The Market Dollar & The IMFS Dollar

With a lot of recent talk about the USD system, and seeming interconnections between US current account deficits, the world’s demand for “safe assets,” USD strength and consequent deindustrialization in the US, I thought it might make sense to look at different USD and current account regimes in the last 50 years. My sense, per the below chart that looks at the real broad USD TWI and the US current deficit is that there are broad multiyear- periods where USD strength coincides with large c/a deficits (i.e., the rest of the world wants the USD exposure from funding the US) and others where RoW does not want that FX exposure. USD strength typically coincides with widening US c/a deficits (as seen in 1981–85 or 1995–200) when accompanied by positive political economy or technological shocks. 1981–85 was marked Reagan tax cuts/unionbusting + Volcker Real Rates and 1995–00 by the Clinton internet boom in the second. Conversely, there are long periods where a falling USD suggests that the rest of the world does not want that exposure to funding widening US c/a deficits and reacts by selling USD, even if it continues to buy US assets — as happened in the early 1970s, the early 1990s and 2001–06. Since throughout the period, the USD has remained “central” in the international monetary and financial system, I believe it is inappropriate to conflate USD centrality with USD strength.

One corollary is that reserve accumulation, typically treated as both a sign and a cause of USD strength, often occurs precisely because the global private sector does not want to accumulate assets or is shedding net assets in a currency. This is not true of the USD alone — think, e.g., of BoJ purchases of USDJPY in the early 2000s or SNB activity in €. I believe this also applies to the 2002–08 period that saw large increases in global reserve accumulation (particularly in USD) by East Asia and by the large petroleum exporters even as the global private sector seemed especially eager to limit or cut USD exposure. Examples of this phenomenon that I remember from the time include not just “Gisele wanting to be paid in Euros, hahaha” but GCC private actors buying EUR hand-over-fist and US cross-border equity funds leaving their overseas exposure significantly underhedged or unhedged.

In contrast to the above, durable, long-term USD rallies seem to be associated with a higher private sector demand in RoW for USD assets, typically driven by a positive rate of return or terms of trade (shale) shock in the US. Conversely, refluxes back into USD driven by RoW covering seeming “USD short” (but actually more like USD square+liquidity/maturity shorts) as in 2008 seem to lead to more violent but shorter-term USD rallies, as in 2008–9 until offset by Fed liquidity. The absence of such Fed liquidity provision would have a Mutually Assured Destruction quality via fire sales of US assets and systemic linkages across the global banking system.

In terms of common market narratives, thinking in terms of the famous USD Smile (i.e., the proposition that USD does well when US doing better than RoW or things are falling apart), it seems that USD rallies in the RHS of the USD smile (US doing well) are more robust and long-lasting. Conversely, the LHS of the USD smile seems to be both shorter in duration (if more violent) & its occurences are more recent phenomena (i.e., not 1970s, late 1980s, 1993–95), rather than extending over entire post-BW era.

Also, the LHS of the USD smile seems to depend on underlying asset-liability mismatches, which IMO reflect maturity/liquidity mismatches rather than outright USD shorts. Thus, the post Lehman USD spike (a classic smile LHS) is often attributed to the higher safe-haven qualities of the USD. But it was more of a scramble to replace the USD funding that European banks relied on to support their ostensibly safe US MBS etc. assets.

Those US assets were perceived ex-ante as being credit-safe, but from the point of view of an EZ investor with € liabilities, they could be FX-safe only if funded in USD or hedged. Instead core EZ found its unhedged ostensibly safe assets in the Eurozone periphery.

This in turn leads to the idea that a safe asset is one that protects from these risks — credit, duration, inflation (which may be different from duration), liquidity, FX, redenomination. There is no such asset that does this for all investors around the world at the same time.

So I would argue that asset safety for an investor is related to the nature of his/her liabilities. In this sense, USD centrality is linked to its safety — the USD is the world’s dominant cross-border safe asset because it’s the world’s dominant denomination of cross-border liabilities. HOWEVER, this is more true for EM with larger portion of liabilities in USD, than it is for DM (+large surplus E. Asian EM) that have more of their liabilities in local currencies. In these instances (EZ banks, JP/TW lifers), USD liabilities are “manufactured” alongside these holdings to make USD assets safer.

Another category of FX-related cross-border liabilities relates commodity importers due to the centrality of USD pricing. There are two distinct periods (1970s, 2005–08 +2009–12) when USD assets were perceived as “unsafe” vs. those liabilities because the Fed was seen as unresponsive to inflation in an international consumption basket whose composition was very different from price stability in the domestic consumption basket targeted by the Fed.

One of the market impacts of this lack of faith in the Fed’s commitment to global (rather than purely US) price stability was reserve diversification into AUD and CAD in the 2000s. Conversely, the USD rally that began in 2014 was arguably not just an artifact of ECB QE, but also of shale — a US-based technological revolution that lowered the US’s external financing requirement, and exorcized worries that USD was a poor store of international value.

The contrast between USD behavior 2002–08 (widening c/a deficits + weakening USD) and post-2014 behavior (narrower c/a deficit due shale oil + USD strengthening) also suggests to me that the US now may have conventional Dutch disease rather than exotic financial variants of same.

The bottom-line is that I believe that there has been too much of a conflation of the related but analytically distinct concepts of USD centrality; safety; privilege (a term associated with Jacques Rueff, and coined in the 1970s by Valery Giscard d’Estaing to refer to US ability to settle deficits in currency it printed); and hegemony (which I take to mean the US’s legal control of the pipelines through which the world’s dominant invoice currency flows). And to reiterate the point made at the outset — over the 50 years since the breakdown of Bretton-Woods 1, the USD in the market has had a more interesting and tumltuous history than the USD of the IMFS.