Italy's mega solar refinancing boom
Solar PV portfolios in Italy are getting bigger and require increasingly larger debt financings. Portfolio refinancings and acquisitions bring unique sets of challenges for sponsors and lenders but solutions have been found. By Carloandrea Meacci, partner and head of energy Italy, Ashurst.
In the last few months Italy has seen a flurry of mega solar refinancings worth in aggregate approximately €3bn, two of which, Kosmos and Galileo, were individually the largest solar financings ever entered into in Europe. The notable transactions are detailed in Table 1.
The purpose of this article is to highlight the key features of a market trend towards larger and larger solar refinancings in Italy, and potentially also in Europe, without any specific reference to any specific transaction.
The Italian solar market
In the last few years, the Italian secondary market for solar deals has been one of the most active in Europe both on the M&A and the financing fronts, and the two are strictly connected as competition for solar assets has increased, which has induced bidders for M&A auctions to optimise the refinancing terms so as to bid more aggressively and win the auction.
There are many reasons why the Italian M&A market for secondary solar deals has been so active:
* Significant potential for aggregation – total solar capacity is approximately 18,000MW but the largest player owns only 800MW and the second largest less than 300MW;
* Room for optimisation. The larger the portfolio, the larger the economies of scale, particularly in terms of reduced O&M costs per MW and reduced refinancing costs;
* Decent risk-adjusted internal rate of return (IRR), also in light of a more stable regulatory framework (for which please see “Regulatory Risk” below);
* Strong appetite by banks to finance such assets;
* Availability of skilled human capital in Italy in the solar business as Italy was among the first countries to develop this industry on a large scale since 2010;
* A recent increase in the merchant power price, which is relevant because solar plants generally benefit not only from a feed-in tariff paid by a government entity called GSE, approximately 75% of the revenues, but also from the sale of electricity to the market, approximately 25% of the revenues;
* Global trend of solar being an opportunity not only for infrastructure funds and utilities but also oil majors and other energy companies, eg ENI, Total, Shell, BP, A2A, China General Nuclear etc;
* Availability of W&I1 insurance policies for solar M&A transactions, which has made the exit by funds easier, thereby increasing the liquidity of the M&A market.
* A natural consequence of such an M&A frenzy is that portfolios become larger and larger, and the larger the portfolios the larger the refinancings.
Key features of mega solar refinancings
The key features of mega solar refinancings include the following:
i) Structure– The usual structure of an Italian solar refinancing is set out in Figure 1.
ii) Facilities and purpose– The facilities typically include:
* A term facility to refinance the special purpose vehicles’ (SPVs’) existing debt and hedging unwinding costs (which are usually significant – possibly up to 10% of the debt to be refinanced – if the swaps were entered into in 2010–2012 when the IRS curve was very different from today). Because the debt to be refinanced is at the SPV level, and not at the Borrower/HoldCo level, as shown in Table 1 the term facility is downstreamed by the Borrower/HoldCo to the SPVs in the form of intercompany loans and/or equity. The term facility is usually amortising with six-month repayment instalments.
* A debt service reserve (DSR) facility to replace the more classical debt service reserve account (DSRA), which is inefficient in terms of sponsor’s IRR because it freezes cash, with a facility that can be drawn if the project does not generate sufficient cash to service the debt. The DSR facility is normally repaid through the waterfall except that in the last period prior to its maturity, usually the last two to three years, a cash collateral is factored into the waterfall so that, if the DSR facility is drawn in that period, its repayment is cash-collateralised.
A difficult negotiation point regarding the DSR facility may be which events of default act as drawstop of such facility. The classical lenders argument is that all events of default should be capable of stopping the drawing of the DSR facility like for all other facilities while the classical borrower argument is that this frustrates the very purpose of the DSR facility, which is to avoid a non-payment that is likely to have been caused by an event of default and therefore if any event of default was capable of stopping a drawdown of the DSR facility then the DSR facility would never be used.
There may be different solutions to this dilemma ranging from the application of all events of default and potential events of default to only the most material actual events of default to the exclusion of some or all potential events of default.
