Confirming existing Swedish security in an amendment and restatement – what the Swedish opinion can and cannot say

A common situation

A group has borrowed from a syndicate under a facilities agreement governed by Norwegian, English or another non-Swedish law. The security package includes pledges over the shares in the Swedish subsidiaries, intra-group loans and bank accounts, all governed by Swedish law. Most of it was perfected at closing. Some of it, as is usual with Swedish account pledges, was not, and that is a question in its own right to which this article returns. A Swedish law firm gave a security opinion at the time. A year or two later the parties agree to extend the maturity, reprice the facilities and add an accordion or a new term loan. The amendment is documented in an amendment and restatement agreement (“ARA”), and lead counsel, on the lenders’ behalf, asks each local counsel for a legal opinion.

The Swedish opinion is, in the first instance, an opinion on the ARA. It covers capacity, due authorisation and execution by the Swedish obligors, and the usual conclusions on choice of law, enforcement of judgments and the like. The Swedish companies may also have signed the ARA as security providers. What the lenders then want to see in the opinion is a line on the existing security. The wording proposed by lead counsel is typically borrowed from the other jurisdictions’ opinions and reads as a confirmation that the security documents continue to constitute valid and perfected first ranking security, “as confirmed in the Amendment and Restatement Agreement”. It is not obvious that the wording does what it is meant to do in the jurisdictions it comes from either. It has the look of language passed from one opinion to the next without anyone, local counsel included, stopping to ask what it confirms, and there seems to be room in the market for thinking the question through.

The Swedish adviser’s reluctance to adopt that wording has to do with three features of Swedish law that the proposed wording passes over, and most of them probably have counterparts in other legal systems, not least the other Nordic ones. A Swedish pledge is accessory to the secured claim. It attaches to specific assets through specific perfection acts. The third is that a company’s decision to secure someone else’s debt is tested, under the Swedish Companies Act (Sw. aktiebolagslagen (2005:551)), against the exposure actually assumed at the time it is assumed, and how a later increase of that exposure is to be assessed is not settled. The sections that follow take each in turn. The question that is easiest to overlook, whether the board minutes on which the original pledge rested can carry the increase, is left to the second article.

The debt side and the asset side

Swedish law pledge agreements in syndicated financings routinely secure “all present and future obligations of the Obligors under the Finance Documents, as amended, restated, supplemented, extended, novated or replaced from time to time, including any increase”. The intention is that obligations added by later amendments are covered without a new pledge. Swedish law gives the clause that effect. A general pledge clause (Sw. generalpantklausul) is valid between commercial parties, the secured obligations are sufficiently identified by the reference to the finance documents, and a pledge that has been granted and perfected in advance and by its terms covers future debt is, on the prevailing view, treated as stipulated in advance for that debt when it arises. On the debt side, therefore, the clause does what it says and widens the circle of obligations the pledge secures.

What the clause cannot do is widen the asset side. Swedish security law is asset-specific and perfection-specific. A share pledge covers the shares represented by the share certificates delivered to the pledgee (Chapter 6, Section 8 of the Companies Act). A pledge over a negotiable promissory note covers that note (Section 22 of the Promissory Notes Act (Sw. skuldebrevslagen (1936:81))). A pledge over a receivable or a bank account covers the receivable or account of which the debtor or the account bank was notified (Section 31). Shares issued after the pledge are not covered until new certificates have been issued and delivered, even if the pledge agreement says “all shares from time to time”. A replacement promissory note, a new intra-group loan under a new agreement or a new bank account requires a new perfection act. A mortgage-based security (Sw. företagsinteckning, fastighetsinteckning) is limited in amount to the mortgage certificate plus the statutory addition, so an increase in the debt beyond that amount is unsecured unless new mortgages are registered and pledged, and that is new security, not confirmed security.

This distinction explains why a confirmation that the security “continues to constitute a perfected security interest over the assets pledged thereunder” is a different and larger statement than a confirmation that the amended obligations are secured on the same footing as the original ones. The first is a statement about the assets as they stand today, which the Swedish adviser has not seen since closing. Share certificates may have been returned, accounts may have been closed and replaced, a subsidiary may have issued shares in a reorganisation, a release may have been given under the intercreditor agreement. None of that would show up in the ARA. The second is a statement about the pledge agreement’s terms and their effect under Swedish law, which the adviser can read and opine on.

