The Space Between

Why you should read this. A sponsor needs $100 million. The senior lenders will not advance it. Selling the company would crystallize a disappointing return, and the fund’s investors want cash. A capital-solutions fund offers a note that pays no cash interest for seven years. Another offers preferred equity. Both appear to buy time. The price of that time is harder to see. The note compounds. The preferred may create a withholding obligation before it produces a dollar of cash. The issuer’s interest deduction may be deferred, limited, or partly denied. A fund-level borrowing can change the tax result again. This alert follows one hypothetical company from financing to exit and ends with three decisions a reader should be able to make: which term creates cash tax, which deduction may be unusable, and which investor cannot hold the instrument efficiently. Read it before the term sheet arrives. The term sheet will assume you already know.

Two things to hold onto while reading. First, the tax bite is not a fixed charge: a variety of structuring strategies can reduce it, and the difference between a well-drafted and a hastily drafted instrument is measured in millions, not basis points. Second, the tax consequences to an investor in a capital-solutions fund differ by the kind of investor it is (U.S. taxable, non-U.S., or tax-exempt), and techniques exist to optimize the result for each. This alert names the problems. The techniques are a shorter conversation with a narrower audience.

The three answers, in numbers.

Cash tax now. Option B, if Parent has earnings and profits. A year-one PIK distribution of $11 million, 40% non-U.S., no treaty, is $1.32 million of withholding before a dollar of cash; a full gross-up is about $1.89 million that year, and the LLC has to be permitted to send it up (Exhibit 11).

Unusable deduction. Option A. About 16% of the discount never comes off; the rest waits for cash. Parent is already $12.6 million over the §163(j) cap on the unitranche alone, so the $49 million released in 2030 competes with that year’s cash interest for a 30% cap (Exhibit 9).

Holder who cannot hold it cheaply. The non-U.S. partner in Westbrook, on the preferred. Separately, the tax-exempt partner, if a later NAV draw is acquisition indebtedness of the book (Exhibit 10).

And the good news. Each of the three is negotiable. Structure, drafting, and the choice of holder can reduce the bite considerably; what cannot be done is to discover it after signing.

This is the space between senior debt and common equity. Mayer Brown’s London finance team published a clear primer on it this week, defining “capital solutions” as “bespoke financing that occupies the space between traditional senior secured debt and common equity” and observing that “what has changed is not the existence of the gap, but its width.”¹ We agree, and we credit the source. Our subject is what the article leaves for another day: what the instruments in that space cost after tax, to whom, and when.

Hybrid capital can solve a timing problem. It can also conceal a solvency problem, or move tax cost from the company to its investors. The right structure is the one that funds a credible business plan, preserves realistic repayment capacity, and produces an acceptable after-tax outcome for each investor who has to hold it. That is our thesis, and it cuts both ways: the same instrument that rescues one company will, in another, let a distribution arrive before the loss becomes visible. And because a capital-solutions fund is itself owned by U.S. taxable, non-U.S., and tax-exempt investors, the after-tax outcome is three answers rather than one. Good structuring narrows the spread among them. Indifferent structuring widens it.

Our question is therefore practical. Which instrument solves the business problem at an acceptable cost to the company and its investors? We follow one company, Northline Services, from financing to exit, then identify the terms that should be settled before a term sheet becomes a commitment. The market context is compressed to what the decision needs. The full tax mechanics sit in a technical appendix at the back.

Executive summary

The gap is real, and its uses are not all equal. Buyout funds hold roughly 32,000 companies worth $3.8 trillion that they have not sold; distributions ran at 14% of net asset value in 2025, a fourth year under 15% (Bain, February 2026). Demand for capital between senior debt and equity persists across the cycle. Its price, volume, and bargaining power do not (Section 02).

The note and the preferred both deliver $98 million; neither is free. On Northline’s facts, a 12% PIK holdco note plus warrants returns Westbrook 13.7% gross in the recovery case, and the sponsor 1.49x. The identical operating case with no new financing leaves the sponsor $480 million at exit instead of $313 million at exit plus $98 million now, an aggregate nominal financing cost of about $69 million before timing (Section 03 and Exhibit 6).