* Additional facilities may sometimes be added to the structure in connection with the potential acquisition of further plants. These facilities may be of two types:
a) Facilities to refinance the debt of new SPVs to be acquired, shown as add-on facilities in Table 1, in which case the structure is very similar to the term facility and the debt is downstreamed to the relevant SPV;
b) Facilities to acquire new SPVs, ie to pay the relevant purchase price, shown as acquisition facilities in Table 1, in which case the debt is used by the Borrower/HoldCo to pay such price and therefore is not downstreamed to the SPV. In this case the structure is very different from the term facility and cannot be secured by the assets of the SPV to be acquired for financial assistance reasons.
Reasons why these facilities are particularly important for sponsors include:
a) Solar portfolios are usually not static due to the continuing evolution of the secondary solar market described above;
b) Adding a facility of this type to the structure avoids having to put in place a new refinancing, or a new tranche of the existing financing, every time a new acquisition is completed;
c) Even if a new tranche of the existing financing was put in place at the time of a new acquisition, it would arguably require a deed of confirmation/extension of each security document entered into at the time of the existing financing so that the security is extended to secure also the new tranche2;
d) Such facilities give the sponsor certainty of funds, or at least certainty of the pricing thereof, which may be a material factor in the sponsor’s M&A strategy particularly when auctions are involved.
However, there are also some notable difficulties raised by facilities of this type such as that:
a) If the projects to be acquired are not known to the lenders at the time of the entry into of the add-on facility, as is usually the case, then the actual drawing of the add-on facility remains subject to satisfactory due diligence and, if not satisfactory in full, this may trigger renegotiations of the terms of the add-on facility, thereby frustrating the very purpose of such facility;
b) Such facilities obviously trigger arrangement fees and commitment fees, which may be a waste of money if the acquisitions contemplated/hoped for at the time of the entry into of such add-on facility are actually not completed.
c) Another facility that may sometimes be structured in parallel is the so-called equity bridge facility. The purpose of this facility is to bridge the equity commitment of the sponsor and is borrowed by the shareholder of the HoldCo to inject in the HoldCo a portion of the equity required by the term facility lenders. By bridging such equity commitment the overall project leverage increases and consequentially the sponsor IRR.
This facility is usually documented separately from the other facilities described above because the borrower is different, it is the shareholder of the HoldCo, the lenders tend to be different – they are usually a smaller group of the lenders granting the other facilities – and the contractual covenants are lighter because the equity bridge facility is usually perceived as a corporate loan. Therefore the tenor tends to be a few years, the pricing tends to be higher than for the term facility and the security package may be limited to a pledge over that borrower’s shares, an account pledge and an assignment by way of security of the sponsor’s equity commitments.
iii) HoldCo financings – The borrower of the facilities described above, other than the equity bridge facility, is usually the holding company of a portfolio of SPVs each of which usually owns a solar plant. HoldCo structures, as opposed to multi-borrower structures, have the significant advantage of simplicity – particularly for large portfolios, as is usually the case in the solar business - and fit particularly well with situations where the HoldCo acts also as the pooler under a cash pooling mechanism.
However, the disadvantage with HoldCo structures is that the debt, which is at the HoldCo level, is separated from the cash generation, which is at the SPVs level, and therefore the security package granted by the SPVs is upstream. This triggers the applicability of limitations deriving from corporate benefit and financial assistance rules under Italian law. The corporate benefit limitations may sometimes be mitigated by capping the maximum amount secured by a SPV to a percentage of the amount of the term facility on-lent to that SPV.
The financial assistance limitations are trickier to mitigate. They may be triggered because the SPVs were acquired with debt being refinanced or more frequently because the downstream of the term facility to the SPVs described above is made in the form of equity as opposed to intercompany loans.
Banks usually prefer that the downstream is made in the form of intercompany loans because (i) for financial assistance reasons, the security package granted by a SPV cannot secure the term facility to the extent that it was downstreamed to that SPV in the form of equity and (ii) it is legally more difficult for SPVs to upstream cash to the HoldCo, so that the HoldCo may repay the term facility, because SPVs do not have an obligation to repay equity, while they have an obligation to repay intercompany loans.
However, sponsors may prefer the downstream in the form of equity because of certain tax benefits, the so-called ACE, which was introduced to address the historical under-capitalisation of Italian companies3. This usually starts a negotiation between banks and sponsors regarding the optimal split between the downstream in the form of equity v. intercompany loan, which may end up for example as 100:0 or 50:50 or 60:40.
iv) Debt sizing – Debt is usually sized on the basis of certain criteria, which may include: debt service coverage ratio (DSCR), which may be minimum, average, historical and/or forecast; loan life coverage ratio (LLCR); debt to equity; electricity price assumptions; probability of exceedance and; tail4.
