Stress Testing the Global Impact of U.S. Rate Increases

Tracy Chan, Analytics Consultant; Elena Daskalova,  Analytics Consultant; Nina Xing, Analytics Consultant; and James Cameron, Analytics Consultant contributed to this article. 

Citing a positive economic outlook, in June the U.S. Federal Reserve raised the Fed Funds rate by 25 basis points to a target range of 1.75% to 2.00%, following March’s 25 bps increase. Between the two hikes, the U.S. 10-year government bond yield peaked at 3.1%, its highest level since 2011. In recent weeks, the yield has hovered just below the key psychological 3% level. Fund managers in U.S. and around the globe now find themselves pondering the question; How much further do bond yields have to rise before triggering the danger zone of a major U.S. share market sell-off?

Historically, the correlation between the U.S. 10-year treasury bond yield and key equity indices, e.g. the S&P 500 and MSCI AC World, is not particularly strong. At times when the bond yield rose (2009, 2013, and 2016), both equity indices increased as well.

Analysts surveyed by FactSet expect to see another two rate hikes from the Fed and the 10-year benchmark at 3.1% by end of 2018. How will this impact global equity/bond markets? Which countries/regions would be most affected? We already saw Argentina seek assistance from the International Monetary Fund in May as a sharp depreciation in the country’s currency caused a surge in the value of its substantial foreign currency debt.

What Happens if the 10-Year Yield Jumps to 6%?

Using FactSet’s Monte Carlo, currency-independent MAC risk model, we performed both event-weighted and time-weighted stress tests on major fixed income and equity indices. Because the 10-year U.S. treasury yield is a key driver for both U.S. and global markets, we wanted to find the impacts of varying the U.S. 10-year treasury bond yield within a 3-6% range. Although a yield jump to 6% is unlikely to happen overnight, the stress test was added to gain insight on how a more severe and sudden increase (a Black Swan event) would affect global markets. In event-weighted stress tests, similar periods to the stress results over history are weighted more heavily, while the time-weighted methodology applies heavier weighting to recent periods.

Since the bond market would be directly affected by a change in the 10-year yield, let’s start with an analysis of what would happen when U.S. issues become more attractive. We chose to test the three different rate scenarios on one of the largest global bond indices, the Barclays Global Aggregate index, which includes approximately 22,000 global bonds and has a market value of $50 trillion.

Overall, an increase in U.S. rates has a negative impact on the global bond market, with a slightly stronger negative effect in the time-weighted scenario. As expected, the bigger the hike, the worse the performance of the index. In an event-weighted stress test, a U.S. 10-year yield of 3.5%, 4% or 6% will result in index performances of -2.38%, -4.63%, and -13.59%, respectively. Time-weighted stress tests of a U.S. 10-year yield at the 3.5% and 4% levels return similar results to the event-weighted scenarios; however, at the 6% level, the time-weighted return is nearly 30 bps lower at -13.88%.

Sector Analysis

The breakdown of the index provides some insights into the reaction to the rate increases. More than 60% of the Barclays Global Aggregate index consists of treasury and government-related bonds, corporates make up 18.5%, and the securitized sector represents 15%.

The securitized sector would be hit hardest in both the event- and time-weighted scenarios. In the extreme event of U.S. 10-year yields reaching 6%, the return of asset- and mortgage-backed bonds plummets to more than -40%. Observations of past performance predict that ABS and MBS would perform worse at the time of the bond yield increase than if the shocks occur during more recent periods:

Country/Region Analysis

Around 44% of the Barclays index is exposed to USD-denominated bonds; 37% of those are issued in the United States, the rest by other countries. The other bond currencies in the top five by weight are EUR, JPY, GBP, and CAD. We observe similar trends for hard-currency bonds, with USD bonds hit the hardest, followed by GBP, JPY, CHF, EUR, and CAD. Less stable currencies like the RUB and CLP are the outperformers under all three scenarios. RUB performs best in the event-weighted stress tests and SEK leads performance in the time-weighted stress tests.

In the case of event-weighted shocks, there is a very strong positive impact on bonds denominated in RUB, AUD, CLP, NOK, NZD, and PLN. The higher the yield of the U.S. 10-year treasury, the more positive the return of bonds issued in those currencies. For example, the return on bonds issued in RUB is predicted to increase to 4.8% when the U.S. 10-year bond hits 3.5% and 27% when the bond jumps to 6%. It might be worth noting that the exposure to RUB in the composite is quite small (around 0.07%). RUB-denominated bonds are notably high-yield, high-risk treasury bonds. The weakening of the RUB following the 2008 financial crisis and the imposition of international sanctions in 2014 have led the government to issue high yield debt. As U.S. rates increase and the USD strengthens, U.S. hard-currency bonds become more expensive and less profitable than issuing debt in RUB. For investors, holding the bonds purchased on discount when the RUB was cheap still promises higher returns than investing into more expensive USD hard-currency bonds for lower yield.

If most recent events are given a higher weight, i.e. the shock is time-weighted, we observe that there are positive returns for bonds issued in ZAR, SEK, CLP, RUB, and NOK. The NZD debt goes from slightly overperforming in the event-weighted scenario to slightly underperforming when taking most recent returns into consideration. While the positive effect is much smaller for RUB and AUD compared to the event-weighted shocks, more recent tendencies predict strong positive impact on SEK, CLP, NOK, as well as less stable currencies like MXN and MYR.
 

A U.S. yield increase has a far more destabilizing effect on bonds issued in CHF, JPY, and CAD in a time-weighted shock environment than in events like the global financial crisis. USD debt is the lowest performer in both environments. EUR-denominated bonds are not hit as hard, however, as they would have been when past events are weighted stronger. The GBP appears to have more stability in recent years, while GBP hard-currency debt fares better than during periods of global instability.

If we take into consideration that many countries issue hard-currency bonds (countries in emerging markets predominantly issue USD-denominated bonds), we can see how higher U.S. 10-year treasury yields will affect local markets. In times similar to the global financial crisis, Europe as a region performs worse but has the best performance relative to the other regions based on current returns. The Americas and Asia are the worst and second-worst performers, respectively, in time-weighted scenarios.

Impact on Equity Markets

We also looked at the impact of increases in the U.S. 10-year bond yield on global equity markets. Counter to the theory that equity markets fall as interest rates increase, our models saw positive gains for most countries. The S&P 500 showed the highest gains (both event- and time-weighted) when comparing North American, European and APAC developed market indices. There were some exceptions, notably the Nikkei 225, which is predicted to fall in response to an increase in U.S. rates.

Conclusion

With U.S. interest rates expected to continue rising over the next 18 months, the rest of the world will be watching. While we can’t predict the impact on global equity and bond markets with 100% certainty, using historical data to model the reactions at a country and regional level helps to give some guidance looking forward. The stress tests we performed here predict declines for fixed income markets, while global equity markets appear more attractive. The key determining factors for the eventual outcome will be the magnitude and cadence of rate changes.