Tax character is decided by terms, not by labels. A seven-year 12% PIK note with warrants, issued by a corporation, meets all three conditions of §163(i): about 16% of its discount is never deductible and the remainder is deductible only when paid. A PIK preferred distribution can be a dividend to a foreign holder, withheld on before any cash exists, depending on earnings and profits and the holder’s status. The same drafting choice moves the deduction, the withholding, and the unrelated business income at once (Section 04).

The reader’s roadmap. Section 01 places the instruments. Section 02 gives the market in one page. Section 03 works Northline, including the downside. Section 04 takes the three seats: issuer, fund, investor. Section 05 lists the terms that change the answer. Section 06 says when we would decline. Section 07 is the signing table. The appendix carries the schedules and authorities.

01 Where the capital sits

A company’s money comes in layers, and the layers are paid back in order. Senior secured debt has the first claim on the assets; common equity is paid last and keeps whatever is left. Everything in between is priced by three dials: when you are paid, what you may do if you are not, and how much of the upside you keep. A fourth question belongs beside them, and it is the one this alert is about: who supplies cash when the tax on the instrument becomes payable?

Loan-to-value here means cumulative debt or exposure relative to enterprise value: a lender at 50% is covered twice over by the worth of the business. Attachment and detachment are where a given layer’s exposure begins and ends. The instruments that live in the band are listed below, in words a board member who is not a financier could use. Two distinctions matter more than the names. The first is where in the structure the money goes in: at the operating company (where the cash is), at a holding company above it (outside the operating company’s collateral, and subject to whatever the senior agreement says about it), or at the fund that owns the holding company (where the investors are). The second is whether the instrument is a loan (a creditor with remedies) or equity (a shareholder with rights by contract). The market word “hybrid” describes an economic blend; for tax, as Section 04 shows, the instrument is one or the other.

Exhibit 2. A short glossary. Drafted by La Presa Partners for this alert. Where an instrument’s remedies or tax treatment are described as deal-specific, they are.

02 Why the demand persists

Three facts explain why sponsors keep reaching for the middle of the structure. First, the exit backlog. Bain’s 2026 report counts about 32,000 unsold buyout-owned companies worth $3.8 trillion, an average hold at exit near seven years, and buyout distributions at 14% of net asset value in 2025, under 15% for the fourth straight year; and, in a Private Equity International survey Bain reports, 53% of investors said undrawn prior commitments constrain their ability to make new ones.² Second, the cost of cash-pay debt. Lincoln International’s first-quarter 2026 data on sponsor-backed companies show average leverage of 5.1 times EBITDA, a simplified fixed-charge coverage ratio of 1.3x, and 19.5% of companies below 1.0x, with PIK at 8.9% of interest dollars; by count, 55.7% of PIK investments carried no PIK at close.³ Third, the refinancing calendar. Amend-and-extend activity cleared $106 billion of maturities in the first half of 2026 and left only $32 billion of U.S. leveraged loans due through 2027, and a large share of leveraged loans and high-yield bonds, on the order of $1 trillion by PitchBook LCD’s count, matures in 2027 through 2029, with the weakest credits concentrated in 2028 and 2029.⁴

Rates are the variable everyone watches and the least reliable of the four. The Federal Reserve raised the funds rate to 3.75–4.00% on September 16, 2026, by a 12–0 vote, and its September projections show 4.1% at the end of 2026 and 2027.⁵ That is a dated fact, and it keeps coverage where it is for now. It is not a forecast; forecasting the Fed is a hobby, not a method. The gap would narrow with stronger cash generation, cheaper senior debt, a reopened sale market, fresh common equity at a lower mark, or sponsors accepting losses. Several of those are cyclical. What is durable is that flexible capital has uses across a cycle: growth capital when senior lenders are full, bridge capital when an exit is delayed, and rescue capital when a structure built for lower rates needs to be rebuilt. Its price, its bargaining power, and its volume will change with the cycle; its existence will not.