The original opinion already reflects this. It does not say that the pledge agreement creates security over whatever shares, loans or accounts the pledgor may come to hold. It says that security is created over the shares, the intra-group loans, the accounts and the mortgage certificates as defined in that opinion, which is to say the shares evidenced by the certificates listed, the loans under the agreements named, the accounts with the numbers given and the mortgage certificates in the amounts stated. For each class it then adds that the pledge agreement provides for further assets of the same kind to become subject to the security, and what has to be done for the security over them to be perfected, such as delivery of new share certificates, a new notice to the debtor or the account bank, or registration of a new mortgage certificate in the agent’s name. The opinion is given on assumptions, among them that the certificates of registration and the pledgor’s own statements of what it owns are correct, and subject to qualifications. The ARA opinion refers to that opinion and incorporates its assumptions and qualifications. The asset side of the ARA opinion is therefore not an open question but a defined one. What the adviser can say is that the obligations as amended stand where the original obligations stood, secured by the same documents to the same extent and on the same terms, whatever that extent is under the original opinion. Anything that has come into existence since is covered by the pledge agreement on its terms and by the original opinion’s description of what perfection requires, and whether those steps have in fact been taken is something the adviser has not examined and should not be read as answering. That is why the question of what has happened to the assets since closing belongs to the pledgor and the security agent rather than to the Swedish adviser.

What that question looks like differs by class of asset. For shares in a company that has issued share certificates, it is whether any shares have been issued since the pledge and, if so, whether certificates for them have been delivered to the agent. For shares registered with a central securities depository, it is whether new shares have been registered on the pledged account or on some other account. For intra-group loans under negotiable promissory notes, it is whether the notes are still held by the agent and whether any loan has been increased by a side letter or replaced by a new note. For intra-group loans under ordinary loan agreements and other simple receivables, it is whether the agreements have been replaced or new loans made under new agreements, since the notice given at closing covers only the agreement it names, and whether the pledgor is still cut off from collecting. For bank accounts, it is whether any pledged account has been closed or replaced and whether accounts have been opened that the finance documents require to be pledged. For business and real property mortgages, it is whether the mortgage certificates are still held by or registered to the agent, whether further mortgages have been registered and pledged to someone else, and whether the increased debt exceeds the mortgage amount with the statutory addition, in which case the excess is unsecured until new mortgages are registered and pledged, and that is new security with the full clawback exposure of new security. The list is not for the opinion. It is what the Swedish adviser asks lead counsel about when the opinion is being prepared, and what lead counsel should ask the pledgors. Two things cut across every class. Whatever the asset, the pledgor must have remained cut off from dealing with it, and whether that has held since closing is something the adviser giving the opinion cannot know. The pledge can also have been affected without any written amendment at all, where the parties’ practice since closing has departed from the terms of the pledge, which is a release or a loss of perfection by conduct rather than by document.

Two further points on the asset side are often lost in the wording lead counsel proposes. The first is that the original opinion also records what was actually perfected when the security was taken, which is not always what the security documents purport to create, and the ARA opinion inherits that picture. Swedish bank account pledges are the obvious example. They are not perfected until the pledgor is deprived of control of the account, which in a financing is at a triggering event, and until then the account pledge is a contractual promise of which “perfected first ranking” cannot be said, however the ARA is worded. The same question arises for pledges over intra-group loans. A pledge over a simple receivable is perfected by notice to the debtor and by the pledgor being cut off from collecting it, and where the finance documents let the pledgor go on receiving payment until a triggering event the position is the same as for the accounts. Where the loan is documented in a negotiable promissory note, perfection depends on the note having been delivered to the pledgee and still being held, and a loan that has since been increased by a side letter, or replaced by a new note, is to that extent unperfected. None of this is changed by the ARA, and none of it can be confirmed by it. A confirmation that puts the new debt on the same footing as the old carries that picture with it, unperfected pledges included, rather than improving on it. The second point is that a release under Swedish law requires no form. It can be express, implied, partial or the automatic consequence of a permitted disposal under the intercreditor agreement. The original opinion’s perfection qualifications protect against a pledge that never obtained, or lost, its proprietary effect. They say nothing about a pledge that has simply ceased to exist by agreement.

