Limited recourse

The basic principles of contract

Formation: capacity and authority · representation · misrepresentation · offer · acceptance · consideration · intention to create legal relations · agreement to agree · privity of contract oral vs written contract · principal · agent Interpretation and change: governing law · mistake · implied term · amendment · assignment · novation Performance: force majeure · promise · waiver · warranty · covenant · sovereign immunity · illegality · severability · good faith · commercially reasonable manner · commercial imperative · indemnity · guarantee Breach: breach · repudiation · causation · remoteness of damage · direct loss · consequential loss · foreseeability · damages · contractual negligence · process agent Remedies: damages · adequacy of damages ·equitable remedies · injunction · specific performance · limited recourse · rescission · estoppel · concurrent liability Not contracts: Restitution — quasi-contract — quasi-agency Index: Click ᐅ to expand: Contract Tort Restitution

Tell me more Sign up for our newsletter — or just get in touch: for ½ a weekly 🍺 you get to consult JC. Ask about it here.

Of a contract, that a debtor’s obligations under it are limited to a defined pool of assets. You see this a lot in repackagings, securitisations and other structured transactions involving espievies. Security and limited recourse are fundamental structural aspects of contracts with special purpose vehicles and investment funds.

The basic idea

Investment funds and structured note issuance vehicles tend to be purpose-built single corporations with no other role in life. They issue shares or units, or secured notes, to investors and with the proceeds, buy securities, make investments and enter swaps, loans and other transactions with their counterparties.

These counterparties will generally be structurally senior to the fund’s investors (either as unsecured creditors, where the investors are shareholders, or as higher-ranking secured creditors, where the investors are also secured noteholders).

Why?

The main reason for limiting a swap dealer’s recourse to the espievie’s assets is not to prevent the swap dealer being paid what it is rightly owed. It is to stop an empty SPV going into formal bankruptcy once all its assets have been liquidated and distributed, according to their contractual priority, to investors. At this point, the espievie has nothing left to pay anyone, so launching a bankruptcy petition is kinda — academic, but sometimes academic stuff is important if you’re a director of an SPV.

Now, why would any creditor want to put an empty espievie — one which has already handed over all its worldly goods — into liquidation? What good would it do? Search me. Why, on the other hand, would the directors of that empty espievie, bereft as it is of worldly goods, be anxious for it not to go into liquidation? Because their personal livelihoods depend on it: being directors of a bankrupt company opens them to allegations of reckless trading, which may bar them from acting as directors to other countries. Since that’s their day job, that’d be a bummer.

But if the espievie’s bankrupt, doesn’t that mean they have been reckless? No. Remember, we are in the parallel universe of SPVs. Unlike normal commercial undertakings, espievies run on autopilot. They are designed to give exposure, exactly, to the pools of assets and liabilities they hold. That’s the deal. Everyone trades with that understanding. Those assets might blow up, but that’s hardly the espievie’s fault. How is it supposed to know? It is a harmless little otter-like creature from Guernsey. The directors are really nominal figures, and are also rather like otters: they outsource trading decisions (if any — in a repackaging, there most likely won’t be any) to an investment manager. The directors are really there to ensure accounts are prepared and a return filed each year, and build little dams out of twigs and rushes.^[1] They are not responsible for the trading strategy that drove the espievie into the wall.^[2]

So all an investment fund’s limited recourse clause really needs to say is:

Creditors’ recourse against the fund will be limited to its assets, rights and claims. Once they have been finally realised and their net proceeds applied to creditors, the fund will owe no further debt and creditors may not take any further steps against it to recover any further sum.

But, as we shall see, sometimes asset managers can be a malign influence, and try to further limit this.

Multi-issuance repackaging vehicles: secured, limited recourse

Secured, limited recourse obligations are de rigueur for multi-issue repackaging SPVs. They save the cost of creating a whole new vehicle for each trade, and really only do by contract what establishing a brand new espievie each time would do through the exigencies of corporation law and the corporate veil. The point is to completely isolate each set of Noteholders from each other. This is a surprisingly narrow point, as we will see, so we should not get carried away for the formalities of security.

With secured, limited recourse obligations there is a quid pro quo: all creditors are known; they are yoked to the same ladder of priorities; they all have agreed to limit their claims to the liquidated value of the secured assets underlying the deal. In return, the espievie grants them a first-ranking security over those assets — mediated between them by the agreed priority structure — and this stopping any interloper happening by and getting its mitts on the espievie’s assets.

