PRIVILEGED AND CONFIDENTIAL ATTORNEY-CLIENT COMMUNICATION / ATTORNEY WORK PRODUCT
This memorandum is a privileged and confidential attorney-client communication and attorney work product. It is intended solely for the use of the addressees and the persons copied hereon. Any unauthorized disclosure, distribution, or copying of this memorandum or any of its contents is strictly prohibited. If you have received this memorandum in error, please notify Calderwood & Harkness LLP immediately and destroy all copies.
CALDERWOOD & HARKNESS LLP 1200 Avenue of the Americas, 42nd Floor New York, New York 10036 Telephone: (212) 555-7800 Facsimile: (212) 555-7801
MEMORANDUM
FROM: Stephanie Okafor, Partner, Calderwood & Harkness LLP Daniel Obi, Lead Associate, Calderwood & Harkness LLP
TO: Marcus Delaine, Chief Executive Officer, Ironhaven Data Centers LLC Rebecca Sung, Chief Financial Officer, Ironhaven Data Centers LLC Jonathan Pryce-Williams, General Counsel & Secretary, Ironhaven Data Centers LLC Natasha Greenfeld, Vice President, Treasury & Capital Markets, Ironhaven Data Centers LLC
CC: Elena Castillo, Partner, Thornfield Capital Partners, LP Graham Ashworth, Principal, Thornfield Capital Partners, LP Philip Hartmann, Billing Partner, Calderwood & Harkness LLP Lisa Morrow, Working Associate, Calderwood & Harkness LLP Sylvia Hartwell, Specialist Partner (Tax), Calderwood & Harkness LLP
DATE: March 25, 2024
RE: Issues List and Negotiation Points — Blackthorn & Sundial Software Initial Draft Credit Agreement dated March 22, 2024 — $950,000,000 Senior Secured Term Loan B Credit Agreement
Matter No.: 1043-00001 Billing Code: IHDC-TLB-2024
Right-click to update Table of Contents
I. Executive Summary
The purpose of this memorandum is to catalogue and analyze the open business and legal issues arising from our review of the initial draft Senior Secured Term Loan B Credit Agreement (the "Agent Draft" or "Draft Credit Agreement") prepared by Blackthorn & Sundial Software (Catherine Ellsworth, Partner, and Derek Huang, Senior Associate) on behalf of Crescent Harbor Bank, N.A., as Administrative Agent (the "Administrative Agent" or "Agent"), and the Lenders party thereto. The Agent Draft was received by Calderwood & Harkness on the afternoon of Friday, March 22, 2024. Our team has conducted a comprehensive review of the Agent Draft over the weekend and has identified the issues discussed in detail below.
Transaction Summary. As the client team is aware, the transaction under negotiation is a $950,000,000 Senior Secured Term Loan B credit facility (the "Facility"), structured as follows:
• Borrower: Ironhaven Data Centers LLC, a Delaware limited liability company ("IHDC" or the "Borrower").
• Co-Borrower: Ironhaven Frankfurt GmbH, a company organized under the laws of the Federal Republic of Germany ("IHDC Frankfurt"), for the €253,571,429 EUR sub-tranche.
• Parent Guarantor: Ironhaven Holdings Inc., a Delaware corporation ("Holdings").
• Administrative Agent: Crescent Harbor Bank, N.A. (the "Administrative Agent").
• Collateral Agent: Pinnacle Trust Company, N.A. (the "Collateral Agent").
• Facility Amount: $950,000,000 in aggregate, consisting of (i) a USD tranche of $675,000,000 and (ii) a EUR tranche of €253,571,429 (approximately $275,000,000 at the applicable exchange rate).
• Maturity: Seven (7) years from the Closing Date, maturing June 3, 2031.
• Interest Rate: SOFR + 425 basis points (USD tranche) / EURIBOR + 425 basis points (EUR tranche), with a 50 basis point SOFR/EURIBOR floor.
• Original Issue discount: 99.50 (50 basis points).
• Amortization: 1.00% per annum (paid quarterly).
• Sponsor: Thornfield Capital Partners, LP, acting through Thornfield Infrastructure Fund III, LP.
Baseline Reference Document. Our analysis uses the Commitment Letter dated February 28, 2024 (the "Commitment Letter"), including the annexed term sheet (the "Term Sheet"), as the baseline for agreed-upon terms between the parties. The Commitment Letter was negotiated over a four-week period and represents a carefully balanced set of economic and structural terms. Our review reveals that the Agent Draft deviates from the Commitment Letter Term Sheet in several material respects, as identified and discussed below.
Analytical Approach. For each issue identified in this memorandum, we set forth: (a) the position taken in the Agent Draft, (b) Calderwood & Harkness's recommended position for Ironhaven (the "Recommended Position"), and (c) a fallback compromise position in the event the Recommended Position is not achievable in negotiation. Each issue is also assigned a priority category as follows:
• Category A — Must Achieve: Issues that are critical to the deal economics, structural soundness, or enforceability of the transaction, where the Recommended Position should be pursued without significant compromise. There are eight (8) Category A issues.
• Category B — Strongly Preferred: Issues that are important to Ironhaven's operational flexibility or risk profile, where negotiation is appropriate but compromise is acceptable within defined parameters. There are twelve (12) Category B issues.
• Category C — Nice-to-Have / Trading Points: Issues that would be beneficial to Ironhaven but that may be conceded in whole or in part as part of the broader negotiation dynamic, used as currency to achieve favorable outcomes on Category A and B issues. There are seven (7) Category C issues.
Overall Assessment. In our professional judgment, the Agent Draft is a relatively aggressive lender-favorable document. While it is not unusual for an initial agent's draft to lean in favor of the lender group, the degree of deviation from the agreed Commitment Letter Term Sheet on several key economic terms — including the leverage ratio, the excess cash flow sweep, the soft call premium, and the change of control definition — is notable and goes beyond what we would characterize as routine "papering" differences. We have identified approximately twenty-seven (27) discrete issues across the Agent Draft, of which eight are Category A. We believe the Commitment Letter provides strong support for Ironhaven's positions on all Category A items, and we are confident that the market for data-center TLB financings supports the positions outlined herein.
