The (un)surprising weakening of the dollar and what could change it

In recent months, the yield spread between nominal 10-year US Treasuries and 10-year Bunds has increased significantly. Somebody who has not been following currency markets over that period could be forgiven for thinking that this should have led to a strengthening of the dollar versus the euro. After all, when US yields are rising and those in Germany and the other eurozone countries are more or less stable or even declining, one could expect capital to flow from the euro area to the US. This has clearly not been the case, as shown in chart 1.

Obviously, many theoretical arguments can be put forward why this assumption should not work, at least not all the time. One is that the yield difference may reflect different developments in terms of inflation and inflation expectations. In such case, it does not justify an appreciation of the higher yielding currency. However, chart 2 shows that the real bond yield differential has increased as of late, but this has not stopped the dollar from weakening. Moreover, it appears that since 2011, the correlation between the real yield difference and the EUR/USD exchange rate has been positive, rather than negative, as the capital flow model of exchange rates would imply.

This shows the complexity of explaining and forecasting exchange rates. Certain theoretical models work sometimes but not all the time[1], so the challenge is to assess which model may apply when producing a forecast or taking an investment decision. Several factors have played a role in the depreciation of the dollar in the past several months. One, the new strategy of the Federal Reserve of targeting average inflation can be considered as a commitment of being dovish for longer whereas there are serious doubts about the ECB following the Fed’s lead when announcing the outcome of its own strategy review later this year. Two, the rising real yield differential could be associated with the huge US fiscal stimulus and the ensuing rising borrowing requirements. In that case, higher yields do not warrant a dollar appreciation. Three, historically, the combination of a very accommodative US monetary policy and improving prospects for the global economy typically have caused capital flows from the US to the rest of the world. This is quite likely again the case at the current juncture, witness the strengthening of several emerging currencies versus the dollar and the strong recent performance of emerging equities. Such a development could indirectly cause an appreciation of the euro versus the dollar to the extent that eurozone investors would be less inclined to invest in emerging markets than their US counterparts. Four, the short-term interest rate differential between the US and the eurozone has narrowed significantly in 2020. For European investors with big historical holdings of US assets, hedging this currency exposure has become very cheap. This increased hedging of existing exposures is likely to weigh on the dollar[2]. Five, the fair value of the dollar versus the euro –based on the long-term relationship between the exchange rate and economic variables- is around 1.34[3], so this level can act as a magnet when other factors are ‘pushing’ the currency in this direction. We expect these factors to remain in place for the foreseeable future and hence the euro to appreciate further versus the dollar towards 1.25 at the end of this year and 1.30 at the end of 2022.

What could change the direction? In theory, a sudden change in guidance from the central banks could play a role. However, this is very unlikely. The Federal Reserve will not be in a hurry to adopt a more hawkish tone whereas the ECB’s policy is already very accommodative. This sets the bar very high for more accommodation. Another factor could be a sudden drop in risk appetite of international investors. A move into a ‘risk-off’ environment is typically associated with a strengthening of the dollar on the back of US investors reducing overseas exposure. However, unless there would be an abrupt, considerable deterioration of the economic outlook, such a risk-off phase should be short-lived. A more intriguing question is whether a further widening of the yield differential and a weakening of the dollar could end up becoming self-correcting. For investors who intend to hold on to their bonds until maturity and are not subject to mark-to-market accounting, at some point the yield gap will become compelling and offer enough protection against a potential further depreciation of the dollar. A 200bp yield gap versus Bunds for a 10-year maturity offers a lot of downside protection for the currency. For marked-to-market investors, this argument does not apply. Moreover, they may consider US treasuries as being riskier than Eurozone bonds based on the outlook for GDP growth and inflation. The biggest risk for a change in direction would be a repetition of the ‘taper tantrum’ of 2013 with the Federal Reserve starting to point towards a possible beginning of the normalisation of its policy. However, such a change in guidance is not to be expected anytime soon: the Fed wants to overshoot its inflation target first.

[1] This point is analysed empirically in great detail in “Exchange rate prediction redux: new models, new data, new currencies”, ECB working paper 2018, February 2017. The authors conclude that “Overall, model/specification/currency combinations that work well in one period will not necessarily work well in another period.”

[2] This argument was provided by Alexander Jekov, FX strategist of BNP Paribas Global Markets.

[3] Calculation provided by BNP Paribas Global Markets.