Debt sizing is particularly important for add-on facilities when the details of the SPVs to be acquired and refinanced are not known at the entry into of the add-on facility. In such cases the financial model is re-run on the basis of such debt sizing criteria as a condition precedent to first drawdown of the add-on facility and only the amount resulting from that re-run is actually disbursed, provided that it does not exceed the committed amount of the add-on facility.
v) Tenor – Tenor is usually an output resulting from the debt-sizing criteria and in particular the tail. Considering that under Italian law the feed-in tariff lasts 20 years from connection to the grid and that the tariff was usually granted between 2010 and 2012, and assuming that the tail is six months, the maturity of refinancings entered into in 2018 may range between 12 and 13 years.
To the extent that the acquisition facility is structured as pari passu with the term facility, eg same tenor, this is a good outcome for the sponsor because usually acquisition facilities may have more conservative terms also because they are secured by a more limited security package than the Term Facility due to the financial assistance limitations described above.
vi) Pricing– Pricing is specific to each single project and confidential so cannot be disclosed or generalised. According to publicly available information – see for example Project Finance International, May 2 2018, page 39 – pricing for one of such deals was 150 basis points over Euribor.
We cannot confirm if this is true or not but generally speaking the market sentiment is that pricing for such deals was competitive. With the current turmoil in financial markets and the increase of the yields of Italian government bonds with consequential increased cost of funding for certain banks, the same level of pricing competitiveness may no longer be achievable at least in the short term.
vii) No recourse – Such facilities are generally non-recourse so the term facilities are basically HoldCo project financings.
viii) Portfolios– These deals are naturally portfolio financings for the existence of multiple SPVs, as noted above. The portfolio approach sometimes has implications not only on the structure but also on the content of the contractual provisions.
Indeed, the size of the portfolio is sometimes perceived per se as a strength/asset and as a result the borrower of large portfolios is perceived as a quasi-utility, despite being a mere holding of SPVs, and the credit risk as quasi-corporate, despite being a project financing facility. This may mean that the documentation may be covenant-lite or at least lighter than classical project finance, for example:
* Many solar plants in Italy may raise due diligence issues that sometimes may be material. We have experienced, however, that issues, however material, affecting a single plant or a number of plants may become immaterial or less material in the context of a portfolio unless they have significant reputational implications. This means that such issues do no need to be fixed as conditions to drawing or do not necessarily require the negotiation of an ad hoc event of default, representation or covenant, which would otherwise most likely be required if the financing was on a stand-alone basis for that plant only;
* Similarly, the negotiation of the events of default in the finance documentation is affected by the size of the portfolio because the classical sponsor’s argument is that certain events of default affecting a single SPV/plant or a series of SPVs/plants cannot drag the whole portfolio into default.
It is always debated where the line is between events of default capable of defaulting the whole portfolio and events of default not capable of such consequences. There may be cases where the line is represented by a threshold of MW affected by the event and/or the extent to which the event breaches certain financial ratios;
* More generally, the covenant-lite portfolio approach may be reflected in fewer and lighter information covenants, representations, events of default and conditions precedent.
There is actually a more general trend towards switching conditions precedent to conditions subsequent, which in turn triggers a difficult negotiation regarding whether the relevant waivers should be subject to unanimous lenders consent – because they were initially supposed to be conditions precedent, which are usually subject to unanimity – or majority lenders consent because they are like events of default, which are usually subject to majority.
Sometimes a compromise is that conditions subsequent are subject to majority lenders consent except for the conditions subsequent related to security which are subject to unanimity or a super-majority.
ix) Due diligence – As a result of the background of the portfolios described above, which were usually built thorough acquisitions rather than developed internally by the sponsor, when a financing of this type is structured usually there are already available quite recent due diligence reports prepared by the sponsor’s M&A counsel in the context of the acquisitions.