Target Timeline. We recommend that Ironhaven's mark-up of the Agent Draft be transmitted to Blackthorn & Sundial Software on or about April 5, 2024. Based on customary negotiation timelines for transactions of this nature, we anticipate reaching substantive agreement on the draft by mid-April 2024, with the objective of circulating a final agreed-form Credit Agreement by May 1, 2024, in advance of the May 15, 2024 targeted signing date and the June 3, 2024 anticipated closing and initial funding.
We respectfully request that the Ironhaven client team and Thornfield Capital Partners review this memorandum and provide comments and approvals in accordance with the action items and target dates set forth in Section XIII below. We are available to discuss any of these issues at the team's convenience and would suggest scheduling a client call for later this week to align on negotiation strategy.
II. Summary Issues Matrix
The following table provides a quick-reference summary of all issues identified in this memorandum. This matrix is intended to facilitate discussion during client management calls and to serve as a tracking tool as the negotiation progresses. Detailed analysis of each issue follows in Sections III through X.
Note: Issue numbers 27 and 28 were added to accommodate the German-law specific items identified as separate issues during our review, bringing the total to 28 discrete issues. For purposes of the Category A/B/C classification, the totals remain 8/12/8 (with the additional German-law item classified as Category C).
III. Financial Covenants
The financial covenant package is, in our view, the single most critical area of negotiation. The Agent Draft has materially tightened the financial covenants relative to the Commitment Letter Term Sheet, and the tightening, if accepted, would leave Ironhaven with dangerously thin headroom during the critical construction phase of the Virginia and Frankfurt campuses. We address each financial covenant issue below.
III.A. Maximum Net Leverage Ratio (Issue No. 1 — Category A)
Agent Draft Position (§7.01(a)). The Agent Draft establishes a Maximum Net Leverage Ratio (defined as Consolidated Total Net Debt to Consolidated Adjusted EBITDA) of 5.50 to 1.00, tested on a quarterly basis commencing with the fiscal quarter ending June 30, 2024. The ratio is flat throughout the life of the Facility with no scheduled step-downs. For purposes of this covenant, "Consolidated Total Net Debt" means all funded indebtedness for borrowed money of the Borrower and its consolidated subsidiaries (including the EUR sub-tranche, capital leases, and purchase money obligations), less unrestricted cash and cash equivalents on hand, on a consolidated basis.
Commitment Letter Baseline. The Commitment Letter Term Sheet provides for a Maximum Net Leverage Ratio of 6.00 to 1.00, with quarterly testing commencing with the fiscal quarter ending September 30, 2024. This was a specifically negotiated point during the Commitment Letter discussions, and we note that it was the subject of considerable back-and-forth between Ironhaven/Thornfield and the Agent before the 6.00x level was agreed. The Commitment Letter is silent on step-downs, which we interpret as a flat 6.00x covenant for the life of the Facility.
Ironhaven Recommended Position. We recommend that Ironhaven insist on the Commitment Letter terms: a Maximum Net Leverage Ratio of 6.00 to 1.00, with the first quarterly test on September 30, 2024 (which represents the first full fiscal quarter following the anticipated June 3, 2024 Closing Date). This recommendation is supported by the following analysis:
Headroom Analysis. Based on the pro forma capitalization at closing, Ironhaven's pro forma Net Leverage Ratio is approximately 5.35x, calculated as follows:
• Pro forma Consolidated Total Net Debt: approximately $887.5 million (consisting of $950 million in Term Loan B borrowings, less approximately $62.5 million of unrestricted cash on hand at closing after payment of transaction costs and OID).
• LTM Consolidated Adjusted EBITDA (through March 31, 2024): approximately $165.8 million.
• Pro Forma Net Leverage Ratio: $887.5M ÷ $165.8M = 5.35x.
At a 5.50x covenant level, Ironhaven would have only approximately 15 basis points (0.15x) of headroom at closing. This level of headroom is insufficient given the construction-phase dynamics discussed below.
Construction-Phase EBITDA Variability. Ironhaven is currently undertaking two major data-center construction projects:
• Ironhaven Ashburn I (Virginia campus, Phase I): 60 MW critical IT load capacity; total estimated construction cost of $475 million; anticipated completion Q4 2025.
• Ironhaven Frankfurt I (Frankfurt campus, Phase I): 40 MW critical IT load capacity; total estimated construction cost of approximately $275 million / €253.6 million; anticipated completion Q1 2026.
During the construction ramp period (expected through late 2025 / early 2026), Ironhaven's Adjusted EBITDA is expected to experience temporary compression due to: (i) increased staffing and pre-operational costs at the new campuses; (ii) ramp-up timing for tenant occupancy (even under the existing pre-lease arrangements with Stratos Cloud Services and Euronova Digital, occupancy will not commence until construction completion); and (iii) potential one-time commissioning and integration expenses. Ironhaven's internal financial model (shared with Calderwood & Harkness on a privileged basis) projects that Adjusted EBITDA may temporarily decline to a trough of approximately $155 million in the fiscal quarter ending March 31, 2025, before recovering as the new campuses begin generating revenue. At a $155 million EBITDA trough, the Net Leverage Ratio would be approximately 5.73x — well within a 6.00x covenant but in technical default under a 5.50x covenant.
Testing Commencement Date. The June 30, 2024 testing commencement date in the Agent Draft is problematic because the Closing Date is anticipated to be June 3, 2024, meaning the first test would capture a period in which only 27 days (June 3–30) of post-closing activity would be included. This creates annualization distortions and does not provide a meaningful compliance test. September 30, 2024 is the appropriate first testing date, as it will be the first full fiscal quarter following the Closing Date and will provide a more representative baseline for ongoing compliance measurement.
Fallback. If the Recommended Position of 6.00x is not achievable, we would accept a compromise of 5.75x, provided that: (i) the testing commencement date remains September 30, 2024, and (ii) a construction-period equity cure right is included in the Credit Agreement (as discussed in Section III.D below). Under no circumstances should Ironhaven accept a 5.50x covenant — the headroom is simply too thin given the construction timeline and EBITDA variability.
