Investor concern about European political risk has sharpened further following the surprise outcome in the US presidential elections. That result, following Brexit, is seen as confirming the surge of an anti-establishment movement fuelled by discontent at the impact of globalisation. There are a number of polls coming up, but the Italian referendum on 4 December and its potential ramifications have been in the spotlight recently.
What is the referendum about?
The referendum is on a proposal to change the electoral system of the Senate, which has been already approved by parliament. Italy’s political system would become a single-house system, where only the lower house ratifies a government and approves most ordinary legislation. The current elected Senate would become a House of Regions and Municipalities, with limited and specific legislative powers and no right of veto. The number of MPs in the Senate would fall from 315 to 100.
The reform to the Senate is part of a broader reform to the electoral system that has been on the agenda since Matteo Renzi took office in 2014. Italy has a perfectly symmetrical parliamentary system (a bicameral system), in which the Lower House and the Senate effectively have the same powers. This tends to result in laws bouncing back from one House to the other, often blocking the legislative process and hindering policy making. A reform to the lower house (the Italicum) was already agreed by parliament but a number of appeals have been filed against it at Italy’s Constitutional Court (see box below). If both reforms of the lower house and Senate are passed, it would make government formation easier in the future and make it easier for future governments to implement their policy agenda.
What is the likely result?
The last opinion polls suggest that the No camp has a slight lead. Our running average of the last five polls stands at 40% for No and 37% for Yes. This essentially means that the outcome is too close to call. The gap is within the margins of statistical error. This is especially the case given the recent poor performance of opinion polls in predicting actual outcomes. In addition, there is a large proportion of the population (23%) that is undecided. The way these voters turn will have big implications for the outcome and increase the uncertainty.
What happens in case of a No vote?
There are two significant possible implications of a No. It would reduce the ability of the future governing party or coalition to pass into law its policy agenda. A future government would likely not have control of both chambers. This would make passing ambitious structural reforms more difficult. However, it is uncertain whether Mr Renzi’s Democratic party (PD), or any other party, has such an agenda anyway. On the other hand, a split in the chambers, would also make it difficult for the populist and eurosceptic Five Star movement (M5S) to push through a referendum on the future of the euro.
The other implication would be political instability. Prime Minister Mateo Renzi has flip-flopped on whether he would stay on in his role in the case of a No vote. However, there is a significant risk he would resign. That could potentially trigger new elections in early 2017. However, we think in that case, it would be more likely that a new Prime Minister would be appointed that would lead an interim government. That government may not last its full term (early 2018) and there could be new elections in the second half of next year in any case. An alternative source of instability in case of a No is that the PD may lose support from smaller groups in the Senate, which could mean that a grand coalition (including Forza Italia) would need to make up the new government.
What happens in case of a Yes vote?
A Yes vote would mean that Prime Minister Renzi’s government would stay in place, likely through to the scheduled election in 2018. An optimistic take is that he would then be free to focus all his energies on a major structural reform programme. However, it remains to be seen whether Mr Renzi wants to pursue aggressive reforms. His track record leaves question marks. Since coming to power, his main achievement on economic policy side is labour market reform, but that was not far-reaching. In addition, the Prime Minister still needs to deal with the Senate in the current state, and he has a waver thin majority. So major reform before new elections does not seem likely. However, if he does win new elections and both the lower house and senate reforms are passed, he would be in a powerful position to implement major reforms in his next term.
There is also another scenario following the Yes vote. The forces of populism are unlikely to go away in a hurry. A new election, even in 2018, could still see the MS5 party win. It would then have (again assuming both parliamentary reforms are passed) the ability to execute its policy agenda, though it may still struggle to push through a euro referendum (see below). So a Yes win may be positive in terms of near term political stability, but could raise the risk of a more dangerous scenario in the not too distant future.
What would be the result of any new election?
A new general election is scheduled in early 2018, but as noted above could occur earlier, given the political instability a No vote could trigger. There are two complications in trying to assess the outcome of any new election. First of all we need to rely on the outcome of the current polls, which may or might not be accurate, and in any case may change up to the election. The second is that outcome will also depend on to whether the electoral system will change (fully or partially).
