Primer and Analysis of the Non-Subsidiary Dropdown

Welcome to the 212th Pari Passu Edition. 

Eight months on, and people are still talking about the innovative asset dropdown in Xerox. The company had created a ‘Non-subsidiary’ by setting up a Joint Venture LLC with the new money providers, giving the lenders majority voting power over the new entity. We have already covered Xerox and the transaction before in March, but we couldn’t help ourselves and had to rejoin the conversation. 

This week, we are passing the pen from our financial analyst to our legal analyst to put the transaction structure under the microscope to understand how the Non-Subsidiary Joint Venture works, and where it sits alongside other landmark dropdown transactions. We do this by looking at the history of both asset dropdowns and non-subsidiary LMEs (yes, Xerox was not the first, actually!), before running through a simple hypothetical transaction to demonstrate how the Xerox-style non-subsidiary dropdown works step-by-step and its use case: to get around popular asset dropdown blockers. 

From there, this write-up takes an analytical and critical turn. We think there are some features of the non-subsidiary dropdown that everyone is missing. So we want to bring them to your attention and contribute to the wider discussion around these transactions. Our first claim is that it is too narrow to equate non-subsidiaries with minority-owned joint venture corporations. If we start from first principles, we can see that a joint venture corporation is a means to accomplish a non-subsidiary dropdown, but it is conceivable that there are other ways to achieve this. Next, we claim that it is worthwhile to explore other options, as joint venture corporations do come with significant legal disadvantages for the borrower. In light of these disadvantages, we propose two alternative non-subsidiary structures, using US and UK law, that a borrower can use to avoid giving the incumbent lender majority voting power / an equity position in the entity with the valuable assets. We then conclude with some thoughts on recent market developments and how to think about Xerox Blockers given what we’ve learned today. We are incredibly excited to bring this analysis to you!

It is important to start from the fundamentals and build up so we can delve into the legal complexity behind non-subsidiary dropdowns that much of the market has been missing out on. However, for those familiar with the history of asset dropdowns, feel free to skip ahead to ‘Historical Context of the Non-Subsidiary’.

There are three overarching types of subsidiaries outlined in common credit documents [1]:

The history of dropdowns, many of which we have covered in our research, typically takes the following structure. The obligor entities move assets to an NGRS and/or Unsub. Then, as old liens are released, the NGRS or Unsubs will raise new debt against the transferred assets on a structurally senior basis to the debt at the obligor entities [4]. The new money raised at the NGRS and Unsubs is then forwarded up to the obligor entity. The transferred asset is commonly the borrower’s intellectual property (IP) because IP is very portable. Unlike a physical asset, it can be transferred between entities and then licensed back to the operating business with little to no operational disruption [2]. In the case of software companies, source code and related patents are the most valuable assets software companies have. Further, IP often lacks a readily observable market price because of its heterogeneous nature, giving the borrower greater flexibility in determining FMV for covenant purposes [2]. Let’s briefly go over two landmark transactions particularly relevant to us.

J. Crew, Unsub Trapdoor: J. Crew was able to create new collateral to support a refinancing due to its credit agreement. The transaction involved the use of three transfer baskets (‘baskets’ are agreed exceptions to covenants that prohibit certain actions for the borrower; an asset transfer exception here): (i) the $150mm common intercompany basket for investments by obligors in NGRS; (ii) the $100mm common general use investment basket into Unsubs; and (iii) the ‘trap door’ basket which permitted investments by NGRS in Unsubs if such investment is financed with the proceeds of other permitted investments [4]. Simply put, “other permitted investments” were the initial $250mm of IP transferred into the NGRS under baskets (i) and (ii), and the trapdoor allowed J.Crew to treat that same $250mm as the funding for a second transfer from the NGRS into the Unsub, without needing another $250mm of fresh investment capacity.

Therefore, the three baskets combined to produce the dropdown: the first two baskets allowed J.Crew to push $250mm of IP into an NGRS, and the third basket allowed the NGRS to make investments to the extent they were “financed with the proceeds received” [4]; the NGRS just received $250mm of IP and so the same $250mm can pass onwards to the Unsub. As the Unsub was not bound by covenants in the credit documentation, there were no limits on such a subsidiary incurring debt.

