement existing stimulus measures. Of these countries, we see QE as most likely in Czech Republic, Slovenia or Morocco, as they all have rates at or near the zero-lower bound. The other countries still have positive (in some cases substantially) nominal and real policy rates, and are thus more likely to ease via the interest rate channel.

Vulnerable EMs and frontier markets must proceed with caution

While the risks from QE have thus far been contained, it is worth asking whether it is necessary in the EM context. In AEs, central banks resorted to QE to overcome the constraint of conducting monetary policy at the zero-lower bound. However, within our sample there were no countries at the zero-lower bound before implementation of QE (though Poland is there now), and only five had nominal policy rates below 2%.

A recent BIS study shows that policy rates across the EM universe averaged 4.9% at the beginning of the year, far above the 0.4% that prevailed in AEs. As such, EMs have cut interest rates by 114bps on average compared to 40bps in AEs.

While higher policy rates could be required to compensate investors for greater risk, there could still be more room to lower interest rates via the traditional interest channel. As such, it is unclear that the benefits of QE are worth the risks. The SARB has adopted that stance, saying that it will not embark on a bona fide QE program until it has run out of policy space using its traditional interest rate levers.

Beyond traditional interest rate policy and QE, there is also a risk that some central banks turn to financial repression (i.e. artificially pushing down government borrowing costs via regulatory channels). In Colombia the government has mandated the purchase of “solidarity bonds”, while Turkey has channelled funding to the public sector by including government debt in banks’ regulated loan-to-deposit ratios. Meanwhile, there is discussion in South Africa of “prescribed asset purchases” for its massive public pension fund.

While financial repression constitutes a form of yield curve control, it tends to have a perverse effect on growth by channelling funding away from the private sector and has a much higher cost in terms of credibility. As such, financial repression should be avoided as a policy tool by EM governments struggling with a large debt overhang.

Overall, massive AE monetary stimulus alongside bolstered EM central bank credibility has opened the door for EMs to experiment with QE for the first time with (so far) positive results. However, EM central banks must be mindful that QE is a largely untested emergency tool (in the EM context, at least), and should continue to drive policy via traditional interest rate channels while limiting the size and duration of QE to manage the risks of inflationary pressure and fiscal dominance.

QE presents potential trade opportunities

From the investor perspective, the initiation or expansion of QE in countries with independent and credible central banks, low and stable inflation, and limited risk of fiscal dominance could present an attractive trade opportunity. EMs that have experimented with QE so far have seen a notable drop in yields amid stable or appreciating currencies. However, with average 10-year domestic yields of 4.3% across our EM sample and inflation averaging 3.2%, yields are barely positive in real terms, and further easing could be limited following rate cuts of c120bps already this year.  

On the other hand, frontier markets with less credible central banks and higher debt burdens that attempt QE should be avoided, as they are likely to see FX depreciation and rising inflation in response. That said, countries across the EM/FM universe are still likely to benefit from the large ancillary liquidity boost occasioned by massive AE stimulus efforts, with EM/FM rates and currencies likely to continue performing positively through year-end. Even without QE, this could provide a welcome liquidity boost for struggling EM/FM economies.