Amendment or novation

A Swedish pledge is accessory to the claim it secures. If the claim is discharged, the pledge is discharged with it, whatever the pledgee holds in its safe. The question whether an ARA amends the existing obligations or discharges them and replaces them with new ones is therefore the first question for the Swedish pledge, and it is not a Swedish law question, since it is decided by the governing law of the facilities agreement.

In the ordinary case the answer is that the ARA is an amendment. The obligations continue, on amended terms, and the general pledge clause’s reference to the finance documents “as amended” is enough. If instead the ARA operated as a novation under its governing law, the consequences for the Swedish security would depend on the wording of the pledge agreement. If the pledge agreement secures the obligations only “as amended”, the original claim is discharged, the pledge falls with it and the new claim is unsecured. If the pledge agreement also captures “replaced” or “novated” obligations, the pledge survives as a contract and covers the new claim, but now as old security for a claim that is new in its entirety, with the consequences for clawback described in the next section. A third-party pledgor that has not consented to the new claim is not bound by the pledge in respect of it, even if the clause formally covers it, because a novation by definition goes beyond what the pledgor undertook under the original claim.

The Swedish opinion cannot answer that question and should not try. It is a question for lead counsel, and the Swedish opinion should be drafted so that it rests on lead counsel’s answer rather than leaving the lenders with an open point about Swedish law that the Swedish adviser could in fact have settled. It helps to be precise about what the question is. An ARA will often contain things that look like replacement, a prepayment of part of the debt, the resignation of an obligor, the transfer of a loan from one borrower to another, and none of them makes the agreement a novation. What matters is whether the facilities agreement remains the same agreement, with the obligations continuing under it on amended terms. The related question, whether the obligations under the agreement as amended fall within the secured obligations as the pledge agreements define them, has two parts. Reading the secured obligations clause is Swedish law, and it is the reading on which the wider confirmation rests. Whether the amended agreement is a finance document, or whatever else the clause points to, under the foreign-law facilities agreement or intercreditor agreement is not, and that part belongs with those who can answer it. Where the whole structure is Swedish, the intercreditor agreement included, the Swedish adviser can answer that question itself. If lead counsel cannot confirm that the ARA is an amendment, the structure needs new Swedish pledge agreements, and a new security opinion to go with them.

Old security for new debt

Chapter 4, Section 12 of the Swedish Bankruptcy Act (Sw. konkurslagen (1987:672)) provides that security given by the debtor later than three months before the reference date (two years in relation to a related party) is recoverable by the estate unless it was stipulated when the debt was incurred, or was transferred without delay after the debt was incurred. Since 2002 the provision also excepts security whose giving can, in the circumstances, nonetheless be regarded as ordinary, on the model of the exception for ordinary payments in Section 10. The exception is construed narrowly, it is directed at security that is not prompted by payment difficulties or an impending insolvency, and it is a fallback rather than a footing. It is of no help for security that is perfected only on a triggering event, since the event is usually a default. The provision is written for the case where a creditor obtains security for a debt that already exists. An ARA that adds debt to a pledge that already exists is, on its face, the mirror image, and the question is whether the pledge is “stipulated when the debt was incurred” in respect of the addition.

The prevailing view in the literature (Lennander, Walin and Millqvist) is that it is. A pledge granted and perfected in advance, which by its terms covers future debt, is stipulated for the new debt when the new debt arises. The lender extends the new credit in reliance on a security that already exists and is already agreed to cover it, and the transfer of the security has taken place before, not after, the debt was incurred. For a credit facility the security is regarded as stipulated at the time of the credit agreement rather than at each drawdown. A perfected share pledge is therefore not ordinarily recoverable merely because an ARA adds new or increased obligations.

That is also the footing on which the Swedish opinion proceeds. The clawback qualification in the original security opinion carries over by reference, and the ARA opinion does not add a qualification of its own for the increase. Three things are nonetheless worth keeping in view. First, the Supreme Court has not ruled on the question for commercial general pledge clauses, and “stipulated” can be read as requiring that the lender, when extending the new credit, actually made it conditional on the security. In practice the ARA’s references to the security do that, but it is not something the adviser has verified. Second, security that was not perfected when it was granted, such as the account pledges and any receivables pledge under which the pledgor may still collect, is not “transferred” until the triggering event, which comes after both the old and the new debt. For that security Section 12 applies in full, regardless of the general pledge question, as the original opinion will already have said. Third, Chapter 4, Section 5, the general avoidance rule for improper transactions where the debtor was insolvent and the counterparty knew or ought to have known, applies in any event, with a five-year limit and no limit at all in relation to related parties.