Pluralsight, NGRS Dropdown: Other than being the first mainstream private credit LME, Pluralsight was also significant as the final destination of the assets was an NGRS, rather than an unsub. The assets remained inside the restricted group, but the new entity did not guarantee the existing private-credit debt. Liens securing the legacy lenders were released from the transferred IP, while financing raised at the NGRS benefited from structural priority over those legacy claims. The point to take away is that obtaining structurally senior debt does not require moving assets outside the restricted group: moving them outside the obligor group may be enough, assuming the restricted group's negative covenants still provide freedom to incur enough debt.

All in all, we can see that transferring assets to NGRS and Unsubs proves to be an effective way to obtain unencumbered collateral to raise new money against. However, the ability to pull such maneuvers off is contingent on the covenants and the respective baskets which allow for assets to move into subsidiaries. Subsequent covenant drafting has included ‘blockers’ named after these landmark transactions, albeit not as prominent given that we are very much still in a covenant-lite era. Still, many companies will have to get creative to find workarounds where their credit agreements contain an array of these blockers.

Despite nearly a decade of innovation, all these transactions had one feature in common. The recipient of the asset always fell into two overarching types of subsidiaries commonly found in credit documents: Unsubs (J. Crew, Envision, and Del Monte) and NGRS (Pluralsight). But what if a borrower still wants to raise new money against valuable assets, while its credit documents contain LME blockers restricting transfers into both Unsubs and NGRS?  Well, some creative lawyers in the Xerox transaction appeared to have challenged an underlying assumption: why should the recipient of the assets be a ‘Subsidiary’ at all? 

At this stage, it’s worth getting granular about what a ‘Subsidiary’ actually means. For the purposes of credit documents and, by extension, LMEs, this is principally a contractual question: the documents themselves define which entities are classified as a Subsidiary. Here is a relatively common, but narrow definition in US credit agreements, found in Xerox’s indentures:  

In Xerox, the legalese largely boils down to identifying whether Xerox owns or controls more than 50% of the votes that can be used to choose the people who manage the entity. This would typically be ordinary shares in a company. If Company X controls more than 50% of the votes attached to equity interests entitled to elect Company Y’s directors, managers or trustees, Y falls within the Subsidiary definition and is a subsidiary of X. But if X owns 50% or less, Y may sit outside it. This creates a potential fourth recipient of dropdown assets: a ‘Non-Subsidiary’ (Non-sub), neither an NGRS nor an Unsub and therefore outside restrictions drafted solely by reference to those categories.

However, before proceeding, because the definition of ‘Subsidiary’ is a function of contract, it is very important to note that Xerox's voting-power structure and the transaction that follows can only explain one way of creating a Non-sub. The definition of Subsidiary depends on the particular test for it in the credit document. To demonstrate, there are two LMEs that preceded Xerox that found other ways to place an entity outside of this ‘Subsidiary perimeter’, so to speak. 

Trinseo

Here is Trinseo’s Subsidiary definition:

Even with a brief skim, you can tell that the definition is far more robust than what was found in Xerox’s documents. Here, there are three ways an entity would qualify as a Subsidiary: (i) having majority voting control (as in Xerox); (ii) ownership of more than half its issued share capital (which is not necessarily the same as having voting rights, i.e., preference shares typically do not come with voting rights); or (iii) control over its management. However, there is an additional qualification at the end. Unless specified otherwise, references to a ‘Subsidiary’ meant a Subsidiary of the Lead Borrower, Trinseo Materials Operating S.C.A.  This reflected how the Lead Borrower sat below separate holding companies, which the credit documents made efforts to exclude from the restricted group. Put simply, if a parent company sitting above the Lead Borrower directly owned another company alongside it, that sister company could remain part of the wider Trinseo corporate group without being a ‘Subsidiary’ for purposes of the credit agreement, because the Lead Borrower itself did not own or control it. By contrast, if the Lead Borrower owned an intermediate company which in turn owned that entity, it would still fall within the definition indirectly.

