euro europe italy mario draghi

Back To 2011? Not Exactly.

Don't panic.

2.5-sigma bund yield increase of 9.6bp Wednesday, red Euribors full circle to pre-Italy debt fears, Eonia now pricing total depo-rate increase of 24bps by end of 2019 and 61bp by end-2020 (using 1m Eonia forward spread).

That’s from Bloomberg’s Tanvir Sandhu and it’s a nice summary of how consequential Wednesday was in terms of shaping the narrative around APP wind down.

That narrative had become hostage to political developments in Italy as some speculated that Draghi would be forced to hold off on slapping a September sell-by date on ECB QE in the interest of ensuring Italy retained market access in the event the country ended up holding new elections.

Obviously, ECB purchases are key when it comes to Italian debt and there was no shortage of speculation as to whether Draghi was trying to send some kind of message to the new government in Italy by deviating from the capital key in favor of Germany last month. When you think about that theory you do want to note that there are technical considerations at play. Here’s a dispassionate take from BNP:

The key focus in May’s data fell on deviations versus the capital key, with Germany’s monthly net purchases coming in 5.3% above the capital key, the highest reading since the start of QE. Meanwhile, monthly net purchases in Italy and France came in 1.5% and 1.7%, respectively, below the capital key, ending months of deviation above the capital key (Chart 3). The absolute changes in monthly net purchases were revealing and suggested that the aforementioned dislocations versus the capital key were likely caused by some of April’s PSPP reinvestments in Germany being carried over to May. The monthly net purchase amounted to EUR 6.9bn in Germany, much higher than the EUR 4.8bn/month average in January-April this year. Meanwhile, in Italy and France the monthly net purchases amounted to EUR 3.6bn and EUR 4.2bn, respectively, broadly in line with the average EUR 3.6bn/month and EUR 4.1bn/month so far this year. Marginally slower monthly net purchase in Italy versus April (EUR 3.6bn in May versus EUR 3.97bn in April) could have been due to liquidity conditions.


You can write your own conspiracy theories there (or preferably not), but what is indisputable is that without ECB support, there are serious questions about what the clearing price would ultimately be for Italian sovereign debt and with 90% of Italian credit trading inside of similar maturity Italian govies, there are concurrent questions about whether spreads would ultimately close the gap with the sovereign in the absence of the CSPP bid, which effectively acts as insurance on Italian corporate obligations.

Well, now that Italy actually has a government (albeit a euro-skeptic government), the prospect of new elections coming at virtually the exact same time as the ECB was planning on winding down APP is off the table. Once the specter of fresh elections was banished, Italian bonds and equities rallied, but as noted on Wednesday, that rally has now stalled as investors and traders attempt to make sense of what comes next and as everyone comes to terms with the reality of the situation which, simply put, is that Five Star and League are going to be at odds with Brussels sooner or later.

Still, things have calmed down enough for folks to start doing postmortems of last week and that’s exactly what BofAML has done in a new piece.

The bank has been keen to note the obvious lately, which is that a key variable for keeping credit spreads suppressed (and for keeping cross-asset volatility suppressed more generally) is ensuring that rates vol. remains anchored. Part and parcel of that effort is effective DM central bank forward guidance. Here’s the bank’s take on that in the context of the Italian drama:

We have been highlighting for months that there are two key risk metrics to follow when assessing credit spreads: rates vol and peripheral spreads to Bunds. Even though rates vol has remained close to the lows (chart 2) we see that peripheral spreads are heavily hit especially in the case of Italy on the back of rising political risks.


As to the “are we back to 2011?” debate (i.e., are we now right back in the eurozone sovereign debt crisis), BofAML notes that while the trajectory of “lo spread” would suggest that we are, credit has remained well behaved. To wit:

The recent market reaction on BTPs has been reminiscent of the 2011 peripheral crisis; chart 4 depicts the trends then and now. Note that not only are the patterns identical, but so are the spread levels. One could think that we are back in 2011…

Not exactly; as this sell-off finds credit more “insulated” vs. rates, we think. It feels that the recent developments on Italian politics have hit the rates market more than credit. Both corporate bond spreads (chart 6) and CDS spreads (chart 5) are much tighter than what we have experienced in 2011.


What accounts for credit’s resilience? Well CSPP, probably. And that’s the irony here. The same thing (i.e., the ECB) that has kept a lid on Italy’s borrowing costs is also ensuring that bloc-wide credit spreads don’t blow out. The chart in the left pane below is straightforward, but the one in the right pane depicts the extent to which iTraxx is basically bulletproof vis-à-vis the BTP-bund spread:


And here’s another way of visualizing the same phenomenon only using spreads on financials (cash and CDS) to illustrate the point:

Resilient Fins

Finally, here’s BofAML resurrecting their “selloff indicator” to suggest that no one should “panic”:

We find that more than half of the market (52%) was under significant pressure (names where spreads widened by more than 10bp over a four weeks period, while the index was also wider). This is the highest reading – but comparable – since Brexit (June 2016) and the February 2016 sell-off, but much lower than the 2011/2012 European sovereign crisis (chart 1).


Another interesting fact is that even though a significant proportion of the market was under stress, dispersion has increased. This highlights that the widening was not uniform across the market, as kurtosis increased. High kurtosis indicates higher dispersion in the market, and thus less contagion risks. We think that this is clearly reflective of the ECB “put”


And see that last bolded bit brings us back full circle to the signaling from the ECB on Wednesday. All of this hinges on how effective Draghi is at convincing the market that he’s capable of calibrating forward guidance around QE to match the evolution of market expectations.

As long as the APP forward guidance (and remember, the ECB won’t hike until at least a couple of quarters after QE ends, so forward guidance on APP is effectively forward guidance on rates, with the only ambiguity being how quickly you think they’ll be inclined to hike once asset purchases end – see Villeroy on that) Italy can be quarantined, so to speak.

That’s it. That’s the end of this post. If you were waiting on some kind of doomsday prediction, go somewhere else. We’re “serious analysis with a healthy dose of cynicism and a dash of profanity” not “doomsayers with an occasional dash of seriousness”.



1 comment on “Back To 2011? Not Exactly.

  1. monkfelonious

    Kurtosis? Jeezus K. Ghandi. I hope I never get kurtosis….

Speak On It

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Skip to toolbar