The key point to absorb here: this is not a material economic modification to the deal. The line it draws, it draws around all the assets underlying the deal: the underlying securities, cashflows deriving from them, the espievie’s rights against custodians and bankers holding them, and its rights against the swap counterparty — everything, tangible or otherwise, of financial value in the transaction is locked down and pledged to secured parties, and the intercreditor arrangements, too, are fully mapped out. This kind of limited recourse, in fact, doesn’t limit recourse: it maps practical recourse, exactly to the totality of assets that the issuer has available for the purpose: all it saves is the unnecessary process of bankrupting a shell company with nothing left in it in any case. Secured limited recourse is like a nomological machine; a model; it is a simplified account where everything works as it should do, there are no unforeseen contingencies, and all outcomes are planned.

We shouldn’t get too hung up about the whys and wherefores of the security structure of a repackaging as long as it is there, it covers all the rights and assets it is meant to cover, and all necessary perfections and execution formalities are observed. For in a repackaging, the security just sits there and will almost certainly never be exercised.

All that tedious business about automatically releasing it to make payments, powers to appointing receivers, calling and collecting in, the trustee’s rights and obligations under the Law of Property Act 1925 and so on — look it is all good stuff; let your trustee lawyer have his day — but as long as it is there, none of it really matters.

Why? Because — unless you have negligently buggered up your ring-fencing and your Trustee has let you: both of these are quite hard to do — the SPV cannot go insolvent. Any repack redemption will be triggered by an external event: a non-payment on an underlying asset or by a failing counterparty or agent. None relate to the solvency or ability to meet its debts of the Issuer itself.

That being the case, once it exists, the security package will never actually do anything: any diminution in value to of the secured assets — will happen regardless of how strong the security is. The security is a formal belt and brace there to fully isolate from each other the noteholders of different series, and even that only matters only when the SPV is bankrupt. Which is, never.

The limited purpose of the security package in a repackaging is widely misunderstood – all it does is defend against unexpected holes in the ring-fencing.

This is why it is de rigueur to accelerate, liquidate and distribute the proceeds of a repackaged note without enforcement of the security.

“Extinction” versus “no debt due”?

When pondering limited recourse language, more passionate students may be interested to learn of an arid discussion practitioners once had about whether it was better to say, upon exhaustion of all assets, that the SPV’s debt was “extinguished”, or “not due” in the first place. This was weakly supported by coal-face lawyers at places like Linklaters (in the “no debt due” camp) and Clifford Chance (“extinction”), vociferously propelled by their partners, but left the rest of us with a hard-to-articulate but quite persistent sense that there really ought to be better things for masters of the universe of structured finance to worry about. Such as, “how on Earth is it that people who fixate about things as tedious as this get paid so much?”

As with all wanton acts of pedantry, this one has an intellectual grounding, but it is so brittle it will make you angry: the concern about “extinction” is that it implies that, at that very instant that final settlement drop drained from the expiring espievie’s veins, the spectral debt was still there, in an amount which, Q.E.D., the company was, theoretically, at that frozen moment in the ineffable hereafter, unable to pay — until in the next infinitesimal subdivision of time, whereupon the debt finally vanished. The concern? That, for just that fleeting angel’s blink, and no more, the espievie was technically insolvent. I know. I know: don’t @ me, folks.

These people prefer to say “no debt is due” on the supposition that this magic spell can have some spooky retrospective effect, banishing phantoms, re-rendering the fresco of time on which the great redeemer has already daubed an arrow, or some such thing.

What difference does it make, in point of practical fact? None at all.

Investment funds

Where you are facing an investment fund held by equity investors it is slightly — but not very — different. Generally, there is no security, since there’s no question of ring-fencing separate pools of assets. (But investment managers can get in the way and steal options, so be on your guard — see below). Limiting recourse to the fund’s entire pool of assets: A provision which says “once all the fund’s assets are gone, you can’t put it into bankruptcy”, looks harmless, seeing as once all the fund’s assets are gone there’s no point putting it into bankruptcy. This is the same place you would be with a single-issue repackaging vehicle: the corporate veil does the work anyway. This provision just keeps the directors of the fund in paid employment. But there’s a subtle cast on this. With no security, and no co-ordination of creditors that is typical of a structured finance deal, with security, a priority of creditors, covenants not to create any other indebtedness and so on, the ecosystem is very much mapped and controlled. In an investment fund, it isn’t. The creditors (competing brokers, prime brokers, swap counterparties, futures clearers and so on) have no idea what each is doing, and there are real benefits to them in the insolvency rules ensuring fair and equitable treatment of creditors in insolvency. The Archegos situation (which as far as I know didn’t involved limited recourse, by the way) illustrates this dynamic pretty well.

Limiting recourse to a pool managed by an agent

On the other hand, limiting recourse to a pool of assets within a single fund entity — say to those managed by a single investment manager (some funds subcontract out the management of their portfolios to multiple asset managers) — being a subset of the total number of assets owned by the fund — is a different story altogether. This, by sleepy market convention, has become a standard part of the furniture, but to the JC and his friends and relations, seems batshit insane.