Supply has scaled to meet it. Blackstone’s fifth Capital Opportunities Fund closed in April 2026 with more than $10 billion of investable capital, Apollo’s Hybrid Value Fund III at about $6.5 billion of commitments in May, and 17Capital’s NAV lending fund at about $7.5 billion including affiliated mandates in March.⁶ Two 2026 deals show the range: QXO’s $3 billion of 4.75% convertible perpetual preferred, acquisition currency for a listed buyer, and Bayer’s €3 billion minority-equity capital solution from Apollo and KKR, with Bayer keeping control.⁷ That is what the decision needs from the market; the rest is color.

Demand for the space between is structural in its uses and cyclical in its price. A reader who takes only one thing from Section 02 should take this: for a suitable borrower hybrid capital is available at scale, so the question is no longer whether it can be raised, but whether a given company can repay it and a given investor can hold it.

03 Northline Services: the choice

Numbers argue better than adjectives, so here is a company. The parties are invented, the figures are rounded, and the case is a template rather than advice. The entities are stated because they decide the tax result.

The entities and the base case. Harbor Fund IV, L.P. is a $1.5 billion 2019-vintage buyout fund, a partnership whose investors include U.S. taxable individuals, U.S. tax-exempt institutions, and non-U.S. investors. It owns 100% of Northline Parent, Inc., a Delaware C corporation, which owns 100% of Northline Services LLC, a wholly owned limited liability company that is disregarded for federal tax purposes. Parent is therefore the taxpayer for the whole operating group, and the LLC’s income, interest, and deductions are Parent’s. (If the operating entity were a partnership or an elected corporation, the §163(j) and characterization analysis below would change; we model the disregarded case because it is the cleanest teaching case and the most common.) Harbor is a fund, not the common parent of a consolidated group; there is no consolidated return above Parent.

Swipe sideways for the other columns.

Exhibit 4. Northline Services, inputs. Hypothetical. Coverage is cash flow before interest divided by cash interest; it omits mandatory amortization ($5 million a year, assumed) and other fixed charges. DPI is cumulative distributions divided by paid-in capital, here assumed at $1.5 billion; it is not a distribution yield on NAV.

Northline is a good business in a capital structure built for a different rate. EBITDA has grown 10%; cash interest has grown 64%. Harbor wants to distribute about $100 million to Fund IV’s investors, which would lift DPI from 0.40x to about 0.47x. There are four ways to do it, and three of them have already said no. A dividend recapitalization at the LLC is blocked by the unitranche’s 6.0x incurrence cap. A continuation vehicle would price the company off a mark Harbor does not accept; on Fund IV’s facts the LPA requires an advisory-committee vote for an affiliate sale, and a new vehicle would carry a second promote. A NAV facility at the fund would require advisory-committee consent under Fund IV’s LPA, which follows ILPA’s July 2024 guidance for distribution-funded facilities.⁸ That leaves the holding company. Harbor causes Parent to seek $100 million, and Westbrook Capital Solutions Fund III, a hybrid fund whose investors are 40% non-U.S., offers two term sheets.

The cash path at closing is short: Westbrook advances $100 million to Parent; Parent uses $2 million for assumed total fees, including any lender consent fee, and distributes $98 million to Harbor. No transfer from the LLC is needed for that. LLC cash matters later, for any gross-up, protective payment, or repayment, and each of those requires a permitted transfer from the LLC to Parent under the unitranche agreement. Whether Parent may borrow at all, grant liens, or issue paper structurally subordinated to the operating-company debt, and whether a mandatory-redemption preferred falls within the agreement’s definition of disqualified equity, are questions the actual agreement answers. In the case we assume the unitranche agreement permits the structure and the relevant distributions.

Swipe sideways for the other columns.

