Macro Analysis /
United States of America

History lessons for EM assets as Fed starts to hike

  • Markets have shifted from pricing c3 Fed hikes this year to c6.5 amid rising inflation, but expectations remain anchored

  • EM assets have typically underperformed going into hiking cycle but outperformed after commencement

  • This could make for better backdrop for EM assets this year, but could be offset by higher fundamental vulnerabilities

History lessons for EM assets as Fed starts to hike
Tellimer Research
14 February 2022
Published byTellimer Research

US interest rates have shot up so far this year amid rising inflation, which rose from 7% to 7.5% yoy in January, a 40-year high and above the 7.3% Bloomberg consensus. The US yield curve has consequently flattened in anticipation of tightening by the Federal Reserve, with 2-year yields rising by 85bps and 10-year yields rising by 50bps ytd and breaching the 2% mark for the first time since July 2019.

10yr yield components

The increase has been driven entirely by higher real yields, with the 10yr inflation-adjusted rate rising by 62bps ytd in anticipation of tighter Fed policy. Meanwhile, inflation expectations have remained well-anchored, with the 10yr breakeven actually falling by 12bps ytd and the 5y5y breakeven fluctuating around 2.2% for the past year.

US breakeven inflation

The market is now pricing in nearly seven hikes of 25bps over the course of 2022 compared to just under three at the end of 2020. While EM central banks have thus far stayed ahead of the curve, a hawkish shift at the Fed could leave EM assets vulnerable if EM central banks fail to preserve a positive real interest rate differential and if the withdrawal of unprecedented global liquidity leads to a rise in risk aversion and capital outflows from EM.

MIPR

With the Fed almost certain to commence its hiking cycle next month, we examine how EM assets have fared during past hiking cycles. We use the JP Morgan EMBI total return index to proxy for hard currency debt, and the MSCI EM Currency total return index to proxy for local currency debt. The EMBI dates back to January 1994 while the MSCI dates back to October 1997, but we truncate the EMBI sample to October 1997 for comparability and to remove distortions from the mid-90s EM financial crisis.

As a starting point, we compare the average monthly returns for each index over the sample period to the returns in the year leading up to and year after the commencement of a hiking cycle (defined as a cyclical trough-to-peak movement in the Fed Funds Rate). We find that the EMBI on average returns 0.69% monthly versus just 0.35% for the MSCI, resulting in significant outperformance by the EMBI over the sample period (up 6.1x versus just 2.7x for the MSCI).

For both asset classes, however, there is evidence of a drop in returns in the 12 months before the commencement of a Fed hiking cycle followed by significant outperformance in 12 months after the cycle commences. This makes intuitive sense given the steady improvements in forward guidance by the Fed, with markets seemingly pricing in the impact of rate hikes on EM assets well in advance and enjoying a strong relief rally after the cycle commences.

EM returns

Looking at this a different way, the EMBI has returned an average of 17% over the 12 months following the commencement of Fed hikes over the last three hiking cycles (1999-2000, 2004-2006, and 2015-2018) versus just 1% in the 12 months leading up to the first hike, while the MSCI has returned an average of 8% in the 12 months after the first hike versus 5% in the 12 months prior. Overall, this paints a relatively sanguine picture for both asset classes as the Fed gets ready for lift-off next month.

Taking a slightly more rigorous approach, we regress changes in the EMBI and MSCI indexes on a Fed dummy (1 if it is in a hiking cycle and 0 if it is not), controlling for changes in commodity prices (proxied by the BCOM index) and risk appetite (proxied by the VIX). For both asset classes, we find that the Fed dummy takes on a negative sign when controlling for BCOM and VIX, but that it is neither quantitatively or statistically significant (especially so for the MSCI), suggesting that other factors are responsible for driving EM returns than the Fed policy stance (also illustrated by the low r2 for both regressions).

Hard currency returns

Local currency returns

Taking a slightly more rigorous approach, we regress changes in the EMBI and MSCI indexes on a Fed dummy (1 if it is in a hiking cycle and 0 if it is not), controlling for changes in commodity prices (proxied by the BCOM index) and risk appetite (proxied by the VIX). For both asset classes, we find that the Fed dummy takes on a negative sign when controlling for BCOM and VIX, but that it is statistically insignificant (and quantitatively insignificant for the MSCI), suggesting that other factors are responsible for driving EM returns than the Fed policy stance (also illustrated by the low r-squared for both regressions). 

Regression table

By restricting the sample to Jan 2005-Jan 2022, we can add in changes to the 1y1d US$ interest rate swap to proxy for expectations of the Fed Funds Rate over the next 12 months. For the EMBI, the 1y1d swap is significant at the 95% confidence level and implies a 1.1% drop in the monthly EMBI return for each 100bps increase in 12-month interest rate expectations, while the Fed dummy becomes significant at the 90% confidence level. Meanwhile, neither the 1y1d swap nor the Fed dummy are statistically or quantitatively significant for the MSCI, suggesting Fed policy and forward guidance have a greater impact on hard currency than local currency assets.

In any case, for both of the expanded regressions, the r-squared remains relatively low (25% for the EMBI and 41% for the MSCI). The natural conclusion is that there is a myriad of fundamental factors that drive EM asset performance beyond what can be explained by Fed policy, commodity prices, and risk sentiment. While we must accept these shortcomings for the time being due to a lack of high-frequency fundamental EM aggregates, this is a possible area for further research. Further, with just 2-3 hiking cycles to go off for each model, any conclusions should be taken with a grain of salt.

Overall, history tells us that EM assets tend to perform better in Fed hiking cycles than they do before the cycle begins. Barring any adverse inflation surprises or disorderly adjustments to Fed policy in the year ahead, therefore, EM assets may be nearing a positive inflection point. However, fundamentals still appear to be the key driver of EM asset performance, and the pandemic-induced growth overhang and rise in debt vulnerabilities in many EM, alongside elevated inflationary pressures and commodity price shocks worldwide, may be difficult headwinds to overcome. Further, if stubbornly high inflation pushes the Fed into an even more hawkish response than expected by markets (which are already pricing in a very hawkish response), EM assets could be in for an adjustment. That said, it remains to be seen if EM assets are at an inflection point or if they will break the historical mould.