So firstly, the investment manager is an agent. An agent isn’t liable at all for any of its principal’s obligations. It is a mere intermediary: the JC has waxed long and hard enough about that elsewhere; suffice to say the concept of agency is one of those things we feel everyone in financial services should know; this is not one to drop-kick to your legal eagles: it is fundamental to the workings of all finance.

So why would an agent seek to limit its principal’s liability to the particular pool of assets that principal has allocated to that agent?

Probably because the principal has said, “I don’t trust you flash city types, with your Sharpe ratios and your intelligent beta. If I am not careful you could put on some insanely cavalier spread play on the seasonal convergence of natural gas futures and blow up my whole fund. I don’t want you to do that. So I am only prepared to risk the assets I give to you, and that is the end of it.”

What the principal is doing here is broadly of a piece with segregated, ring-fenced repackaging. She is saying, “swap dudes: to stop my agent, against whom you are trading, going properly postal and blowing up my whole operation I am limiting it, and therefore you, to, this bucket of assets. Cut your cloth accordingly.”

And that would be fine, if it were like a ring-fenced repack. But it is not quite: for one thing, poor swap dealer has no security over the pool. It gets no “quid” for its “quo”. It is limited to that pool of assets, but it has no priority over them, as against other general creditors of the fund, as it would if it had security. It may find itself not only limited to the pool of assets but even then only recovering cents in the dollar on them. Double whammy. You could fix that by having the fund represent that all other creditors are similarly limited to other pools of assets, so every creditor has its own dedicated bucket — but that is messy and unreliable. Are there really no other creditors? What about people claiming under a tort?^[3] Granting security is much cleaner and neater — but you’ll never get it. Somehow, asset managers have won this battle. Swap dealers the world over run this structural risk. One day this might come back to nip them on their bottoms, like an angry black swan. Who can say?

Note that the principal — or more likely the agent — is engaged in some dissembling here. So the principal wants its agent on a short leash. Fine; understood. Fair. But whose problem should that be? Who should carry the can when an asset manager exceeds its mandate, goes crazy-ape bonkers, or just royally messes up? Not, we would submit, an arm’s length swap dealer trading in good faith, for value and without notice of turpitude. The incentives are all wrong. Isn’t the principal going to be inclined to keep a tight rein on the value of that pool, rather than, as it should, on its agent?

The general principles of agency, we submit, say this should be the principal’s problem. Choose your agents wisely, and monitor them. If not the principal (“I did! I even conducted due dilly! I monitored! Daily!”), then the agent’. For if the agent’s own principal can’t be expected to have rooted out this canker, what chance did a poor old swap dealer have?

The one person whose problem it should not be is the poor, weather-beaten old swap dealer. Yet this is what agent-pool recourse limitation does. It transfers agent screw-up risk — perhaps a second-loss risk, but still a material risk, since you have unreasonably limited the first-loss to an arbitrary number — to the swap dealer. It is hard to understand why a swap dealer would ever agree to this. The answer likely to come: “Well, all our other counterparties have agreed this.” Alas, in this particular case, the agent is probably right.

The asset pool is indeterminate

Secondly, a pool of assets for the time being allocated to an investment manager is kind of nebulous. What the client giveth, the client can taketh away. If the client’s asset manager has gone rogue, that is exactly the time at which it will be anxiously raking its assets back. So the swap dealer facing that pool of assets — who has been faithfully handling and executing all orders competently and in good faith, of course — may find that nice big juicy bucket of assets to which it has limited its recourse, suddenly has a hole in it.

Now you might extract a covenant from your asset manager — even better, from its principal — not to precipitously whip the rug away just as things are getting jiggy — but don’t bet on it. They are likely to appeal to the commercial imperative — fair enough, the JC has a healthy respect for that — but bear in mind it is rather predicated on the iterated prisoner’s dilemma — that there will be another time, there will be more business to do; that the revenue opportunities from cooperating into the infinite & unknowable future far outweigh the value of assassinating the bird sitting in the bush this very instant. But that calculus changes, mightily, for the period between when your Rotary Mazda begins to slide sideways and when, at its current vector, it will hit that oncoming truck.

This is liability cap, not a credit mitigant

Thirdly, and critically: here is where your investment manager might be stealing a put option. A limitation to “a specified pool of assets under management” is, make no mistake, a limitation on the numeric value of your claim. Your exposure is “the greater of the net mark-to-market value to you of your portfolio of transactions and the net market value of this bucket of assets.” This is a sort of put option. A cynical agent could exercise it, it if got in a hole, by failing to pay. That is not what limited recourse is meant to do. Accepting this kind of limitation ought to be trading decision, not a credit decision. Ask yourself why your credit team, rather than trading, are being asked to approve this.