The latest polls suggest that PD would still be the biggest party. PD is currently polling at 33%, while MS5 is at 28%. Still given the uncertainty of the polls and the potential for swings, MS5 still has a realistic chance of winning. If the electoral reform of the Lower House is passed before the new election, then PD or MS5 would emerge as the dominant force in the lower house as the system ensures that winning party has more than 50% of the vote. If the senate reform has also passed (so in case of a Yes vote in the referendum) the winning party in the lower house would have significant power to pass through its legislative agenda. However, if the reform does not pass, dominance in the Lower House may not amount to much, as the Senate would remain able to frustrate the government.
If the electoral reform of the lower house is blocked by the Constitutional court (or if it is significantly watered down) Italy may well find it very difficult to form a new government. This is because the electoral law under which Renzi’s government was elected, which also had a winner premium to help ensure a majority, was ruled unconstitutional by the Constitutional Court in 2014. If this law is not replaced by the Italicum, any new election would be decided by full proportional representation. This means that the winning party would have to form a coalition given current polling. The next biggest party in the polls is Lego Nord (12.1%) followed by Forza Italia (11.5%). Assuming that PD and MS5 would not want to form a coalition, the other combinations look problematic, and would in any case need to involve a multitude of small parties. In this case, a ‘Spain scenario’ where coalition negotiations are drawn out and new elections become necessary would seem likely.
What is Five Star’s policy on Europe?
The Five Star movement is against Italy’s membership of the euro, but not necessarily of its membership of the EU. Last month, Luigi Di Maio, a MS5 party leader in the lower house of parliament who is often seen as a possible future prime minister, set out the party’s position on the euro in a recent interview. He said he would like to ‘see a European referendum on the euro, to see other countries starting to talk about it’. He added that it would be ‘a consultative referendum’. He also noted that Italy should explore other alternatives to the euro and mentioned a return to the lira, as well as the (more fanciful) idea of splitting the eurozone into different currency areas.
Would a No vote start the countdown for Italy’s euro exit?
We do not think that a No vote would increase the chances of an Italian euro exit. It actually might make it less likely. The nightmare scenario for financial markets in simple terms is that a No triggers early elections, MS5 wins, it holds an in-out euro referendum, which leads to a vote for ITEXIT. However, in case of a No vote, the situation would be more complex. As described above, without parliamentary reform, MS5 would struggle to form a government and to pass the legislation to hold the referendum.
In many ways a YES vote followed by a MS5 win in 2018 could actually be the scenario which implies the bigger risk of ITEXIT. This is because MS5 would have more legislative power. However, even then there would be significant hurdles to overcome. The constitution does not allow referenda on pulling out of international treaties, though it does allow advisory referenda. However, to launch an advisory referendum, there needs to be a two-thirds support in parliament currently (though this could change eventually). Even assuming the Italicum reform sticks and MS5 wins the election, they would still struggle to achieve that. MS5 would then have 340 seats. Given current polling, the other Eurosceptic party Lega Nord would have around 55 seats. So it would need to increase its share of the vote significantly (from the current 12.5% to around 16.5%) to push the combined MS5-Lega Nord to the two-thirds majority necessary. If it does not get that majority, it would need to a referendum to hold the advisory referendum.
If it MS5 were to manage to hold a consultative referendum, then the public would need to vote to leave the euro. Most polls suggest a majority of the Italian public favour staying in, though support has diminished and the gap is now small. If the public did vote to leave, the government could then use that referendum as a mandate to start the process of a euro exit.
What is the most likely scenario?
There are a lot of possible scenarios, given the uncertainty about polls and elections, as well as the feedback loops between electoral reforms and the implications for government formation and policymaking. We have set out a scenario tree in Figure 3. Based on opinion polls, and the anti-establishment trends more generally, we judge that a No outcome (55%) is more likely than Yes (45%). The Italicum law would also most likely be watered down in such a scenario, meaning that future government formation as well as policymaking would be difficult. A Yes outcome would make less likely that the Italicum is watered down, given that Mr Renzi is a stronger supporter of it. However, other reforms before the next election (early 2018) would probably not be likely given that the current situation in terms of parliament would remain in place until then and Mr Renzi would not likely take measures that could hit his popularity in the run up to the poll.