Exhibit 5. Two term sheets for the same $100 million. Hypothetical. The options are alternatives; Option B does not sit behind a note that was never issued. Coupons are chosen from 2025–26 disclosed prints (Calderys 11.75%/12.50% PIK toggle; People Corporation 10–11% talk) and practitioner descriptions of private preferred.

Four years later: the financing comparison. It is 2030. In the recovery case Northline’s EBITDA has reached $70 million, the unitranche has amortized to $290 million, and a strategic buyer pays 11 times, or $770 million. The table shows both options and, for comparison, the same operating case with no new financing at all.

Swipe sideways for the other columns.

Exhibit 6. The same company, four years on. Hypothetical. Accretion at 12% and 11% compounded annually; warrant proceeds are 3% of residual common on a fully diluted basis; Westbrook’s returns are unlevered security returns on single cash flows, before fund fees, carry, and investor tax. Harbor’s cash flows occur on different dates and are shown in aggregate dollars, not as an IRR.

Three things the table says. First, the recovery is driven by EBITDA growth, a higher multiple, and senior paydown. The financing did not create it. Second, Option A reduces Harbor’s exit proceeds by $167 million ($480 million less $313 million); after crediting the $98 million received in 2026, the aggregate nominal financing cost is about $69 million, which is not automatically an economic loss: whether the early cash was worth it depends on what Fund IV’s investors did with it and on their discount rate, and the comparison must be made on a common valuation date. On a common date the cost is smaller than $69 million looks: discounting the 2030 proceeds to 2026 at an illustrative 12%, the no-financing case is worth about $305 million today and Option A about $199 million plus the $98 million in hand, or $297 million, so the financing costs roughly $8 million of present value at that rate, and less at a higher one. Third, the preferred is cheaper for Harbor by $15.2 million before tax. Section 04 finishes that comparison, and finishes it with a verb.

The downside, and the compounding. A term sheet that only works in the recovery case is not yet a term sheet. The table below shows what happens to the same instruments if the 2030 exit is at eight times $55 million (the company stands still) or eight times $45 million (it goes backwards). It also shows why Westbrook’s claim is not protected by $275 million of sponsor equity: at closing, $550 million of enterprise value less $310 million of senior debt and the $100 million hybrid leaves about $140 million of residual common before fees. Harbor’s $275 million is a historical cost. It protects no one.

Swipe sideways for the other columns.

Exhibit 7. The same note in three worlds. Hypothetical. Immediate realization at the stated enterprise value; no insolvency costs, no further senior claims, no tax, no interim payments. Illustrative sensitivities, not forecasts. In the lower-value case the lender recovers its capital but not its full contractual claim; in the stress case it loses principal.

In the lower-value case the company is worth $440 million against $447 million of senior debt and accreted note: Westbrook is nearly whole, Harbor’s common is worth nothing, and the $98 million paid out in 2026 is all Fund IV’s investors will ever see from Northline. That is the case in which hybrid capital let a distribution arrive before a loss became visible. Whether the parties accept that exposure is a question of their agreed risk tolerance, and Section 06 says where we would draw the line. Either way, the time to know which case you are in is 2026. A 12% PIK balance rises from $100 million to $157.4 million in four years and $221.1 million in seven; EBITDA in the recovery case grows about 6.2% a year. At $290 million of senior debt, enterprise value of about $447 million is needed in 2030 to satisfy senior debt and the note before costs (Exhibit 8).

04 Three perspectives: issuer, fund, investor

Hybrid instruments are hybrid for the market and, as a rule, one thing or the other for tax: each component of a unit is classified as debt or as equity, and the discipline consists of deciding which on day one, writing every term to match, and keeping it that way through the amendment and the exit. What follows takes Northline’s two options through three seats. Each seat has its own techniques for softening the result, and most of them work only if applied before the term sheet is signed; we name the problems here and leave the techniques to the engagement. The full schedules and authorities are in the appendix; a general counsel or CFO should be able to use this section without opening it.⁹