III.B. Minimum Interest Coverage Ratio (Issue No. 2 — Category A)
Agent Draft Position (§7.01(b)). The Agent Draft establishes a Minimum Interest Coverage Ratio (defined as Consolidated Adjusted EBITDA to Consolidated Cash Interest Expense) of 2.25 to 1.00, tested on a quarterly basis commencing with the fiscal quarter ending June 30, 2024.
Commitment Letter Baseline. The Commitment Letter Term Sheet provides for a Minimum Interest Coverage Ratio of 2.00 to 1.00, with quarterly testing commencing with the fiscal quarter ending September 30, 2024.
Ironhaven Recommended Position. We recommend that Ironhaven insist on the Commitment Letter terms: a Minimum Interest Coverage Ratio of 2.00 to 1.00, with the first quarterly test on September 30, 2024.
Interest Cost Analysis. At the applicable interest rate of SOFR + 425 basis points, with a 50 basis point SOFR floor, and assuming current market SOFR at approximately 5.33%, the all-in interest rate on the USD tranche would be approximately 9.58% (5.33% + 4.25%). On the full $950 million Facility (assuming a blended rate across the USD and EUR tranches), annualized cash interest expense is estimated at approximately $91 million.
The Adjusted EBITDA of $165.8 million (LTM March 31, 2024) yields a crude interest coverage ratio of approximately 1.82x on a gross basis. However, the Consolidated Cash Interest Expense calculation under the Agent Draft definition is net of interest income and includes certain adjustments, and the Adjusted EBITDA calculation includes the addbacks discussed in Section III.C below. On a properly calculated basis with the addbacks we are requesting, we project the Interest Coverage Ratio at closing to be approximately 2.15x. This provides adequate headroom above a 2.00x covenant but would be in breach of a 2.25x covenant during the EBITDA trough period.
At the projected EBITDA trough of $155 million (fiscal quarter ending March 31, 2025), the Interest Coverage Ratio would be approximately 1.70x on a gross basis — below both the 2.00x and 2.25x levels. However, with the appropriate EBITDA addbacks (including the run-rate addback for contracted pre-lease revenue from Stratos and Euronova), we project the ratio would be approximately 2.10x, providing adequate but not excessive headroom above a 2.00x covenant. A 2.25x covenant would leave no room for any variance from projections.
Fallback. We recommend no further compromise on this issue. The 2.00x ratio is the Commitment Letter term and is the market standard for construction-phase term loan B facilities in the data-center sector. Based on our review of comparable transactions (summarized in Appendix B), the median minimum interest coverage ratio for recent data-center TLBs is 2.00x, with a range of 1.75x to 2.25x. The 2.25x level in the Agent Draft is at the very top of the market range and is inconsistent with the construction-phase profile of this transaction.
III.C. Definition of "Consolidated Adjusted EBITDA" (Issue No. 3 — Category A)
Agent Draft Position (§1.01 — Definitions). The Agent Draft includes a definition of "Consolidated Adjusted EBITDA" that is, in our assessment, unusually narrow and omits several standard borrower-favorable addbacks that are customarily included in sponsor-backed term loan B facilities, particularly those involving construction-phase businesses. Specifically, the Agent Draft definition:
1. Omits any addback for pro forma cost savings from contracted but not-yet-commenced tenant leases. Ironhaven has executed letters of intent with two anchor tenants: (a) a pre-lease with Stratos Cloud Services for 30 MW of critical IT load at the Ashburn I campus, at an annual rental rate of approximately $14.4 million; and (b) a pre-lease with Euronova Digital for 20 MW of critical IT load at the Frankfurt I campus, at an annual rental rate of approximately €9.8 million (approximately $10.7 million at current exchange rates). These contracted revenue streams will commence upon tenant occupancy following construction completion but are not reflected in historical EBITDA. The absence of a pro forma addback for this contracted revenue materially understates the Borrower's earnings capacity and artificially tightens the financial covenants.
2. Caps the addback for non-recurring charges at 10% of EBITDA. The market range for this cap in comparable TLB facilities is 15% to 25% of EBITDA. A 10% cap is unduly restrictive and may not accommodate the non-recurring construction-phase costs that Ironhaven is expected to incur.
3. Omits any addback for non-cash stock or equity-based compensation. This is a standard addback in virtually all sponsor-backed credit facilities and its omission is unusual.
4. Omits any "run-rate" or "synergies" addback for new capacity coming online. In the data-center sector, where new capacity additions are the primary driver of revenue growth, a run-rate addback for contracted new capacity is customary and essential to the proper functioning of the financial covenants.
Ironhaven Recommended Position. We recommend that Ironhaven request the following modifications to the Consolidated Adjusted EBITDA definition:
(a) Pro forma addback for contracted tenant revenue. Include an addback for the annualized contracted revenue under signed pre-lease letters of intent (specifically, the Stratos Cloud Services and Euronova Digital pre-leases) once the applicable tenants have taken occupancy of the relevant facility. This addback should be calculated on a pro forma, annualized basis and should be available from the quarter in which tenant occupancy commences.
(b) Increased cap on non-recurring charge addbacks. Increase the cap on addbacks for non-recurring, unusual, or extraordinary charges from 10% to 20% of Consolidated Adjusted EBITDA (before giving effect to such addbacks). This is within the market range and provides appropriate flexibility during the construction phase.
(c) Non-cash equity-based compensation addback. Include an addback for non-cash stock-based and equity-based compensation expense incurred by the Borrower and its subsidiaries. This is a universal feature of sponsor-backed credit facilities and its inclusion is not a matter of negotiation in the ordinary course.
(d) Run-rate addback for new contracted capacity. Include a "run-rate" addback for new data-center capacity that has been contracted under binding lease agreements, reflecting the annualized revenue that such capacity is expected to generate at stabilized occupancy levels. This addback should be capped at 20% of historical (trailing four-quarter) Consolidated Adjusted EBITDA to prevent abuse while providing meaningful covenant relief during the construction ramp.
(e) Sponsor management fee exclusion. To the extent Thornfield Capital Partners charges management, monitoring, or advisory fees to Ironhaven or its subsidiaries, such fees should be expressly excluded from any deductions in the EBITDA calculation.