Thus, when Trinseo completed its September 2023 pari-plus financing, it placed the new primary obligor, Trinseo LuxCo Finance SPV, at a sister level outside the legacy borrower's ownership chain. The SPV remained within the wider Trinseo corporate family but was not owned or controlled by the Lead Borrower and therefore was not one of its Subsidiaries; this has been otherwise described as a “non-restricted affiliate” company [10]. As for why Trinseo bothered with a Non-sub structure, it comes back to the credit documents: An NGRS remained constrained by Trinseo’s debt covenants and lacked capacity to incur the $1.1bn new-money loan, while an Unsub would be subject to separate restrictions that would complicate its ability to hold the intercompany claim against the existing obligor group. The Non-sub company was not subject to the covenants affecting restricted and unrestricted subsidiaries. The LuxCo SPV could therefore incur the new debt outside the legacy covenant package and on-lend the proceeds into the restricted group. 

Robertshaw

Robertshaw is a particularly important case for us to understand as it laid the foundation for Xerox’s non-subsidiary LME. But first, it is important to understand the context and the objectives of the transaction briefly before looking at Robertshaw’s Subsidiary definition. 

Following an aggressive uptier in May 2023, Invesco held a majority of Robertshaw’s first-out and second-out notes [17]. This was significant because Invesco obtained “Required Lender” status, which allowed it to bind the rest of the lenders on certain matters under the Super-Priority Credit Agreement (SPCA) [11]. Invesco used its controlling position in the January 2024 Chapter 11 bankruptcy filing, where Invesco lined itself up to be the DIP lender and ‘Stalking Horse Credit Bidder’ (the opening bidder in a bankruptcy sale, using its secured debt claim as part of the purchase price). The ad hoc group (AHG) of Bain, Canyon, and Eaton Vance had a strong incentive to avoid this position, as little to no value would have trickled down to them in the event of a credit bid.

Evidently, the AHG needed to find a way around Invesco’s “Required Lender” status. Authorizing Robertshaw to issue new debt to dilute Invesco’s voting share was not an option. Rather, the AHG pursued a strategy that was the inverse of the Wesco/Incora transaction [10]. Instead of increasing the amount of debt held by the ad hoc group, Robertshaw was to reduce the amount of debt held by Invesco. As Invesco held a disproportionate share of the First-Out loans, repaying that tranche removed far more of Invesco’s voting debt than the AHG’s, which would give the AHG “Required Lender” status. The problem, however, is that Robertshaw needed the cash to do this. Raising financing within the restricted group was constrained by the SPCA’s covenants on additional indebtedness and its blockers against manipulating lender voting rights.

The next move then was to raise the financing outside of the restricted group. However, a host of LME blockers, though, meant that virtually all of Robertshaw’s subsidiaries would have been within this restricted group. The solution? Well, it was in Robertshaw’s definition of a Subsidiary:

At first glance, the first limb of the definition of a Subsidiary looks very similar to Xerox’s definition in Figure 1. An entity was a subsidiary where more than 50% of its voting power was owned or controlled by the would-be parent company.  The key theme that runs from Trinseo to here is that a ‘non-subsidiary’ is often a powerful way to avoid being caught within the restricted group (a non-sub is not invariably outside the reach of covenants; we cover this in detail later in the write-up). If a covenant applied only to the borrower and its Subsidiaries (obligor, restricted, and unrestricted), a financing vehicle structured so that it did not satisfy the contractual Subsidiary definition could fall outside that covenant perimeter. Robertshaw sought to exploit the same idea under its SPCA. Its solution was to create a new vehicle that could sit outside of the ‘subsidiary perimeter’ of the SPCA, so that it would be exempt from the covenants. 