The issuer: is the coupon deductible, and can Parent use it? Option A is debt if its terms and substance say so. There is no single bright-line test for an unrelated-party financing; §385 supplies the statutory framework and factors, and contemplates that an instrument can be part debt and part stock. The label is not the answer, and it is not irrelevant either. In Hewlett-Packard Co. v. Commissioner the Ninth Circuit listed the names given to the certificates among its factors, then weighed them against the substance. A fixed return, a put to a third party, and “highly predictable income” pointed to debt despite the “preferred stock” label. Other circuits keep their own lists, and there is no universal test.¹⁰ What the factors reward is a sum certain at a fixed date, creditor remedies, a realistic expectation of repayment from the issuer’s own cash flows, and holders who are not also the shareholders. Northline’s note has a seven-year maturity, accelerates on default, is held by an unrelated fund, and is expected to be repaid from a sale or refinancing that the recovery case supports; the warrants are a separately classified equity component of the unit. That supports debt treatment of the note at issuance. It does not guarantee it: an instrument that is debt when signed is judged on its terms and the facts at signing, and a later decline in the business does not by itself convert it to equity, but a note whose repayment was never plausible is a different instrument from the start. Rating-agency equity credit and accounting classification are separate analyses and answer separate questions.

Option A’s PIK interest is original issue discount, and the note is an AHYDO. A PIK payment is not a payment; the accrued interest is discount. The discount accrues under the constant-yield method, but the AHYDO rules permanently deny part of the deduction and defer the balance until payment. Section 163(i) defines an applicable high yield discount obligation as a corporate note that (A) matures in more than five years, (B) yields at least the applicable federal rate plus five points, and (C) carries “significant” discount, meaning that at an accrual period ending after the fifth anniversary the accrued and unpaid discount exceeds one year’s yield on the issue price. The base-case Option A note contains no mandatory cash-payment provision, and it meets all three conditions. Because the $100 million is sold with $3 million of warrants, the note’s issue price is $97 million and its yield to maturity is 12.49%, not 12%. With the September 2026 mid-term AFR of 4.49%, the disqualified yield is 12.49% less 10.49% (AFR plus six points), and the disqualified fraction is 16.0% of every accrual of discount.¹¹ That slice is permanently nondeductible. The remainder is deductible only when paid in cash.

Swipe sideways for the other columns.

Exhibit 9. Northline’s Option A note under §163(i), no cash-payment provision. Tax balances are adjusted issue price; the contractual balance at the year-four exit is $157.4 million. Hypothetical. Issue price $97 million after the $3 million warrant allocation; yield 12.49%; annual accrual periods ending on issue-date anniversaries. AFR 4.49%. At the year-four retirement the contractual balance is $157.4 million against an adjusted issue price of $155.3 million: $58.3 million of discount has accrued, $9.3 million of it permanently disallowed and $49.0 million released on payment, and the $2.0 million paid above adjusted issue price is repurchase premium, generally deductible as interest under Treas. Reg. §1.163-7(c) subject to the same limitations, for a potential $51.0 million before §163(j).

The market’s answer is a protected note: a mandatory cash-payment provision that prevents the note from having significant discount. Its timing and amount must be tested against the note’s actual accrual periods and all scheduled payments. Section 163(i)(2) looks at accrual periods ending after the fifth anniversary; with annual periods ending on anniversaries, the first such period ends at year six, when the unpaid discount on the unchanged schedule would be about $99.5 million, exceeding the $12.1 million allowance by about $87.4 million (at year five the unpaid discount is $77.7 million and the excess about $65.6 million). A contract can require an earlier protective payment, but it must say so, and the payment changes the balance, the later PIK, and the tax yield, so the protected note needs its own schedule and its own yield computation. It also needs a demonstrated source of permitted cash: Parent has none of its own, so the provision is only as good as the LLC’s ability to upstream that amount under the unitranche in 2031 or 2032. In one line: a payment large enough to keep unpaid discount under one year’s yield at the first period ending after the fifth anniversary, sourced from a permitted upstream, on a recomputed yield. The uncured seven-year note modeled in Exhibit 9 contains no such provision. The other ways out are a maturity of five years or less, which changes the credit, or an issuer that is not a corporation, which does not escape the rule: a partnership issuer applies the AHYDO deduction rules at the level of its corporate partners, and a disregarded issuer’s federal tax owner is the taxpayer.¹²