Fallback. If the full suite of addbacks is not achievable, we would accept the following compromise:
• Non-recurring charge addback cap at 15% of EBITDA (rather than 20%);
• Narrower run-rate addback tied only to executed, binding lease agreements (not letters of intent) with tenants that satisfy a credit approval standard (e.g., investment-grade or approved by the Administrative Agent); and
• Retention of the non-cash equity-based compensation addback, which is not a matter that should require compromise.
We note that the EBITDA definition is foundational to the entire financial covenant framework and to numerous other provisions of the Credit Agreement (including the Excess Cash Flow sweep calculations, the leverage tests embedded in the restricted payments, incremental facility, and permitted acquisition covenants, and the interest coverage ratio). Any gains achieved on the individual covenant levels (Sections III.A and III.B above) will be significantly undermined if the EBITDA definition itself is unduly restrictive.
III.D. Equity Cure Right (Issue No. 4 — Category B)
Agent Draft Position. The Agent Draft does not include any equity cure right mechanism. This is a notable omission for a sponsor-backed term loan B facility.
Ironhaven Recommended Position. We recommend that Ironhaven request the inclusion of a customary sponsor equity cure right, which would allow Thornfield Capital Partners, LP (acting through Thornfield Infrastructure Fund III, LP, or its affiliates) to contribute equity to Ironhaven Holdings Inc. in order to cure a financial covenant default. The contributed equity amount would be deemed to increase Consolidated Adjusted EBITDA for purposes of the Net Leverage Ratio and/or reduce Consolidated Total Net Debt for testing purposes, at the Borrower's election. The equity cure right should be subject to the following customary limitations:
(i) Frequency limit: No more than two (2) equity cures in any consecutive four-quarter testing period.
(ii) Lifetime limit: No more than five (5) equity cures during the entire life of the Facility.
(iii) Timing: The equity cure contribution must be made within ten (10) business days after delivery of the compliance certificate reflecting the financial covenant default.
(iv) No over-cure: The equity cure contribution shall be limited to the minimum amount necessary to bring the applicable financial covenant ratio into compliance (i.e., no excess contribution that would artificially depress the leverage ratio).
(v) No consecutive cures: The equity cure right may not be exercised for more than two consecutive fiscal quarters.
Rationale. Equity cure rights are standard features in sponsor-backed term loan B facilities and are particularly appropriate in the context of a construction-phase borrower where EBITDA variability is anticipated. The equity cure provides a safety valve that protects both the Borrower (by avoiding a technical default during a temporary EBITDA compression) and the Lenders (by ensuring that additional equity is injected to support the credit). Thornfield Infrastructure Fund III, LP has committed capital of approximately $4.8 billion, with substantial uncalled commitments available to support portfolio companies. The equity cure mechanism ensures that Thornfield's significant financial resources are available to backstop Ironhaven's covenant compliance during the construction phase.
Fallback. If the full Recommended Position is not achievable, we would accept a compromise of: (i) a maximum of two (2) cures per consecutive four-quarter period, and (ii) a maximum of four (4) cures during the lifetime of the Facility (rather than five). We should not compromise on the ten-business-day cure period or the no-over-cure limitation, as these are standard market terms.
III.E. Maximum Annual Capital Expenditure Covenant (Issue No. 5 — Category B)
Agent Draft Position (§7.02). The Agent Draft includes a Maximum Annual Capital Expenditure covenant of $300,000,000 per fiscal year for FY2024 and each fiscal year thereafter. No carry-forward of unused capital expenditure capacity from one fiscal year to the next is permitted.
Commitment Letter Baseline. The Commitment Letter Term Sheet provides for a Maximum Annual Capital Expenditure covenant of $325,000,000 for FY2024, stepping down to $250,000,000 for FY2025 and each fiscal year thereafter, with a 25% carry-forward of unused amounts from the immediately preceding fiscal year.
Ironhaven Recommended Position. We recommend that Ironhaven insist on the Commitment Letter terms. The structure of the Commitment Letter CapEx covenant was specifically designed to accommodate the front-loaded construction spend profile of the Virginia and Frankfurt campuses:
• The Virginia campus Phase I construction budget alone is $475 million over the construction period (approximately FY2024-Q4 2025).
• The Frankfurt campus Phase I construction budget is approximately $275 million / €253.6 million over the construction period (approximately FY2024-Q1 2026).
• Total combined construction capital expenditure for the two campuses is approximately $750 million, of which a significant portion will be incurred in FY2024 and FY2025.
• A flat $300 million annual cap with no carry-forward is unworkable given this construction profile. The $325 million FY2024 cap in the Commitment Letter, combined with the 25% carry-forward mechanism, provides the necessary flexibility to manage construction timing across fiscal years.
• The step-down to $250 million for FY2025 and beyond (with the carry-forward) appropriately reflects the reduction in capital expenditure needs as the Phase I construction is completed.
Fallback. If the stepped structure is not achievable, we would accept a flat $300 million cap for FY2024, provided that the 25% carry-forward provision is included (permitting 25% of unused annual CapEx capacity to be carried forward and used in the immediately following fiscal year). The carry-forward is essential and should not be negotiated away regardless of the annual cap level.
IV. Change of Control Definition and Mandatory Prepayment
IV.A. Change of Control — Definition (Issue No. 6 — Category A)
Agent Draft Position (§1.01 — Definitions). The Agent Draft defines "Change of Control" as the occurrence of any of the following events:
(a) Thornfield Capital Partners, LP (or its Affiliates, as defined) ceases to hold, directly or indirectly, more than 50% of the issued and outstanding voting equity interests of Ironhaven Holdings Inc.;
(b) Any "person" or "group" (within the meaning of Sections 13(d) and 14(d)(2) of the Securities Exchange Act of 1934, as amended) other than Thornfield Capital Partners, LP (or its Affiliates) acquires, directly or indirectly, more than 35% of the issued and outstanding voting equity interests of Ironhaven Holdings Inc.; or
(c) A majority of the board of directors (or equivalent governing body) of Ironhaven Holdings Inc. ceases to consist of individuals who (i) were members of such board on the Closing Date, or (ii) were nominated for election, or elected, to such board with the approval of a majority of the individuals then constituting "continuing directors" (i.e., the directors described in clause (i) or previously approved under this clause (ii)).
