10-Year Treasuries: The Ultimate Test for Emerging Markets?

The Thai economy in the 1990’s was growing at a rate that would draw envy from even the CCP of the modern day, with growth averaging 9% in the early part of the decade [1]. Neighbouring countries including Indonesia, South Korea and Malaysia were all following the Thai lead, with the spectacle labelled as an economic miracle by multilateral institutions. Although perhaps surprising, such expansionary proceedings were natural under traditional growth theory such as the Solow Model, which states that countries further below their steady state of output per capita grow faster than counterparts closer to equilibrium: to put it simply, East Asian countries were only beginning to mechanise their economies and leverage the ubiquity of globalisation for their own benefit, with cheap labour attracting multinationals and export-driven growth leading the way.

But most of the growth in Thailand was predicated upon cheap credit propping up the mortgage sector, as well as a relatively expensive Baht pegged against the Dollar. When the real estate sector began to tremor, the Baht went with it, portending a contagion throughout the developing world as currencies and other financial markets sold off: the East Asian Crisis had begun. To incorporate more Macro theory: the Thai economy had simply overshot its steady state and its consumption and capital stock had reached unsustainable levels. Traditional Real Business Cycle Theory would suggest that consumption and capital stock therefore had to decrease.

Although the operations of the Fed in the mid-90’s was not the sole cause of the crisis, in recent times US monetary policy has often exacerbated (or even caused selloffs) in emerging markets. With US 10-year Treasuries seeing a rise of around 80 basis points since the turn of the year, emerging markets have seen inflows turn to outflows [4]. Hot money leads dynamics and with increasing US rates, the Dollar has risen against its basket of EM currencies. This in turn exacerbates selloffs due to increases in interest payments on USD-denominated EM debt. 2021’s episode has been analysed in a similar fashion to the Taper Tantrum episode of 2013, when 10-year yields rose when Chairman Bernanke announced a reduction in the Fed’s balance sheet. Although it could be argued that a lower-for-longer case is still likely given that US breakeven are predicting an initial spike in CPI, followed by inflation just above the 2% average target of the Fed [5]. However, the relationship between the 10-year yield and emerging markets is – on the surface – seen and hypothesized to be negative. But how strong is this effect? And what role does the dollar play?

Using the Data

I aimed to discern the impacts of rates and the dollar on emerging markets via a Vector Autoregressive Model (VAR(p)), a model in econometrics that uses a system of equations to analyse the effect of lagged variables on other variables (including the dependent variable itself). Mathematically, this can be shown as the following system of equations:

Where:

Z(t) is the dollar index (on a trade-weighted basis) against a basket of EM currencies [6];

X (t) is the 10-year Treasury yield [7]; and

Y (t) is the MSCI Emerging Markets Index [8].

They are all plotted on the graph below.

Left axis: Dollar-EM Index and MSCI EM Index ; Right axis: 10-year Treasury yield

As discernible, the effect of one variable on another variable is through a collection of direct and indirect effects throughout the system of equations, captured by an Impulse Response Function, which is the cumulative result of a shock to one variable by one standard deviation.

Daily data was collected from FRED and MSCI from 30^th January 2012 to 11^th March 2021, which incorporates shocks such as the aforementioned Taper Tantrum and the onset of the COVID-19 pandemic. Therefore, a holistic view is possible. After transforming the data and testing against key criteria, a VAR(7) model was used to discern the output, thus p = 7 in the above system of equations. Although the model’s residuals were heteroskedastic and not normally distributed, the model possessed no serial autocorrelation.

To assess the effect on the percentage change of the MSCI EM Index from a shock in first difference of 10-Year Treasuries – that is, a shock of 4.5 basis points – the Impulse Response Function is used, as shown below:

To much surprise, an upwards shock in the 10-year yield actually leads to a positive (but small and muted) upwards movement in the MSCI Emerging Market Index, giving a contrarian consensus compared to much market commentary regarding the effects of rising rates on emerging market indices – this effect is also statistically significant. This is perhaps shown by the fact that the transmission of shocks between rates and the Dollar are very close to zero. Brief analysis shows that the dollar index and 10-year rates are not co-integrated, and hence this could be the stumbling block. But there’s also an elephant in the room: China.

The Role of China

Despite the Chinese economy being a behemoth worth over $14 trillion, it is included in the MSCI EM Index and is therefore heavily weighted (39.48%). As China has only recently returned to the USD-denominated debt market, as well as possessing strong domestic firms which make it an economic superpower, it is arguably impervious to US rate rises as its domestic bond market has offered attractive yield – an almost oasis amongst the lower-yielding fixed income world underpinned by the largesse of QE. Therefore, the effect of rate rises on the Index is muted due to the dominance of China in the Index. One would definitely hypothesise that an ex-China index would react significantly differently.

Bond Market Volatility

The introduction of QE has lowered 10-year yield volatility (below). With extra liquidity sloshing around in markets, coupled with the lower-for-longer consensus adopted by policymakers and supplemented with tools such as forward guidance, the Treasury market is – aside from brief shocks such as COVID and the reflation trade – significantly less volatile in recent times. Therefore, as MSCI data is only used from 2012 onwards, the standard deviation used for the 10-year shock is smaller than a pre-QE figure. This could have skewed the findings.

Yield (and therefore price) volatility has fallen significantly since the introduction of QE. Source: FRED

What Next for EM?

It is still evident that the Dollar drives some part of EM flows. However, it is important to consider fundamentals as the world recovers from the pandemic. Part of the reflation trade has been due to strong economic growth in the US, with the Fed revising its forecast to 6.5% compared to December’s estimate of 4.2% [9]. Emerging economies are struggling: Brazil’s COVID crisis (as a result of incredibly relaxed restrictions in place) is becoming worse by the day, vaccine inequality is very evident [10], and emerging economies in the Euro area will have been affected by the EU’s vaccine hiccups. Only via addressing the urgent need for mass scale vaccination in the world’s emerging (and most populous) economies will we see an improvement from a fundamental perspective. Although this is already evident in countries which have handled the pandemic competently, such as Vietnam and Thailand: the differential in growth should not take long to recover.

More information on:

Solow Growth Model (using intensive labour) [11]
Ramsey Business Cycle Model [12]
Vector autoregressive models [13]