It has therefore become market practice that lenders rely on such reports and do not require their counsel to carry out new due diligence on the portfolio. The role of lenders counsel is usually limited to the review of such due diligence reports (and not the underlying documents) and sometimes may extend to the preparation of a risk matrix, which may sometimes be limited to the most material aspects such as permits and tariff/GSE inspections.
x) Security package – The security package for this type of transaction is the standard security package for a solar project financing and includes:
* At SPV level: a mortgage, an assignment by way of security of the feed-in tariff5, an accounts pledge, a special privilege over movable assets, an assignment by way of security of contractual receivables – insurances, hedging receivables etc – and a pledge over the SPV shares;
* At HoldCo level: a pledge over the HoldCo shares, an accounts pledge, an assignment by way of security of intercompany loan receivables and, if applicable, the share purchase agreement and related W&I insurance policy and hedging receivables. If there are intermediate HoldCos between the Borrower and the SPVs, the intermediate HoldCos usually grant similar security, in particular share pledges and accounts pledges.
Key features of the security package include the following:
* Under Italian law, HoldCo security is full – ie secures all HoldCo’s debt (term facility, DSR facility, add-on facility etc – without any material limitation, while the SPVs security, being upstream, usually secures the HoldCo debt only up to a percentage of the HoldCo debt on-lent to that SPV.
If for example the downstream of the term facility to a SPV is 50% in the form of intercompany loan, and 50% in the form of equity, and the maximum amount secured by that SPV is 200% of the intercompany loan granted to that SPV – ie 200% of the portion of the term facility downstreamed to that SPV in the form of intercompany loan – this means that such SPV secures 100% of the term facility downstreamed to that SPV, which in turn means that, in aggregate, all SPVs secure 100% of the term facility;
* When an acquisition facility is structured, the SPV acquired with that facility cannot secure that facility for financial assistance reasons;
* Specific additional caveats/limitations may apply with reference to Italian law requirements applicable to the assignment of the feed-in tariff – which must rigidly follow the template provided by GSE – and the special privilege over movable assets, which according to certain interpretations cannot secure the debt of an entity other than the grantor of that security and therefore according to them the special privilege granted by a SPV cannot secure at all the HoldCo debt;
* If signing and closing are not simultaneous then the SPV security package can only be taken at closing, and not at signing, when the existing security package is released by the existing banks, otherwise it would be second ranking.
This means, like for any refinancing, that security is entered into at closing but perfected/enforceable vis-à-vis third parties only at a later stage, in which case perfection becomes a condition subsequent. Even more aggressively, particularly when financial close and M&A close are simultaneous and therefore even the mere entry into of the security package may be difficult to achieve on the closing date, market practice has sometimes evolved towards allowing the borrower to enter into (and not only perfect) part of the SPV security package after closing.
xi) Certain funds – Certain funds is a typical acquisition finance concept whereby the only events of default that are capable of stopping the first advance of an acquisition facility are only the most material events of default, eg insolvency etc. This concept is unusual in the project finance world because there is normally no such fundamental need to reach financial close on a certain date, unlike in a M&A context where failure to pay the purchase price on the scheduled date may trigger the payment of material penalties.
However, there may be cases where the project finance closing coincides with the M&A closing in order to avoid triggering the change of control provisions under the existing financings – which may be a critical factor of the M&A strategy of a bidder to avoid making the proposed share purchase agreement subject to receiving the banks’ consent to the change of control. In such cases, the parties sometimes agree that also the closing of the project financing, ie the first advance of the term facility, is subject to a certain funds mechanism.
xii) Distributions– Distribution conditions in this type of financing are subject to the ordinary regime that one would expect to see in any solar project financing such as compliance with ratios, no events of default etc. However, one feature of the mega solar refinancings may be that they may actually create a significant cash windfall to be distributed, for example because the DSRA cash collateral is replaced by the DSR facility and/or the leverage is increased – eg as a result of a shorter tail than the existing financing – and/or the new financing wipes away a lock-up situation, if the previous distribution conditions were tighter.
In such cases, a special and potentially large distribution may become available at, or shortly after, closing and it is always debated the extent to which it is subject to the same or lighter distribution conditions than ordinary distributions. Generally, the closer it occurs to financial close, the more flexibility may be consented on a case-by-case basis.
xiii) Regulatory risk – The first question from any debt or equity investor in any Italian regulated asset like solar is always the extent of the regulatory risk. These mega solar refinancings make no difference in this respect. This is partially as a result of the more general country risk and partially as a result of the specific retroactive cut to solar subsidies enacted in 2014, which was upheld by the Constitutional Court in 2016.
Generally speaking, we understand that the sentiment among the investor community seeking Italian solar assets is, or at least has been for a while, that such risk is reasonable or at least proportional to the premium that a solar deal in Italy commands over a solar deal in other countries.