Commitment Letter Baseline. The Commitment Letter Term Sheet defines "Change of Control" solely by reference to clause (a) above — that is, a change of control occurs only when Thornfield Capital Partners, LP (or its Affiliates) ceases to hold, directly or indirectly, more than 50% of the voting equity of Ironhaven Holdings Inc. Neither clause (b) (the 35% minority accumulation trigger) nor clause (c) (the continuing directors clause) appears in the Commitment Letter.
Ironhaven Recommended Position. We recommend that Ironhaven reject clauses (b) and (c) of the Agent Draft definition and insist on the Commitment Letter formulation:
Clause (b) — 35% Minority Accumulation Trigger. This provision is not contained in the Commitment Letter and is inappropriate for a sponsor-backed private credit structure. Ironhaven Holdings Inc. is a privately held corporation whose equity is controlled by Thornfield Infrastructure Fund III, LP. There is no public market for the equity, and no mechanism by which a hostile third party could accumulate a 35% stake without Thornfield's knowledge and consent. The 35% trigger would, however, create operational complications in the event that Thornfield wishes to bring in a co-investor or permit management equity holders to increase their ownership stakes. So long as Thornfield maintains its >50% controlling position, the Lenders' interests are adequately protected.
Clause (c) — Continuing Directors. The continuing directors provision is a standard feature of investment-grade or public-company credit facilities, where the publicly traded nature of the equity creates the possibility that a hostile bidder could gain board control through a proxy contest without acquiring a majority of the equity. This concern is entirely inapplicable in the present context, where Ironhaven Holdings is privately held and board composition is directly controlled by Thornfield through its majority equity position. Including a continuing directors clause in a sponsor-backed private credit facility is outside market norms and could inadvertently trigger a change of control in routine board-refreshment scenarios.
Retained Provision. Clause (a), as formulated in the Commitment Letter, should be retained as the sole change of control trigger. A change of control occurs when Thornfield (or its Affiliates) ceases to hold, directly or indirectly, more than 50% of the voting equity of Ironhaven Holdings Inc. This is the standard market approach for sponsor-backed term loan B facilities, as confirmed by our review of comparable recent transactions.
Fallback. If Blackthorn & Brightmoor Systems insists on the inclusion of a secondary trigger beyond the >50% ownership test, we would accept a 35% "hostile acquiror" clause — but only if it is expressly structured to carve out the following from the definition of a triggering "person" or "group":
(i) Thornfield Capital Partners, LP and all of its affiliates and managed funds; (ii) Any co-investor investing alongside Thornfield with Thornfield's written consent; (iii) Management equity holders (including Marcus Delaine, Rebecca Sung, Jonathan Pryce-Williams, and any other officers or employees holding equity in Holdings or its subsidiaries); and (iv) Any portfolio company of Thornfield Capital Partners.
Under no circumstances should clause (c) (continuing directors) be included in the definition.
IV.B. Change of Control — Mandatory Prepayment Terms (Issue No. 7 — Category B)
Agent Draft Position (§2.11(b)). Upon a Change of Control (as defined in the Agent Draft), the Borrower is required to offer to prepay the entire outstanding principal amount of the Term Loans at a price equal to 101% of par (i.e., a 1% premium), plus all accrued and unpaid interest to the date of prepayment. The Borrower must provide 30 days' prior written notice to the Administrative Agent and the Lenders. All Lenders are entitled to exercise a "put right" requiring prepayment of their respective portions of the Term Loans.
Commitment Letter Baseline. The Commitment Letter Term Sheet provides for prepayment at 101% of outstanding principal upon a Change of Control. The Commitment Letter does not address the specific mechanics of the prepayment process or the interaction with the soft call premium.
Ironhaven Recommended Position. We recommend that Ironhaven accept the 101% prepayment premium, which is consistent with the Commitment Letter. However, we recommend that Ironhaven negotiate the following additional terms:
(a) Declining Lender Mechanism. Each Lender should have the individual right to elect to decline the prepayment and remain in the Facility under the new ownership structure (a "declining lender" concept). This is preferable to a mandatory total prepayment structure, because it preserves the possibility that some or all of the Lenders may be willing to continue under the new ownership, thereby reducing the cash outflow required upon a change of control. Under this construct, the Borrower would deliver a change-of-control prepayment notice, and each Lender would have 30 days to elect whether to accept prepayment (at 101%) or to remain in the Facility.
(b) Notice Period. The 30-day notice period proposed in the Agent Draft is acceptable.
(c) No Premium Stacking. The Agent Draft should explicitly provide that the 101% Change of Control prepayment premium is in lieu of (and not in addition to) the 101 soft call premium applicable during the first six months after closing (through December 3, 2024, under our Recommended Position on soft call). If a Change of Control were to occur during the soft call period, the Borrower should not be required to pay both a 101 change-of-control premium and a 101 soft call premium (totaling a 2% premium). The 101% Change of Control premium should be the sole premium payable, regardless of whether the event occurs during or after the soft call period.
Fallback. If the declining lender mechanism is not achievable, we would accept a mandatory total prepayment structure (all Lenders must be repaid), but we insist on the no-stacking clarification regarding the change of control premium and the soft call premium. This is a matter of simple economics — the Change of Control premium was negotiated to compensate Lenders for the credit risk of new ownership, not to be an additional windfall on top of the repricing protection premium.
V. Mandatory Prepayment — Excess Cash Flow Sweep
V.A. Excess Cash Flow Sweep Percentages and Step-Downs (Issue No. 8 — Category A)
Agent Draft Position (§2.11(c)). The Agent Draft requires the Borrower to apply the following percentages of Excess Cash Flow ("ECF") to the mandatory prepayment of the Term Loans on an annual basis:
Commitment Letter Baseline. The Commitment Letter Term Sheet provides for the following ECF sweep percentages:
Ironhaven Recommended Position. We recommend that Ironhaven insist on the Commitment Letter levels without compromise. This is a Category A issue for the following reasons:
The Agent Draft deviates from the Commitment Letter in two critical respects: (i) the top-tier sweep percentage is 100% rather than 75%, and (ii) the bottom-tier sweep percentage is 50% rather than 25%. The net effect is that, at every leverage level, the Agent Draft requires a sweep that is 25 percentage points higher than the Commitment Letter level. This is a material economic deviation that was not agreed in the Commitment Letter and has significant operational and financial implications for Ironhaven.
