EnergySource | Issue 20 19 Jul 2018 Italian renewables: A bond evolution

All the above bonds have been listed on ExtraMOT Pro, the professional segment of the multilateral trading facility ExtraMOT, regulated by the Italian Stock Exchange and dedicated to the listing of bonds and commercial paper. Such a bond listing is simpler, quicker and cheaper than listing the equity of a company, also in terms of initial and ongoing disclosure requirements.

Following the introduction of the new rules at the height of the credit crisis, the market showed significant interest in this product in late 2014 and early 2015. This was followed by a pause in late 2015 and early 2016 when banks became more active, but activity resumed in the second half of 2016 with a new wave of issuances.

We understand that the above pause was generated by, among other things, the ECB’s quantitative easing that flooded the banking market with liquidity, making it more difficult for bondholders to compete with banks – for example, in terms of pricing and tenor.

However, quantitative easing sooner or later will expire while Basel III rules on capital requirements for banks will not (and may actually become even tougher under Basel IV). In the medium to long-term we expect this product to be more frequently used because it will be a more efficient use of capital for qualified investors and infrastructure debt funds than for banks to provide long-term financing for infrastructure projects.

Structures

All the renewables bonds issued to-date have been amortising, some of them have floating interest rates – albeit partially hedged under separate hedging agreements – while others have fixed interest rates. All such bond deals have been refinancings of portfolios of brownfield solar PV assets grouped under a holdco that issued the bond. They all benefited from a cash pooling mechanism where the cash pooler was the holdco and the ratios were calculated at the holdco/ aggregate level. The debt has usually been sized on the basis of a tail calculated on the basis of the weighted average expiry date of the incentives.

In terms of tranching of the debt, we have seen both pure bonds with no bank tranches and more commonly hybrid structures: that is, a combination of pari passu bond tranches and bank tranches regulated among other things by a common terms agreement.

The hybrid structure is generally used when there are liquidity concerns – as to whether there is sufficient appetite for the asset in the bond market – and when certain bondholders desire that banks share the risk. The need for ancillary facilities traditionally provided only by banks such as a VAT facility, not required when completion is not recent, or a liquidity facility, in lieu of a debt service reserve account, may also play a role in the selection of the hybrid structure.

In terms of rating, most of the renewables bonds issued so far have been unrated because sophisticated bondholders usually rely solely on their internal rating and because rating agencies may require a higher debt service cover ratio (DSCR) than banks and bondholders in order to grant investment grade ratings and such higher DSCRs may not be compatible in terms of debt sizing with the hedging unwinding costs (up to 10 per cent) usually associated with these refinancing deals. However, the absence of a rating sometimes makes it impossible for certain types of bondholders to subscribe to the bond.

In terms of corporate structure, all renewables bonds have been issued at the holdco level for a number of reasons. First, renewables plants are usually small and held by different SPVs grouped in a portfolio under a holdco and it would be practically impossible to issue as many bonds as the SPVs. Second, the tax regime associated with downstreaming of the bond proceeds from holdco to the SPVs – to allow the SPVs to refinance their debt – in the form of equity or quasi-equity (versamenti in conto aumento di capitale) is eligible for a notional interest deduction (the ACE tax relief). However, this equity/quasi-equity downstream triggers increased rigidity and difficulty in upstreaming the cash from the SPVs to holdco to service the holdco debt, i.e. the bond. The typical structure is depicted in figure 2.

Figure 2 – Corporate bond structure 

Players

In terms of types of bondholders, the subscribers have usually been infrastructure debt funds whose investors have frequently included insurance companies. Insurance companies have sometimes subscribed to the notes directly. In each case the investment rationale was the appetite for infrastructure assets generating stable returns over the long term to allow take-and-hold investors that have no stated intent to sell down the bond in the short/medium-term.

Interestingly, although all such bonds have been listed on a regulated market or multilateral trading facility, the rationale for the listing was not linked to the liquidity or marketability of the bond – which is not a key requirement for a take-and-hold investor – but was rather driven by a legal analysis.

Indeed, bonds listed on a regulated market are not subject to the risk of an unfavourable interpretation of certain legal restrictions, such as the need for the subscribers to be subject to prudential supervision and the liability, in case of insolvency of the issuer, of the initial subscriber vis-à-vis the ultimate bondholder, which is not a professional investor.

Security package

In terms of security package, these bonds benefit from the same security package usually applicable in a bank project financing – share pledge, mortgage, special privilege, accounts pledge, assignment of receivables by way of security including the feed-in tariff – except that the assignment of the feed-in tariff in favour of bondholders is not expressly contemplated by the relevant GSE¹ rules. However, we are not aware of any objection so far by GSE to such assignments and we do not see any rational reason for such objection.

Renewables sources

In terms of technology, all the initial renewables project bonds have been in the solar sector, but most recently, at the end of 2017, there was a bond issue in the wind farm sector (see figure 3).

Figure 3 – The Glennmont wind farm portfolio bond issue

Ashurst advised BNP Paribas Securities Services as noteholders' representative, security agent, calculation agent and paying agent in connection with the project bond granted to CEF 3 Wind Energy, a company controlled by Glennmont Partners and PGGM Vermogensbeheer, to refinance part of CEF 3 Wind Energy's wind farm portfolio, which is one of Italy’s largest operating wind portfolios with 245 MW of installed capacity. 