Such perception of proportionality is obviously very subjective to each investor and may vary greatly from time to time. For example, there have been periods until recently when the risk perception was becoming lower and lower such that some investors were querying whether a bubble was being formed and risks not appropriately priced.
And there have been periods including currently when, together with the surge of the Italian government bond yields, investors have queried whether such premium actually rewards the risk sufficiently. However, even such concerns for certain investors present an opportunity for other investors in terms of reduced competition for assets.
In addition, within the broader infrastructure asset class, renewables have been less affected by certain recent waves of concern because there seems to be continuing and actually increasing government support for renewables, as evidenced by the fact that for the first time in years a new subsidy regime for renewables will be launched shortly.
In addition, the above investor sentiment towards the Italian solar business relies on the fact that the support for renewables in Italy is grounded not only on the classical argument that it is “good to be green” but also on structural and stronger factors than in many other countries such as:
* Renewables are an essential instrument to increase Italy’s energy autonomy. Indeed, Italy does not have the natural resources – oil, gas, coal – to produce electricity that other countries have, and a referendum banned nuclear a few years ago, so Italy must import fuel to produce electricity. Renewables present an opportunity to produce electricity autonomously, without importing fuel from any other country;
* Italian subsidies to renewables are not paid by the government but by the consumers through electricity bills so that the government has no reason to cut subsidies as a tool to improve public finances;
* The rationale for the 2014 cut to the solar subsidies was indeed to reduce electricity bills but with the oil price at current levels – and no expectations of material increases in the near future – there appears to be no impelling reason for the government to make further cuts;
* The aggregate curve of incentives paid to renewable energy plants has peaked and is declining in any event due to the natural expiry of old incentives;
* Paradoxically, the Constitutional Court decision confirming the tariff cuts has brought a degree of stability to the market because many market players actually feared that a declaration of unconstitutionality of the cuts by the Constitutional Court could have prompted the government to take other indirect measures to react to the reinstatement of the old (higher) tariffs;
* The government was careful, at the time of sizing the cuts, to ensure that the projects would not default under their financings;
* The Government is actually finalising a plan to introduce new subsidies for greenfield solar projects under a regime whereby the maximum amount of available subsidies will be capped and therefore with full control over the total spend, differently from the previous regime.
What’s next – grid parity
Although there is still a large potential for aggregation of brownfield subsidised solar assets in Italy and the upcoming new solar subsidy regime will help the development of subsidised greenfield projects, the consensus among solar investors seems to be that the next frontier of the solar business in Italy and elsewhere is grid parity, ie developing unsubsidised solar plants.
This is easier in Italy than in many other countries because of the higher level of irradiation and is becoming possible, or at least less difficult, because of the continuing decrease of the cost of the technology.
This will require the development of a long-term PPA market, which is starting in Italy - like in, and following, Spain - with offtakers starting to consider innovative structures such as availability fees secured by second ranking security packages and/or floored prices with openings for upsides, lenders starting to consider financings with different layers of PPAs – short-term, mid-term, long-term – for the same project, equity investors starting to consider to bridge development costs until the PPA market will have developed to a level where prices are more competitive and mezzanine investors starting to consider how to fill the above gaps.
So … see you hopefully at next grid parity mega solar refi!
1 – W&I insurance policies are policies designed to indemnify a purchaser if certain representations and warranties of the seller under an acquisition agreement are not true.
2 – This would be burdensome and expensive, eg notarial fees and legal fees, because the existing security package is usually very extensive. In some cases this may not even be possible, eg for the feed-in tariff assignment an extension of the security is not contemplated by the rigid forms provided by GSE and therefore the security must be released and re-taken in its entirety, with the risk that the insolvency claw-back period re-starts and is longer – 12 versus six months – because the security is no longer entered into simultaneously with the original debt.
3 – According to publicly available information, it is however expected that the 2019 budget law will abolish ACE in its entirety so, if this will be confirmed, ACE may be a less relevant factor in the future.
4 – The tail is the time lag between the final maturity of the financing and the expiry of the tariffs and therefore represents a buffer if issues arise during the loan’s life requiring an extension of the repayment profile.
5 – This is sometimes waived by the lenders to avoid the risk deriving from certain case law precedents whereby GSE may claim the return of the feed-in tariff (eg if it has been revoked) from the banks if the tariff has been assigned to the banks by way of security.