At a 100% top-tier sweep, Ironhaven would retain zero Excess Cash Flow for discretionary purposes during the period when its Net Leverage Ratio exceeds 4.50x — which, based on projections, will be the entirety of FY2024 and likely most of FY2025 during the construction phase. The inability to retain any Excess Cash Flow during this period would deprive Ironhaven of cash needed for: (a) discretionary capital investments beyond the CapEx covenant limits; (b) tax distributions to equity holders; (c) management fee payments to Thornfield Capital Partners; (d) reserve funding; and (e) ordinary-course working capital.
Moreover, the Commitment Letter ECF sweep percentages (75/50/25) were specifically negotiated during the Commitment Letter process. Crescent Harbor Bank and the initial lender group agreed to these levels with full knowledge of Ironhaven's projected cash flow profile and construction timeline. The Agent Draft's departure from these agreed terms is not a "papering" difference — it is a fundamental change to the deal economics.
Fallback. We recommend no fallback on this issue. The Commitment Letter terms are market standard for sponsor-backed term loan B facilities, and the 75/50/25 sweep structure was specifically negotiated and agreed. Any deviation from the Commitment Letter sweep percentages should be treated as a breach of the Commitment Letter and addressed accordingly. We note that the Commitment Letter contains a "market flex" provision, but the flex is limited to pricing adjustments (interest rate and OID) and does not extend to structural terms such as the ECF sweep percentages.
V.B. Definition of "Excess Cash Flow" (Issue No. 9 — Category B)
Agent Draft Position (§1.01 — Definitions). The Agent Draft defines "Excess Cash Flow" in a manner that limits the deductions from Consolidated Net Income to the following items:
(i) Mandatory amortization payments on the Term Loans (1.00% per annum, or approximately $2.375 million per quarter in aggregate); (ii) Voluntary prepayments of the Term Loans; and (iii) Permitted capital expenditures (i.e., capital expenditures up to the annual covenant limit of $300 million under the Agent Draft).
Notably, the Agent Draft definition does not include deductions for: (a) cash taxes actually paid; (b) capital expenditures in excess of the covenant limit (i.e., capital expenditures actually made, regardless of whether they exceed the permitted amount); (c) cash restructuring, integration, or transaction costs; (d) cash contributions to pension or benefit plans; (e) payments on capital leases or purchase money indebtedness; (f) permitted acquisitions funded with internally generated cash; or (g) a "builder basket" mechanism.
Ironhaven Recommended Position. We recommend that Ironhaven request an expansion of the deductions from Excess Cash Flow to include the following:
(a) Cash taxes actually paid (as opposed to tax expense accrued). The distinction between cash taxes paid and accrued tax expense is critical — accrual-basis tax expense may differ significantly from actual cash tax payments due to timing differences, deferred tax assets and liabilities, and the utilization of net operating loss carryforwards. The ECF calculation should reflect the actual cash consumed by tax obligations.
(b) Capital expenditures actually made — without reference to the annual covenant limit. The ECF definition should deduct all capital expenditures actually incurred during the fiscal year, regardless of whether they fall within or exceed the annual covenant limit. The purpose of the ECF sweep is to capture the Borrower's free cash flow after all operating and investment activities; artificially capping the CapEx deduction at the covenant limit would overstate Excess Cash Flow in any year in which the Borrower incurs CapEx above the limit (which may occur with Requisite Lender consent or under exception baskets).
(c) Cash restructuring, integration, and transaction costs. These are genuine cash outlays that reduce the Borrower's available cash and should be deducted from ECF.
(d) Cash contributions to pension and benefit plans. Any contributions to defined-benefit pension plans, 401(k) matching contributions, or other employee benefit plan funding obligations should be deductible.
(e) Payments on capital leases and purchase money indebtedness. Ironhaven utilizes capital leases for certain equipment at its existing data-center facilities and may enter into additional capital leases or purchase money financing arrangements. Cash payments on these obligations should reduce ECF.
(f) Permitted acquisitions funded with internally generated cash. If Ironhaven funds a permitted acquisition using cash flow rather than incremental debt or equity, the cash spent on the acquisition should be deducted from ECF. Otherwise, the ECF sweep mechanism creates a perverse incentive to fund acquisitions with debt rather than cash.
(g) Builder basket. We recommend a cumulative "builder basket" credit, commencing at $25 million on the Closing Date and growing by 50% of retained Excess Cash Flow (after giving effect to the sweep) in each subsequent fiscal year. The builder basket would provide a pool of retained earnings available for discretionary purposes, including restricted payments and investments, without requiring compliance with the financial covenants at the time of use.
Fallback. If the full suite of deductions is not achievable, we would insist at a minimum on: (i) cash taxes actually paid; and (ii) all capital expenditures actually made (without reference to the annual cap). These two items represent the most significant cash outlays that are omitted from the Agent Draft definition and their inclusion is essential to a fair ECF calculation.
V.C. Excess Cash Flow Payment Timing and De Minimis Threshold (Issue No. 10 — Category C)
Agent Draft Position (§2.11(c)). The Agent Draft requires the annual ECF sweep payment to be made within 90 days after the end of each fiscal year. No de minimis threshold is included — meaning that any positive amount of Excess Cash Flow, no matter how small, must be swept.
Ironhaven Recommended Position. We recommend that Ironhaven accept the 90-day payment timeline, which is market standard, but request the inclusion of a $5 million de minimis threshold. Under this threshold, if Excess Cash Flow for the applicable fiscal year is less than $5 million, no sweep payment would be required. The de minimis threshold avoids the administrative burden and transaction costs of processing immaterial sweep payments and is a standard feature in comparable TLB facilities.
