As anyone who has been following the VIX, US CDS (which is quite interesting as the US catastrophe trade appears to have become selling CDS to fund gold purchases in euros: more on that eventually), or stock markets in general has grown to appreciate all too well, no matter the amount of perceived risks, the market continues to shrug off any bad news: after all, the Bernanke put means that the greater the systemic shock, the higher the likelihood that the Fed will get involved yet again and push up all risk assets. However, the same can not be said about the dollar. The currency which in 2011 has traded like anything but the world's reserve currency is less than 1 point away from 2009 lows. But that could be just the beginning. Citi's head of FX has released a not warning about the potential coming avalanche to the greenback should debt ceiling negotiations hit a snag: "what we are looking at here is very much the tail risk event that the
debt ceiling negotiations unexpectedly hit an impasse. The question is
what the impact would be on USD." Englander's summary observations: 1) The USD will be in big trouble if investors get the sense that the
debt ceiling negotiations have gone beyond the expected choreography
into a zone where there is perceived risk to US credit; 2) More broadly, we think FX markets are increasing the attention they pay to fiscal sustainability relative to monetary policy; 3) The FX response may be non-linear so G10 countries may have a
false sense of security in seeing little FX response to deterioration so
far. Then again, perhaps a major step down in the dollar is precisely what the Fed wants...
Englander's summary view:
- The 2011 budget impasse was resolved with little markets impact
- If a breach of the US debt ceiling comes into question the impact will be larger
- Since September 2009, a 100bps increase in CDS has been associated with 8% currency weakness …
- … but there also is a level effect as poorer credits have faced larger FX pressures
The USD dodged the 2011 budget bullet last weekend and is now facing the debt ceiling cannonball. Although the debt ceiling is not normally considered a tool by which fiscal consolidation is achieved, it seems likely that there will be a fair amount of brinkmanship before the debt ceiling is raised. Investors and most economists expect political posturing as the debt ceiling debate drags on through late May and June, but no event that affects perceptions of US credit quality. So far US yields are showing no pressure and US CDS is trading well within the range of the last 15 months.
So what we are looking at here is very much the tail risk event that the debt ceiling negotiations unexpectedly hit an impasse. The question is what the impact would be on USD.
Below we present a cross-section analysis across more than 30 major currencies on what impact a deterioration in credit has on the currency. The analysis covers the period September 2009 to the present – September 2009 was the trough in spreads between financial crisis related sovereign credit concerns and the European sovereign credit blow-up, so it gives us a good base for comparison.
We find a surprisingly strong impact of credit deterioration on currencies. A 100bps increase in CDS has been associated with 8% currency weakness over this period, but we also find that countries that started the episode with higher CDS tended to fall more or appreciate less than others. Having a September 2009 CDS level 100bps higher in the cross-section is associated with 7% less subsequent appreciation, given no further CDS deterioration. Hence poor credit seems to represent a headwind to appreciation beyond any additional deterioration.
We have three takeaways from this.
1) The USD will be in big trouble if investors get the sense that the debt ceiling negotiations have gone beyond the expected choreography into a zone where there is perceived risk to US credit;
2) More broadly, we think FX markets are increasing the attention they pay to fiscal sustainability relative to monetary policy;
3) The FX response may be non-linear so G10 countries may have a false sense of security in seeing little FX response to deterioration so far.
The debt ceiling negotiations
We have little to add on the debt ceiling negotiations except to reiterate that there is very little evident concern in either FX or FI markets. Both 10- and 30-year yields are in the year’s range, the 30-year Treasury auction on April 14 was very well received and there is no stress evident on CDS. Last weeks budget negotiations probably had a small negative impact on USD, and the concern about the process is a lingering negative, but the impact of any concrete debt ceiling risks would be much higher.
Impact of credit deterioration
We regress the Sep 2009 to April 2011 change in the value of 33 currencies against the beginning level of their sovereign CDS and the change in the sovereign CDS, the correlation of the currency with the S&P and a dummy to distinguish between G10 and non-G10 currencies.
The strongest association by far is with the CDS variables. The results imply that a currency with a 100bp higher CDS in Sep 2009 tended to appreciate about 8% less through Apr 2011, even if the CDS did not move, while a currency with its sovereign CDS rising by 100bps tended to appreciate about 9% less. (We ran our regressions using USD as a base but in a cross-section regression the base currency does not matter.)
The measure of riskiness that we used, the correlation with the S&P, was marginally significant, indicating that over this long period, the CDS coefficient did not reflect the appetite for risk to any great degree. Being in the G10 had a modest negative effect and being heavily correlated with risk had a marginally significant positive effect but they did not affect the coefficients or results greatly.
Figure 1 presents the actual currency changes as well the changes predicted by the model specified above. For such a simple equation it explains a fair amount of the currency variation. Most importantly for the USD it seems to have quite a bit of explanatory power for the currencies that have performed the poorest versus those that have performed the best. In some regressions we included levels and changes in swap spreads, but these did not seem to provide significant additional explanatory power.
We did not include the EUR in our regression since the combination of national sovereign risks was not necessarily linear. However, we did calculate a debt weighted average of CDS and CDS changes and used the estimated coefficients to get an estimated EUR impact. Interestingly the actual change in the EUR very much matched the predicted and the 96bp increase in debt-weighted CDS suggests that the EUR would be about 8% (11 big figures) higher had the EUR sovereign CDS retained their September 2009 level. That would explain much of the deterioration in EURXXX crosses against commodity and other risk-correlated currencies.
Visually we observe that the currency effect seems to become significantly larger when the CDS level was above about 90bp. Moreover the currency impact seems to kick in as soon as there is any deterioration. We do not want to emphasize this, but it comes back to the risk that there is a false security in having seen no effects till now.
This provides preliminary indication that FX markets have come to focus on debt and creditworthiness in addition to the standard macro variables. It suggests both potential upside for the EUR and other currencies that get their sovereign debt situation under control and significant downside for USD and JPY if markets ever begin to price in concrete risk that debt will become unsustainable.