Even though they are far from their historical troughs of 2009 and 2011-2012, EUR/USD cross-currency (XCCY) basis swaps remain stubbornly negative. In this post I explore the potential drivers of the spreads and their likely directions for the next few months.
Cross-currency basis swaps are quoted as USD Libor versus the Euribor plus/minus a spread. According the BIS, "a basis swap spread of x basis points indicates that a counterparty wanting to swap U.S. dollars for a foreign currency loan must pay x basis points above/below the benchmark floating rate on foreign currency funds in return for US dollar Libor." Given the EUR/USD 5-year basis, one must invest EUR and earn EUR-34 in order to borrow at U.S. Libor flat.
Corporate Issuance and XCCY Basis
The relationship between issuance hedging and cross currency basis swap works as follows: If a European company issues a bond denominated in USD (a "yankee") and wants change the nature and currency of its liability payments, it will first go for a traditional fixed/Libor interest rate swap, where it will receive fixed USD and pay Libor USD. It will then enter into an FX cross-currency swap to hedge its stream of interest and principal payments from USD to EUR. It will lend USD and borrow EUR, receiving USD Libor and paying Euribor + a basis. A rise in swapped USD issuance will lead to a narrower (less negative) basis since the implicit borrowing in EUR increases.
On the other hand, a US company issuing abroad will need to change is floating EUR liability into USD, thus pay USD Libor and receive Euribor + the basis. A rise in swapped EUR issuance will lead to a wider (negative) basis since the implicit borrowing in USD increases.
- USD/yankee issuance results in a narrower negative basis
- EUR/reverse yankee issuance results in a wider negative basis
The chart below would confirm this view, even though the percentage of yankee or reverse yankee bonds is unknown. The 5-year XCCY basis tends to widen (become more negative) when the relative issuance of corporate bonds is stronger in Europe than in the U.S.
Two serious considerations:
- Why didn't the relationship hold before the crisis?
- An opportunistic approach would suggest the arbitrage of issuance currencies is driven by the level of the basis. Therefore, the relative issuance flows would follow, not drive, the current level of BS. We are facing a chicken-and-egg problem here.
Carry Trade and Basis
An FX carry trade strategy rests on a payoff that involve the expected change in the currency pair and the (positive) spread between interest rates of the foreign (*) and funding currency. Even if this formula is close to that of a XCCY swap, the link between the basis and the level of the exchange rate might be dubious since it is well known that forwards are very poor forecasters of future spot levels.
Yet, the chart below shows that in the past, the total return of a rolled carry-trade position on the EUR/USD has been closely linked to the short-tenor XCCY basis. Interestingly enough though, at a time when the EUR has been the favored funding currency for FX carry trades, thanks to the ECB's QE, the link seems to have weakened; the payoff of a EUR-funded carry trade has been more loosely connected to the XCCY basis swap.
The reason why carry strategies and the basis of cross currency swap might go hand in hand is linked to a third factor: the relative supply of money.
1. The chart below shows that the huge increase in the relative supply of euros by the ECB weakened the EUR/USD. The timing is not perfect, but the directionality is decent.
2. The rise in the relative supply of euros against USD could also be a cause for a wider (more <0) XCCY basis (since one can borrow in EUR + swap to USD - akin to reverse yankee issuance). This is what the chart below implies. If the ratio of the Fed to the ECB balance sheet is an indicator or the relative supply of each currency (assuming for instance that the credit multiplier is not gripped):
- an increase in the ratio would mean a provision of U.S. dollars: lower basis (less negative)
- a decline would either mean a dearth of U.S. funding of an excess supply of Euros: wider basis
Much more than carry trades or FX moves, the relative provision of EUR and USD by central banks looks like a serious candidate for explaining XCCY basis swaps. The stability of the relationship between balance sheet ratios and the BS XCCY basis since 2008 is telling. It might indeed appear almost obvious that a currency that is abundantly provided by a central bank would come along with a wider basis.
The Japan example of the late 1990s may suggest otherwise yet, as the widening of the Fed/BOJ ratio led to a temporary, not structurally, wider XCCY basis for the USD/JPY.
As often with XCCY swaps, would-be fundamental drivers tend to fail to provide a robust, long-lasting structural explanation. As we have shown in a previous post, there are several factors are not only time-varying but can also vary in nature.
One of those risks is the bank risk. The dollar shortage of both 2008 and late 2011 where at the core of the strain on XCCY basis swap. The relative performance of banking stocks on both sides of the Atlantic fails to provide any clue on what has been going on since (I use S&P 500 and Euro Stoxx 50 bank indexes).
The issue with the time-varying nature of explanatory factors is that of potential ad hoc modeling. A simple model based on money market curves spreads and the CDS Subfin Markit index would capture more that 70% of the total variance of the XCCY swap since 2011, but as can be seen below, the explanatory power would have faded since late 2014.
We thus need a new, likely temporary, scapegoat. Introducing the ratio of the Fed to the ECB balance sheets increases significantly the explanatory power of the model. It would suggest that as long as we have diverging trends in the sized of central banks' balance sheets, the EUR/USD XCCY swap basis will remain deeply negative.
The model of 5-year XCCY basis swap presented below is based on several drivers mentioned above: the spread between U.S Libor (3-month) and the Euribor of the same maturity; the relative return of Europe vs. U.S. banks (bank risk); the ratio of the Fed-to-ECB balance sheet (relative supply of currency); an index of political uncertainty in Europe (capturing Greece's fate notably); and the European CDS/Cash basis (proxy of the buoyancy of the primary market).
The output is visible in the chart below and provides a decent explanatory power. The model does not fail to capture the entirety of the recent decline in the XCCY basis.
Assuming that: (1) the ECB will increase the pace of its PSPP as soon as December (from 60b EUR per month to 80b), (2) the Fed will tighten as early as December 2015, (3) there will be no significant under/over performance of U.S. banks against their European counterparts, and (4) the overflow of corporate issuance will ease somewhat in Europe (tighter CDS/cash spread), there is room for a little contraction of the EUR/USD XCCY basis, but it should still remain far from its pre-crisis levels.
This post provides an update of the recent developments of the EUR/USD cross currency swap market. I show that the drivers of the spread present a major problem of inconsistency. Their explanatory power is either time-varying or fraught with some dubious causality. In spite of those hurdles I propose a simple model of the EUR/USD XCCY 5-year basis. Based on my forecasts for the major inputs, we can expect a slight tightening for the next few months without significant reversion towards the pre-crisis levels.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.