Fallback. We would accept a de minimis threshold of $2.5 million as a compromise. While this is below our preferred level, it still provides meaningful relief for immaterial ECF amounts.
VI. Voluntary Prepayment and Repricing Protection
VI.A. Soft Call Protection Period and Premium (Issue No. 11 — Category B)
Agent Draft Position (§2.06(a)). The Agent Draft includes a soft call (repricing protection) provision requiring a premium of 102 (2% of the principal amount prepaid) for any voluntary prepayment or repricing transaction occurring within twelve (12) months after the Closing Date (i.e., through June 3, 2025). After the twelve-month soft call period, voluntary prepayments may be made at par.
Commitment Letter Baseline. The Commitment Letter Term Sheet provides for a soft call premium of 101 (1% of the principal amount prepaid) for voluntary prepayments or repricing transactions occurring within six (6) months after the Closing Date (i.e., through December 3, 2024). After the six-month soft call period, voluntary prepayments may be made at par.
Ironhaven Recommended Position. We recommend that Ironhaven insist on the Commitment Letter terms: a 101 premium for 6 months after closing. The Agent Draft's 102 premium for 12 months represents a significant economic deviation from the agreed terms:
Economic Impact. On the full $950 million Facility, a 2% soft call premium would amount to approximately $19 million — a substantial cost that significantly constrains Ironhaven's ability to refinance the Facility during the first year if market conditions become favorable. By contrast, the 1% premium under the Commitment Letter would be approximately $9.5 million. The additional $9.5 million in premium represents an unwarranted transfer of value from the Borrower to the Lenders that was not part of the agreed transaction economics.
Market Context. In the current interest rate environment, with the market expectation for Federal Reserve rate cuts in the second half of 2024 and into 2025, the ability to refinance or reprice the Facility within the first year is a valuable option for the Borrower. A 12-month lockout at a 2% premium significantly diminishes this optionality. The 6-month / 101 structure in the Commitment Letter reflects the market standard for data-center TLBs and appropriately balances the Lenders' interest in initial yield stability against the Borrower's interest in refinancing flexibility.
Fallback. We recommend accepting the Commitment Letter terms of 101 for 6 months without further compromise. This is a clearly negotiated Commitment Letter provision, and any departure should be resisted.
VI.B. MFN Sunset (Issue No. 12 — Category C)
Agent Draft Position (§2.14). The Agent Draft includes a "most favored nation" ("MFN") provision applicable to incremental facilities. Under this provision, any incremental term loan facility incurred under the Credit Agreement must be priced within 50 basis points of the existing Term Loan B (i.e., SOFR/EURIBOR + 425 bps); if the incremental facility is priced more than 50 bps above the existing facility, the interest rate on the existing facility must be increased to within 50 bps of the incremental facility rate. Critically, the MFN provision in the Agent Draft is perpetual — it does not sunset and remains in effect for the entire life of the Facility.
Ironhaven Recommended Position. We recommend that Ironhaven request either: (a) a 12-month sunset on the MFN provision (such that the MFN applies only to incremental facilities incurred through June 3, 2025, and thereafter Ironhaven may incur incremental facilities at any pricing without triggering a rate step-up on the existing Facility), or alternatively (b) an increase in the MFN threshold from 50 basis points to 75 basis points if the sunset is not achievable. A perpetual MFN effectively constrains the Borrower's incremental financing flexibility for the entire seven-year term of the Facility, which is inappropriate given that the credit profile and market conditions will evolve materially over that period.
Fallback. We would accept an 18-month sunset (through December 3, 2025) with the 50 basis point threshold as a compromise.
VII. Representations and Warranties
VII.A. OFAC and Sanctions Representations (Issue No. 13 — Category A)
Agent Draft Position (§5.18). The Agent Draft contains detailed representations and warranties regarding compliance with sanctions administered by the Office of Foreign Assets Control of the U.S. Department of the Treasury ("OFAC"), the European Union, and the United Kingdom. Specifically, the Agent Draft requires each Loan Party to represent that:
(a) No Loan Party, nor any director, officer, employee, or, to the knowledge of the Borrower, agent or affiliate thereof, is a "Sanctioned Person" (defined broadly to include any person listed on any sanctions list maintained by OFAC, the EU, or UK OFSI, or any person organized in, located in, or resident in a Sanctioned Country);
(b) No Loan Party has engaged in any transaction, investment, or dealing with any Sanctioned Person or in any Sanctioned Country;
(c) No part of the proceeds of the Term Loans will be used, directly or indirectly, in violation of any sanctions; and
(d) Each Loan Party shall provide quarterly compliance certificates confirming ongoing compliance with applicable sanctions.
Ironhaven Recommended Position. We acknowledge that robust sanctions representations are appropriate in this transaction given its cross-border nature. The Facility involves a German co-borrower (IHDC Frankfurt), the EUR sub-tranche is denominated in euros and will be funded through the European syndicate, and the lender group includes NordBrücke AG (a German bank) and Osprey Insurance Capital, Ltd. (a Bermuda-domiciled entity with a UK-person nexus). Accordingly, compliance with OFAC, EU, and UK sanctions regimes is not merely a legal formality but a genuine business concern.
That said, we recommend the following modifications to the sanctions representations to bring them within market norms:
(a) Knowledge qualifier for agents and affiliates. The "to the knowledge of the Borrower" qualifier currently applies only to agents and affiliates. This qualifier should be retained without modification. We do not object to the absence of a knowledge qualifier for directors, officers, and employees of the Loan Parties, as the Borrower is expected to have direct knowledge of such persons' sanctions status.
(b) Controlled affiliates only. The reference to "affiliates" in the sanctions representation should be limited to controlled affiliates of the Loan Parties only. As currently drafted, the representation could be read to extend to the broader Thornfield Capital Partners portfolio, including portfolio companies in which Ironhaven has no involvement or visibility. This is inappropriate — Ironhaven cannot reasonably represent as to the sanctions compliance of unrelated Thornfield portfolio companies over which it exercises no control.