This is the first project bond in the wind sector in Italy. The finance structure includes a bank loan and a project bond, both senior secured, having an aggregate value of €190 million. The €170 million unrated bond consists of a single fixed rate tranche and is listed on the Italian ExtraMotpro segment with national and international institutional investors. 

A bond was also issued in late 2016 in the water sector under the project bond regime rather than the mini-bond regime because it was not a hydroelectric power plant but rather a water management system, which was interpreted by the relevant counsel as falling within the key infrastructure sectors mentioned above.²

The key reasons why all renewables project bonds so far have been in the solar sector despite recent feed-in tariff cuts by the Italian Government include the following:

• there are many more solar plants than wind farms – approximately 19,000 MW versus 9,000 MW – and solar plants tend to be smaller;
• the feed-in tariff cuts pushed sponsors to refinance their projects in order to partially offset the reduction in tariff revenues with a reduction in the cost of debt service and avoid dividend lock-ups; and
• the M&A market for solar plants has been more active than for wind farms, which in turn triggered the refinancing push.

Reasons for bond versus loan

Why might sponsors and banks prefer a bond versus a loan in connection with the refinancing of a renewables portfolio? There have been times when the main reason was liquidity or price, but this is no longer the case in the current market. Additional reasons nowadays include the following:

• a bond may be more covenant-lite than a typical project finance loan – for example, in terms of ongoing reporting requirements and in terms of change of control, as sometimes bondholders accept that the controlling sponsor simply needs to fall within certain categories of investor as opposed to maintaining the original sponsor;
• the fact that renewables bonds are listed requires a minimum level of disclosure – a fact that might be welcomed by future potential equity investors that are used to more transparency, and such transparency might be helpful also in case of an IPO of the solar portfolio; and
• there is a general perception that in the long term – particularly after the end of the ECB's quantitative easing – the future of infrastructure financing is in the capital markets, so selecting today a bond structure as opposed to a loan structure is a good way to prepare the team/human capital for the technology of the future.

On the other hand, bonds, compared to conventional loans, have a number of disadvantages, such  as prepayment fees that can be phased but very rarely waived, and the potential complexity of the waiver process. However, such complexity is in practice often mitigated by the following factors:

• the agent in a bond deal may have more discretion to act independently than in a loan deal;
• the number of bondholders for this type of project is usually quite limited (a handful in many cases, as opposed to tens or hundreds in typical corporate bonds); and
• bondholders have the "take-and-hold" approach mentioned above, so the relationship with bondholders is ultimately not so dissimilar to that with banks.

Regulatory risk

The main question for any credit committee including bondholders investing in Italian renewables is the regulatory risk, particularly after the retroactive solar feed-in tariff cuts introduced in 2015 and the Constitutional Court decision in December 2016 confirming such cuts. This point is not specific to bonds versus loans, nor to financing versus M&A, but it is still worth mentioning due to its materiality. There are various reasons why investors have achieved a degree of comfort in this respect:

• unlike in other countries perceived as similar (e.g. Spain) Italian subsidies to renewables are not funded by the Government through general taxation but passed down to energy consumers through electricity bills, so the Government is not incentivised to cut subsidies to improve public finances;
• the rationale for the cut was therefore to reduce electricity bills, to improve the competitiveness of doing business in Italy. However, with the oil price at lower levels, there appears to be no compelling reason for the Government to make further cuts;
• the aggregate curve of incentives paid to renewable energy producers has started to decline in any event due to the natural expiry of old incentives and the phasing out of new incentives for new plants;
• paradoxically, the Constitutional Court decision confirming the tariff cuts has brought a degree of stability to the market because many market players feared that a declaration of unconstitutionality of the cuts by the court could have prompted the Government to take other indirect measures in response to the reinstatement of the old higher tariffs;
• the Government was careful, at the time of formulating the size of the cuts, to ensure that projects would not default and therefore lenders would not be hit directly by the tariff cuts.   Indeed, most projects we have seen went into lock-up, (i.e. no dividend distributions) but not in default (i.e. a debt service cover ration of less than 1) as a result of the cuts;
• with the tariff cuts the Government has lost credibility vis-à-vis the international investor community at a time when Italy is eager to increase international investments and no doubt the Government is mindful of the fact that further cuts may jeopardise appetite for assets in the country;
• the retroactive cuts applied only to solar plants, which, in the absence of any caps/ceilings on capacity, rose to a staggering total of approximately 19,000 MW; and
• some of the largest Italian companies, such as Eni and Enel, both state-owned, have announced programmes to invest in renewables in Italy.

Despite all the above, Italy is still perceived as more risky than Northern European countries and this risk commands a premium. There is a perception by some market players that this premium is worth the risk, with a view to balancing the overall return of a European portfolio of infrastructure assets.

A final connected trend we have seen more recently is that as more international investors look to Italy for debt or M&A deals in the renewables or other infrastructure sectors, competition for Italian assets tends to increase and the premium mentioned above tends to reduce, leading to more sophisticated investors seeking deals at the edges of the classic infrastructure world – for example, in telecoms, real estate with an infrastructure angle or greenfield renewables projects.

Special thanks for their contribution to Fabio Balza, Head of Tax at Ashurst in Milan and Umberto Antonelli, Counsel at Ashurst in Milan. 

This is an updated version of an article first published in April 2017 in the Project Finance International Global Energy Report.

1. GSE is the government entity that administers the feed-in tariffs for renewable projects in Italy and issues rules relating to their administration, including in relation to the assignment of the tariffs by way of security.
2. This is the €100 million bond issued by Metropolitana Milanese SpA in relation to the Milan water system.