RS Funding was created to be the ‘non-subsidiary’. Importantly, we can see that the ownership structure was split. Investor Holdings held 100% of voting power in RS Funding, while Robertshaw held 100% of its economic interests. The logic was that if the SPCA defined a ‘Subsidiary’ primarily by voting control, RS Funding could fall outside that definition, even if Robertshaw remained economically exposed to it. This enabled Robertshaw to move its assets to RS Funding to raise the financing and pass the cash back to Robertshaw. Robertshaw would then use that cash to repay enough First-Out debt to push Invesco below the Required Lender threshold. This ultimately allowed the AHG to use its status as the “Required Lender” to amend the credit document and push towards a restructuring deal that would have been more favorable to them. 

However, it was not smooth sailing from there. Those of you with a keen eye would have noticed that the SPCA Subsidiary definition in Figure 4 had a second limb. That is, “any subsidiary of Holdings other than an Unrestricted Subsidiary.” Without getting too deep into the legal arguments, because “any subsidiary” was in lowercase, Judge Christopher Lopez held that it had to be given its ordinary meaning (as opposed to any technical / contractually defined one of “Subsidiary”), which meant that RS Funding was a subsidiary notwithstanding the deliberate separation of voting and economic ownership [12]. 

Despite this, the transaction was not unwound. This is because the SPCA itself had set out what would happen if unauthorized debt was incurred. It prescribed that in the event such debt was incurred, only a monetary damages remedy would be available. As the court will respect the bargain the parties actually struck, arguments that the transaction should be unwound as a matter of necessity because damages were an inadequate remedy failed [12]. It is immaterial that absent the prescribed remedy in the SPCA, the transaction would have been unwound. All Invesco received was a monetary damages claim, and so the AHG kept its “Required Lender” majority, and the court approved a credit-bid sale of Robertshaw’s assets to the AHG, despite Invesco originally owning the majority of the notes after the May 2023 uptier. 

Ultimately, the key takeaway for us is that there is no single corporate structure that makes an entity a Non-sub. Whether an entity falls outside the Subsidiary perimeter depends entirely on how “Subsidiary” is defined in the relevant credit documents. Trinseo achieved this by placing its financing vehicle outside the Lead Borrower’s ownership chain; Robertshaw attempted to achieve it by separating voting control from economic ownership. 

Xerox would take the latter idea and combine it with an asset dropdown. Before turning to our walkthrough of the Xerox-style non-sub dropdown, it is worth discussing what subsidiary typically means in English credit agreements for our international readers. While the definition of a subsidiary in facility agreements is also a function of contract, like in the US context, in England there is more of a convention to define ‘Subsidiary’ by reference to what is given in the Companies Act 2006. Nothing forces the relevant parties to take these definitions provided in the statute; it is rather just market practice. These are found in sections 1159 and 1162. We will return to the s1159 definition later in our analysis portion of the write-up. 

Evidently, the English statutory concepts are generally more robust than the simple majority-voting test we saw in Xerox. Section 1159 provides that a company can also be a subsidiary where its parent has the right to appoint or remove a majority of its board, or control a majority of its voting rights through an agreement with other members, in addition to holding a majority of voting rights outright. Section 1162 goes even further. The broader concept of a “subsidiary undertaking” additionally captures relationships involving a right to exercise dominant influence, actual dominant influence or control, and even undertakings managed on a unified basis. For our purposes, we still must consider the specific credit documents definition of a subsidiary in any case, but we generally can take away that English law credit agreements are likely to have a wider definition of a subsidiary, especially where they incorporate the holistic s1162 definition, making it harder to pull off the non-subsidiary structure. 

Now that we have covered the history of asset dropdowns and non-subsidiary LMEs, we are ready to delve into the non-subsidiary dropdown! We have devised a hypothetical to understand the logic and mechanics of the transaction structure.