(c) Defined sanctions lists. The definition of "Sanctioned Person" should be limited to persons listed on specifically identified sanctions lists: (i) the OFAC Specially Designated Nationals and blocked Persons List (SDN List) and the Sectoral Sanctions Identifications List (SSI List); (ii) the EU Consolidated Sanctions List maintained pursuant to Council Regulation (EU) No. 269/2014; and (iii) the UK Office of Financial Sanctions Implementation (OFSI) Consolidated List. The Agent Draft's reference to broader, undefined "sanctions" creates uncertainty as to scope and could inadvertently capture regulatory frameworks that are not intended to be within the scope of these representations.
(d) Inadvertent violation cure. Include a carve-out for inadvertent or de minimis violations of sanctions laws that are promptly cured upon discovery. This carve-out is increasingly standard in cross-border credit facilities and reflects the practical reality that inadvertent technical violations can occur in complex multinational organizations.
Ongoing Compliance Workstream. We note that Calderwood & Harkness's sanctions and KYC compliance workstream, led by Daniel Obi, is ongoing. OFAC and EU screening of all Lender parties and all Borrower-side entities and individuals is expected to be completed by mid-April 2024. No issues have been identified to date, and we do not anticipate any concerns.
Fallback. We would accept the sanctions representations substantially as drafted in the Agent Draft if: (i) the knowledge qualifier for agents and affiliates is retained; and (ii) the scope of "affiliates" is limited to controlled affiliates of the Loan Parties. The defined-sanctions-lists limitation and the inadvertent-cure carve-out are strongly preferred but may be conceded if necessary.
VII.B. Material Adverse Effect Qualifier (Issue No. 14 — Category B)
Agent Draft Position (Various Sections). The Agent Draft contains a comprehensive set of representations and warranties in Sections 5.01 through 5.20. We have identified that several representations that customarily carry a Material Adverse Effect ("MAE") qualification are drafted on an unqualified, absolute basis in the Agent Draft. Specifically, the following representations lack an MAE qualifier:
• Accuracy of financial statements and financial information (§5.04)
• Absence of undisclosed litigation or proceedings (§5.06)
• Environmental compliance (§5.09)
• Tax compliance (§5.10)
• Labor matters (§5.11)
• ERISA compliance (§5.12)
• Accuracy of projections and forecasts (§5.13)
• Compliance with all laws and regulations (§5.15)
Without an MAE qualifier, these representations create strict liability for any inaccuracy, no matter how immaterial. A minor, inconsequential tax underpayment or an immaterial environmental condition could, on a strict reading, constitute a breach of representation, triggering an Event of Default and potentially accelerating the Facility.
Ironhaven Recommended Position. We recommend that Ironhaven request the insertion of an MAE qualifier into all representations where such qualification is market standard. The MAE qualifier should provide that the representation is accurate except where the failure to be accurate would not reasonably be expected to have a Material Adverse Effect. The definition of "Material Adverse Effect" should be customary (i.e., a material adverse effect on the business, assets, operations, or financial condition of the Borrower and its subsidiaries, taken as a whole, or on the ability of the Loan Parties to perform their obligations under the Loan Documents).
Fallback. We would accept unqualified (absolute) representations for the following limited categories, which are customarily not MAE-qualified:
(i) Corporate existence and good standing; (ii) Corporate power and authority to enter into the Loan Documents; (iii) Due authorization, execution, and delivery of the Loan Documents; (iv) No conflict with organizational documents; and (v) OFAC/sanctions compliance (as discussed in Section VII.A).
All other representations should carry MAE qualifiers.
VII.C. Accuracy of CIM / Information Warranty (Issue No. 15 — Category B)
Agent Draft Position (§5.21). The Agent Draft includes a broad information warranty in which the Borrower represents that all information contained in the Confidential Information Memorandum (the "CIM") dated March 14, 2024 (prepared by Crescent Harbor Bank, N.A., in its capacity as Lead Arranger, and reviewed by Calderwood & Harkness for legal accuracy) is accurate in all material respects and contains no material misstatement or omission of a material fact necessary to make the statements therein not misleading.
Ironhaven Recommended Position. We recommend the following modifications to the information warranty:
(a) Limitation to Borrower-furnished information. The representation should be limited to information that was furnished in writing by or on behalf of the Borrower to Crescent Harbor Bank for inclusion in the CIM. The CIM was prepared by Crescent Harbor Bank as Lead Arranger, and significant portions of the CIM (including market commentary, industry analysis, and comparable transaction summaries) were prepared by the bank's syndication team without input from the Borrower. It would be inappropriate to hold the Borrower responsible for the accuracy of information prepared by the Lead Arranger.
(b) Projections and forward-looking statements exclusion. Projections, estimates, budgets, forecasts, and other forward-looking statements contained in the CIM should be expressly excluded from the "no misstatement or omission" representation. Instead, the Borrower should represent only that such projections and forward-looking information were prepared in good faith based on assumptions that were reasonable at the time furnished. By their nature, projections are inherently uncertain and subject to significant revision as actual results unfold; a flat "no misstatement" representation for forward-looking information would expose the Borrower to liability for any deviation from projections, which is an inappropriate allocation of risk.
(c) Express acknowledgment of CIM preparation responsibility. The information warranty should include an express acknowledgment that the CIM was prepared by Crescent Harbor Bank, N.A., in its capacity as Lead Arranger, and that the Borrower's responsibility is limited to the factual information it specifically provided for inclusion in the CIM.
Fallback. We would accept a broader information warranty (covering the CIM as a whole, rather than only Borrower-furnished information) if the projections and forward-looking statements exclusion is included. This is a critical protection — the distinction between factual representations and forward-looking projections is fundamental to the proper allocation of risk in a credit facility.
VIII. Covenants — Affirmative and Negative
VIII.A. Restricted Payments / Distributions (Issue No. 16 — Category A)
Agent Draft Position (§7.06). The Agent Draft prohibits all Restricted Payments (defined to include dividends, distributions, management fees, sponsor distributions, and any other payments on account of equity interests) unless:
(a) The Net Leverage Ratio, on a pro forma basis after giving effect to the Restricted Payment, is at or below 4.00 to 1.00; and
(b) The aggregate amount of all Restricted Payments in any fiscal year does not exceed $15,000,000.