Ambrogio Cesa-Bianchi and Fernando Eguren-Martin
In March 2020, the Covid-19 (Covid) outbreak turned the world upside down. With economies virtually shut, financial markets were an exception and remained open. However, it was not business as usual for them: the increased need to meet immediate obligations, and a more generalised increase in risk aversion, led investors to liquidate positions in favour of hard old cash. In a recent Staff Working Paper we pose that investors did not seek any type of cash but rather that the world witnessed a ‘dash for dollars’. We show that the resulting race for dollars went beyond exchange rate markets and led to selling pressure on dollar bonds in corporate bond markets, which experienced particularly large increases in spreads.
The main objective of our research is to exploit the heterogeneous spread dynamics across corporate bonds to learn about the nature of the underlying shock hitting financial markets during the Covid turmoil, as well as its transmission mechanism. We analyse changes in spreads from a large global data set of corporate bonds issued by more than 2,000 firms in more than 50 countries. In particular, we link the dynamics of these bond spreads during March 2020 to their underlying characteristics, including their maturity and currency of denomination. Importantly, by focusing on within-firm variation, we are able to avoid selection issues that would arise if firms of certain type tended to issue bonds of certain characteristics.
We analyse the change in spreads between end-February, when market conditions were still relatively tranquil, and 20 March, at the height of tensions and just before the Federal Reserve intervened in corporate bond markets. Three main results stand out. First, for dollar bonds, spreads increased more at short maturities, in line with previous findings. One possible interpretation for this result is that markets experienced a ‘dash for cash’, in which investors in need of cash started by selling their most liquid assets in order to minimise fire-sales losses – generating what has been dubbed a ‘pecking order of liquidity’. Given that short-term bonds tend to be more liquid, selling pressure could put downward pressure on their prices (upward pressure on their spreads), which could explain our first finding.
Second, spreads of dollar bonds increased by more than non-dollar bonds, independently of their maturity. This finding could still be rationalised by the pecking order of liquidity mechanism laid out above. Liquidity is a defining feature of an international currency such as the US dollar. Thus it is possible that investors sold their liquid dollar assets first, putting downward pressure on their prices relative to non-dollar assets.
However, thirdly, we also find that for non-dollar bonds it is long-term bond spreads that saw larger increases. This reversal in the relation between maturity and bond spreads (relative to the dollar sample) is at odds with a general dash for cash-driven sell-off of liquid assets. In fact, one would expect more liquid short-term bonds to be subject to a larger selling pressure (ie a larger increase in spreads) for each currency in isolation. Something else must be going on.
What can explain our empirical findings? Our interpretation is that the world did not witness a dash for cash in general but a ‘dash for dollars’ in particular. Investors did not sell dollar assets because of their superior liquidity but because of the need to obtain cash dollars, for reasons that are ultimately related to the role of the US dollar as a dominant currency in the international monetary and financial system.
There are two prominent features of the dollar dominance that could be related to our findings. First, the US dollar is the most widespread currency for international security issuance and cross-border banking, as well as portfolio holdings of international investors. Dollar dominance therefore implies that, globally, a larger share of balance sheets is denominated in US dollars than in any other (non-domestic) currency. This, in turn, means that agents may face more obligations to be met in US dollars than in other foreign currencies. If these obligations have to be suddenly met (or, equally, if the probability of having to meet these obligations in the near future increases), a demand for cash dollars would arise. Another defining feature of a dominant currency is its perceived safety, which is derived from the fact that it offers a natural hedge in times of crisis. If agents in need of precautionary cash have the choice of securing cash in dollars or in other currencies, they could choose to do so in dollars as the US dollar tends to appreciate in crisis times (ie it is a so-called safe haven currency).
To further corroborate our interpretation, as well as its generality beyond the Covid episode, we run an additional exercise exploiting spread dynamics around a different period of financial market stress, namely the global financial crisis. While less pervasive than today, the dollar still played a dominant role back in 2008. If our interpretation is correct, we should therefore observe similar dynamics to those obtained over the Covid period. Indeed, we find that our results hold around the Lehman collapse, lending support to the dash for dollars interpretation, and suggesting it could be a more general feature prevalent during periods of stress.
In summary, we show that currency of denomination is a key variable for explaining corporate spread heterogeneity in periods of stress. In particular, the US dollar’s role in explaining corporate spread dynamics speaks to its status as dominant currency in the international monetary and financial system. What are the lessons policymakers can draw from our findings? A dash for dollars can have financial stability consequences for the balance sheets of both investors holding US dollar assets and corporates displaying US dollar liabilities. For example, investors with holdings skewed towards US dollar-denominated assets could see larger losses than otherwise during periods of stress. In turn, corporates with funding needs would need to pay higher prices for financing if issuing bonds in US dollars (typically done to cater for investors’ demand). Additionally, this feature of US dollar-denominated corporate bonds could affect their pricing outside crisis times, as investors tend to require higher risk premia to hold assets that fall in value during ‘bad’ times. The provision of US dollars abroad by the Federal Reserve during crisis times, including via the expansion of its network of central bank swap lines, shows that these issues are indeed part of the set of concerns under the consideration of policymakers.
If you want to get in touch, please email us at firstname.lastname@example.org or leave a comment below.
Comments will only appear once approved by a moderator, and are only published where a full name is supplied. Bank Underground is a blog for Bank of England staff to share views that challenge – or support – prevailing policy orthodoxies. The views expressed here are those of the authors, and are not necessarily those of the Bank of England, or its policy committees.