Before I go into the details of our argument, just a reminder. Certainly, Italy has well-identified problems: sluggish growth, very high public debt, etc. The political crisis, one more, puts Italy at risk. However, we have always had a more nuanced view on Italy, which also has assets that are often forgotten, cf. our MyStratWeekly “Advocacy for Italy” of 22 February 2021. The Italian economic situation is more complex than what is too often presented.
Rates: political risk is only very partially taken into account
Mario Draghi has therefore bowed out, Italy is heading for elections on 25 September with in all probability a very fragmented chamber that should not lead to a stable government. In short, the commedia dell’arte politique, put in parentheses by the Draghi episode, is back. The “Draghi premium” on interest rates has therefore partly disappeared and Italian rates have risen significantly as shown in the chart below.
However, account must be taken of the fact that peripheral spreads have generally deviated since the beginning of the year with the withdrawal of the ECB and the end of the QE. A risk premium model in this market leads us to believe that peripheral spreads were abnormally contracted during the QE period (which was also one of the QE goals) and that they have returned to normal. What about Italy? On the chart below we have the Italian spreads as seen on the markets.
The “Europe Risk” is the level of Italian spreads if they had reacted purely according to their usual beta in the market. It is clear that during the period 2018-2019, the last period of significant political stress, the Italian spreads deviated more than expected by the model. This is an idiosyncratic risk specific to Italy and therefore has nothing to do with the movements of other countries. This specific risk premium was then around 100 bps over the period. Paradoxically, at present, there is really no specific Italian risk premium. While spreads have increased over the past year, they are largely consistent with other countries.
This is at least what our modelling with an Italian idiosyncratic risk tells us, which is currently estimated at just under 20 bps. To give a basis for comparison, the specific Italian risk, during the Conte I government, from 1 June 2018 to 5 September 2019, was 95 bps. On the other hand, since the beginning of the year, under the Draghi government, Italy has been treating on average a dozen bps more expensive. So that’s our estimate of the “Draghi premium”.
The fact that Italian spreads have reacted, to a very large extent, in line with the spreads of other countries leads to the conclusion that the market only marginally includes a specific Italian risk. Either the market assumes that the political problem is largely negligible, or it assumes that the TPI announced by the ECB is credible.
Two hypotheses that seem very debatable to us. In the absence of any specific risk, Italian spreads should currently be in the 200-230bps range. At the time of writing, we are at 230. In case of major political risk, as we saw in 2018, the target would be above 300 bps.
Actions: same thing
We use the IT30 index of the thirty largest market capitalizations. This index underperformed the Euro Stoxx by almost 6%. Signs that political risk is therefore taken into account by the market?
Here again, the reality is more subtle. The Italian index is highly concentrated in several sectors, with a very strong weighting in Banks, Utilities and Energy. Reflecting sector performance, the index has only underperformed 1.2% YTD. So the underperformance of the Italian index is much more related to the fact, no luck, that the index was concentrated in the wrong sectors. Here too the specific Italian risk is therefore very marginally taken into account. It should be noted, however, that even if we adjust for the sectoral composition, the index nevertheless underperformed by 2.5% over the last two weeks. The market is starting to wake up!
Oddly, and despite the recent mediocre performance, it seems that the specific risk of Italian politics is not really taken into account. Put another way, in case of a bad surprise (a poll, an untimely announcement of a policy, etc.), the risk is that Italian assets will suffer much more.