A strong government beyond the next election would be very likely following a Yes vote. Given current opinion polls a strong PD government would be the more likely and it would then have the possibility to be a reformist government that tackled Italy’s economic problems. However, that is not a given, as that also would depend on the willingness of that government to follow an aggressive reform path, which could still lead to strong opposition from trade unions and other vested interests. Alternatively, an MS5 government could win the elections and take power. However, it would be a small probability within that scenario that they could enough support in parliament to hold a referendum on the euro, given the higher hurdle to pass constitutional laws.
How severe are Italy’s economic problems?
Italy badly needs an economic reform programme to boost its potential growth rate. Its potential growth is generally estimated at close to zero due to ageing and weak productivity growth (see Figure 4). Surveys of international competitiveness suggest it is structurally one of the weakest economy’s in the eurozone, with only Portugal and Greece ranked more poorly. The low potential growth rate exacerbates the country’s two other economic problems: its mountains of government debt and non-performing loans.
Assuming that nominal growth averages 2.5% in coming years, that the primary surplus gradually rises from 1.5% now to 2.5% and interest payments roughly trend at current levels. That leaves the debt ratio trending down only slightly to around 130% GDP in 2025 from 133% now. Furthermore, Italy’s debt sustainability could come into question in the case of even relatively moderate shocks. For instance, assuming 1% slower nominal growth and 1% higher interest rate, would see the debt ratio rising to 160% GDP over that horizon. Arguably the nominal GDP growth we assume is rather ‘generous’ given current potential growth estimates and trends in inflation.
At the same time, Italy’s banking sector needs more capital given the high level of non-provisioned loans. We estimate that if the banking sector sells its NPLs at 25-35%, given the current provisioning, this would imply a capital short-fall of EUR 88-124bn. This amounts to 5.5-7.8% GDP. This would significantly increase the government debt ratio if there was a direct recapitalisation following a bail-in. Up until now, the government has been trying to find private sector solutions to recapitalise its banks, but there are serious question marks about investor appetite.
What are the market implications?
The immediate reaction of Italian government bonds will be to sell-off in a No and rally on a Yes but in both cases spreads will remain elevated. The 10y spread over Italy could move towards 200bp in the first instance in the case of a No, and back towards 140bp in case of a Yes vote. However, these early moves would probably to some extent unwind (especially in case of the initial reaction following a Yes). It would likely become clear that a No vote would not immediately open the door for a MS5 government, while a Yes vote would not lead to much reform in the next year, while it could eventually put MS5 in the driving seat after the next elections. Crucially, we think the key issue is Italy’s economic vulnerabilities, which will remain in place over the next year whatever the result of the poll.
The ECB’s ongoing QE policy should limit the upside for Italy’s government bond yields. Given current sovereign credit ratings, Italy has a quite a buffer before all four agencies place its debt in the sub-investment grade category that would make its bonds ineligible for ECB asset purchases. The ECB bases itself on the highest rating, which currently is given by Fitch, which is three notches away from sub-investment grade (though with a negative outlook).
Reuters reported that the ECB is ready to temporarily step up purchases of Italian government bonds if the outcome of the referendum on Sunday leads to a surge in the country’s bond yields. It cites four unidentified ECB officials who also noted that the move would not necessarily need Governing Council approval. The ECB already deviates from the capital key to make substitute purchases to make up for not being able to make targets for countries where holdings have reached the issuer limit or for other technical reasons. However, this would be a relatively modest and temporary phenomenon because it cannot sharply and persistently deviate from the capital key under the current rules of the programme. Indeed, the Reuters report quotes the officials saying such a policy would be limited to ‘days or weeks to counter any immediate volatility’. If Italy needed long-term support, it would need to officially ask for help according to the report. This would presumably be via the OMT, though that would require Italy entering a reform programme, which would be politically very challenging.
However, an economic shock that leads to a sharp deterioration in the outlook for growth and government debt could increase concerns that there would be significant downgrades in the future. Alternatively, in the scenario where MS5 did get into government following parliamentary reform and did manage to hold a euro referendum, this would also obviously be a major game changer that would see Italy’s spread over Germany explode. An additional element, is that if investors do become worried, a surge in yields would also lead to a sharp deterioration in the government debt outlook, so could become a self-fulfilling prophesy.
On the positive side, following the next elections, a reformist PD government in a reformed parliament, could lead to a rise in growth expectations, leading to a virtuous circle of lower spreads and an improving debt outlook.