Consider Company ABC, a sponsor-backed software company. Five years ago, ABC was generating $200mm EBITDA. However, customer churn due to new competitors entering its market resulted in its EBITDA declining to $120mm. As its performance began to deteriorate, ABC completed a stressed amend-and-extend (A&E) with its existing lenders. In exchange for maturity relief, those lenders significantly tightened the credit documents. A few years after, management has to put in place a new product rollout and turnaround program to improve performance over the next two years but needs cash to operate till then. It only has $35mm in cash and is burning $75mm a year. Furthermore, $250mm of unsecured notes are maturing in 24 months following the initial A&E, and those are trading around 40 cents on the dollar. Its capital structure is as follows:

Company ABC has 10x gross leverage, and trading prices clearly indicate the business is worth a single-digit EBITDA multiple and the sponsor is deeply out of the money. There is one valuable asset left in the business: ABC’s IP, which management values at approximately $300mm. However, that IP already forms part of the collateral securing ABC’s First-Lien Debt. Management calculates that ABC needs approximately $175mm of total liquidity to execute its turnaround. It already has $45mm of pro forma liquidity, meaning it needs to raise another $130mm. Around $75mm would fund projected cash burn, while $100mm could theoretically repurchase the entire $250mm of unsecured notes at their current 40-cent trading price (40% of 250 = 100; we are ignoring the price impact of initial purchases, for simplicity purposes). If successful, ABC would therefore obtain sufficient runway while eliminating its nearest maturity.

Unfortunately, most conventional options are unattractive. ABC’s sponsor is unlikely to inject further equity considering their deeply out-of-the-money position, pressure from LPs to avoid ‘throwing good money after bad money’, and the strong possibility that any equity injection will turn into a gift to creditors. Refinancing through unsecured debt is similarly unrealistic when ABC’s existing unsecured notes trade at 40 cents on the dollar, while its 8x secured leverage leaves little room for additional pari passu secured borrowing. An amend-and-extend or distressed exchange could push out the 2028 maturity but would not itself provide the cash required to fund the turnaround. Chapter 11 could solve both problems through DIP financing and a balance-sheet restructuring, but at substantial cost and with the court supervision, creditor leverage, execution risk of an in-court process, and the crystallization of the sponsor loss. ABC therefore wants an out-of-court financing solution.

Avid Pari Passu readers would know to suggest unlocking ABC’s remaining collateral through an asset dropdown: transfer the $300mm of IP to an Unsub or NGRS, release the existing liens, and raise structurally senior debt against it. After all, the fact pattern is not so dissimilar to J. Crew. But, after ABC’s stressed A&E, its lenders put in place an array of LME blockers, which can be thought of as ABC's consideration to obtain maturity relief. This is a common feature of post-LME or stressed amendment processes; lenders will often tighten baskets and add the blockers to close standard and prevalent LME transaction structures. In the amended credit agreement, there is a J. Crew blocker that prohibits transfers of Material IP to Unsubs,  and a Pluralsight-style blocker that restricts transfers of Material IP to NGRS.

What can ABC do? ABC has valuable collateral that investors would be willing to lend against. In fact, third-party investor XYZ has made its interest known. Well, what can we learn from Trinseo and Robertshaw? In Trinseo, the non-sub was principally used as an intercompany financing vehicle. In Robertshaw, the non-sub was created to find a way around the “Required Lender” status. In our hypothetical, the end is to have a recipient vehicle that could become the destination for the valuable IP and raise new money against it. Yet, in all these situations, there is the same problem. The Obligor/NGRS/Unsub are all ‘Subsidiaries’ that have restrictions or are affected by covenants that make them unworkable for the given goal. Luckily, Trinseo and Robertshaw have shown us we can work around these problems by creating a non-sub by paying close attention to the credit agreements, even if the ends of the transaction are different. So, let’s have a look at how Subsidiary is defined here: 

Defining a subsidiary, or at least an aspect of it, by reference to majority voting control is commonplace in US credit documents [1]. It is therefore an appropriate definition for our example. 

Now, how is it that the Joint Venture (JV) Non-Sub gets around this Subsidiary definition? Well, ABC can transfer its IP to a joint venture in which it does not hold majority voting control. If ABC owns 49% of the voting power and Investor XYZ owns 51%, the JV is not a ‘Subsidiary’ of ABC under the definition above. The recipient JV company is thus unaffected by the LME blockers. 

We have broken down this transaction structure into six steps with diagrams to make it easy to follow.

Step 1: Company ABC and Investor XYZ form JVCo. Let’s start with the following structure chart.