Two drafting points follow. A security confirmation signed by the pledgors should be drafted as a confirmation of existing security, never as a re-grant, since a re-grant invites the argument that it is new security for old debt, which is the classic Section 12 case. If the pledge agreement’s secured obligations clause does not in fact reach the new debt, the problem is one of scope rather than clawback. The pledge simply does not secure the new debt, and no confirmation changes that. It is a question of the pledge agreements’ terms, and it is the reason the pledge agreements have to be read before the wider confirmation can be given. If the clause does not carry increases, the increase cannot be said to be secured on the same footing as the original debt, however the confirmation is worded.

The third-party security provider

In most group financings the Swedish pledgor is not the borrower. A Swedish holding company pledges the shares in its subsidiary, or a subsidiary pledges its intra-group receivable, to secure the parent’s debt. Swedish law applies the principles of suretyship (Sw. borgen) by analogy to a pledge for another’s debt. A surety’s liability cannot be extended without its consent, and amendments that materially increase the surety’s risk, such as an increase in the amount, an extension of the term or the addition of new borrowers, do not bind the surety merely because the original undertaking was widely worded. A general pledge clause is a form of advance consent to future amendments, and such consent is valid, but it is construed restrictively where the result is surprising or burdensome, with Section 36 of the Contracts Act (Sw. avtalslagen (1915:218)) in the background. The analogy to a general guarantee (Sw. generalborgen) is apt. It works, but it is not the same as consent to the particular increase.

The simplest solution is for the Swedish security providers to be party to the ARA, or to sign a separate security confirmation. Either gives the pledgor’s express consent to the amended obligations and removes the suretyship argument. It is also the only way the pledgor’s board comes to consider the increase at all, which is where the second article begins. Security providers are made party to the ARA as a matter of course in most financings, but it does happen that the borrower side resists, usually to save board meetings. The lenders should then understand what they are giving up. Without the pledgor’s consent the suretyship argument stays open for the increase, the pledgor’s board has not considered it, and the Swedish opinion cannot say that the increase is secured on the same footing as the original debt. The most it can say is that the ARA does not of itself affect the security, which is a statement about the ARA, not about the increase.

A consent, however, is not a statement of fact. A security confirmation clause in the ARA tells the Swedish adviser that the pledgor agrees that its pledge secures the amended obligations. It does not tell the adviser that the share certificates are still with the security agent, that no shares have been issued or that no release has been given. The opinion does not rest on those facts being verified, and it cannot. It is drafted so that it does not depend on them, and whether anything has been released since closing is a question the lenders and lead counsel have to put to the pledgor and the security agent outside the opinion. A statement from the pledgor that nothing has been released is not current practice, and one almost never sees it. It is rather something the market could consider, since it would say a good deal more than the formalities certificates directors routinely sign. Those began life as personal statements standing beside what the company itself certifies, and have been watered down to the point where they are personal without personal liability. If a statement on releases is given it can be recorded, but the opinion should not depend on it, and it will usually have to be qualified for the releases and amendments that have in fact occurred since closing, which turns a comfort into a list of exceptions. Whether the instruments are still with the security agent, whether an account has been closed or replaced and whether shares have been issued are questions of a different kind. They are not for a certificate. They belong to lead counsel’s due diligence with the pledgors, and the opinion has to be drafted so that it does not answer them by implication.

What the Swedish opinion can say

The Swedish opinion on an ARA does not restate the original security opinion. It refers to it, incorporates its assumptions, limitations and qualifications, and states that nothing in the new opinion is a restatement of the old one or a confirmation of any change, waiver or release that may have occurred since. This is better for everyone than a restatement. The adviser avoids bringing down facts it has not re-examined, and the lenders keep an opinion whose reservations are the ones they accepted at closing. On that footing there are two things the opinion can say about the existing security, and they are different in kind.

The narrower is that the ARA does not, of itself, affect the existing security. That is a statement about cause and effect. It is true even if a release has occurred, since the ARA did not cause it, and its one point of exposure is novation, which is why that question has to go to lead counsel even for the narrower confirmation. The wider is that the obligations as amended, the increase included, are secured by the existing security documents to the same extent and on the same terms as the original obligations were. The narrower is the normal form, and in most amendments it is all the lenders need. The general pledge clause and the original opinion already do the work, and an extension or a repricing adds nothing for the Swedish adviser to say. The wider confirmation is for the amendment that actually increases the pledgors’ exposure, and it is not an upgrade to be had for the asking. It is a different opinion, with conditions of its own. It is a statement on the debt side, and it is deliberately relative. It says nothing about what the security is, whether it is perfected or what it covers. If an account pledge was unperfected for the old debt it is unperfected for the new, and whatever the original opinion was silent about, related rights, interest on intra-group loans, stays outside. Written that way the confirmation needs no proviso about perfection, and it sits comfortably next to an express statement that the opinion says nothing about the status or scope of the security itself. A confirmation that the security “will extend to secure” the new obligations is a different statement. It is about the security interest, it collides with that exclusion, and it can only be rescued by assumptions, which are the weakest protection an opinion offers, because they can be challenged on the ground that the adviser had reason not to make them. An opinion that assumes away what the lenders asked it to confirm has not helped them. Its conditions are these. The pledge agreements, and usually the intercreditor agreement they point to, have to be examined for the secured obligations clause, the pledgors have to be party to the ARA or sign a confirmation, each pledgor’s board has to resolve on the ARA itself, and lead counsel has to confirm that the ARA is an amendment. What it does not need is a new set of qualifications. The reservations that matter, on clawback, on corporate benefit and on security that was never perfected, are already in the original opinion and travel with it by reference. Adding them again for the increase would suggest that they did not apply before, and it would oblige the adviser to go back through the original opinion for every unperfected pledge each time an amendment comes along. None of this is unreasonable, but it has to be said early, because each item is something the lenders or the borrower have to produce.

What the opinion will not say, at either level, is that the existing security “continues to constitute a legal, valid, binding and perfected first ranking security interest over the assets pledged thereunder”. That is the original opinion re-dated. It cannot be true of security that was never perfected, such as the account pledges, and it is an asset-side statement the adviser has no basis to make. Where there is something new to opine on, new pledge agreements, new pledgors, new mortgages because the increased debt exceeds the mortgage amount, or defects in the original security that are to be cured, the right instrument is a new security opinion on the new documents rather than an ARA opinion with additions, with the ARA opinion given alongside it for the security that is unaffected.

Conclusions

The Swedish law on the point is friendlier than the opinion discussion suggests. A properly drafted general pledge clause carries amended and increased obligations without a new pledge, a perfected pledge is not ordinarily recoverable merely because the secured debt has grown, and the ARA, if it is an amendment, does nothing to the security. The difficulties are of a different kind. The lenders are asking for a confirmation about assets the adviser has not seen since closing, and the confirmation is only as good as the pledgor’s own decision to give it.

For lead counsel, the practical consequence is that the wording used in other jurisdictions will not travel to Sweden unchanged, and that the Swedish opinion will instead say that the security is not adversely affected by the ARA and, where the amendment increases the pledgors’ exposure and they have consented to it in the proper form, that the amended obligations are secured to the same extent, and on the same terms, as the original ones. The assumptions and reservations are not a retreat. They allocate the asset-side facts to those who can verify them, the pledgor and the security agent, and leave the Swedish adviser to say only what it can actually say. For the security agent, whether anything has been released or replaced since closing is its own question to ask, and it is better asked before the opinion is drafted than read into it afterwards. For the borrower, the lesson is that an amendment which increases a Swedish pledgor’s exposure needs the Swedish pledgor’s signature and a board resolution on that agreement. Why the resolution has to be on that agreement, and why the blanket authorisation from closing will not do, is the subject of the second article.

For the Swedish adviser, finally, there is reason to say early, and in writing, whether the wider confirmation is on the table at all and, if it is, what it requires, namely pledgors party to the ARA, minutes on the ARA from each pledgor, and lead counsel’s confirmation that the ARA is an amendment. The alternative is a confirmation that extends the security while recording that the board never considered the extension, and that is a document that will be hard to explain later.