Whether Parent can use the deduction is a separate question, and the answer is mostly not yet. The OBBBA restored the EBITDA-based §163(j) cap for taxable years beginning after December 31, 2024, and the IRS has since issued Rev. Proc. 2026-17 on post-OBBBA elections and Fact Sheet FS-2026-14 on the amended limitation; what remains outstanding is a conforming amendment to Treas. Reg. §1.163(j)-1(b)(1), whose text still carries the 2022 sunset.¹³ We assume Parent’s adjusted taxable income (tax EBITDA, which is not book EBITDA) is $52 million in 2026, so the cap is $15.6 million. Parent’s business interest is the LLC’s $28.2 million of cash interest, all of it Parent’s because the LLC is disregarded. The cap already binds by $12.6 million before the holdco note exists. Current-year interest is applied first and prior disallowed interest carries forward behind it, so the note’s $49.0 million of deferred discount, when it becomes deductible in 2030, competes with 2030’s cash interest for a cap of 30% of that year’s ATI (assumed $66 million, or $19.8 million). Much of it will carry forward again. The case assumes a sale of Parent stock with no deemed asset-sale election and retirement of the note from the sale proceeds at closing, so Parent keeps its business and its taxpayer identity; the buyer changes, the carryforward stays with Parent, and its value depends on post-acquisition adjusted taxable income and debt levels, on §382 if the sale is an ownership change, and on what the buyer agrees to pay for tax attributes. What can be said on these facts is that the deduction is delayed to the year of payment and that its use at or after closing is uncertain. On these facts the sponsor is buying optionality, not a scheduled shield, and Exhibit 9 should not be booked as a credit. This is the point we would make to Harbor’s deal team: the deduction is the sponsor’s argument for the note, and it is an argument about a contingent future benefit. The withholding, below, is the fund’s argument, and it is about cash due now.

Option B produces no deduction, and stock-settlement terms can eliminate Option A’s. Preferred equity is equity; there is no interest to deduct. Separately, §163(l) denies any deduction on corporate debt where a substantial amount of principal or interest is required to be paid or converted into issuer equity, is payable in or convertible into equity at the issuer’s option, is measured by the value of that equity, or is part of an arrangement reasonably expected to produce one of those results. A holder’s conversion right counts only if exercise is substantially certain. Northline’s note pays PIK in additional notes, not in shares, and its warrants are a separate instrument; it is outside §163(l). A PIK payable in stock, or a note with a mandatory conversion at maturity, would not be, and the deduction would be zero from the first day.

The fund: what Westbrook is doing, and with whose money. Westbrook holds Option A as a lender and Option B as a shareholder, and the difference governs its investors. As a lender, its interest income is ordinary; the discount accrues into income each year whether or not cash arrives, so Westbrook’s taxable U.S. investors pay tax on income they have not received for four years (exempt investors and foreign investors holding qualifying portfolio interest do not). A corporate investor in Westbrook may treat the dividend-equivalent portion of the disqualified 16% slice as a dividend for purposes of the dividends-received deduction. As a preferred holder, Westbrook’s PIK distributions are governed by §305: dividends to the extent of Parent’s current and accumulated earnings and profits, otherwise a return of basis. Section 03 assumed $20 million of E&P before the 2026 distribution. For the withholding illustration that follows we adopt a separate sensitivity, full E&P coverage of each year’s PIK distribution, because it is the case in which the cash problem is largest; a supported annual E&P forecast would refine both the character and the withholding, and an E&P estimate is not by itself a license to withhold less.

Westbrook’s own activity also matters. A fund that originates loans directly to U.S. borrowers, negotiates terms, and takes structuring or commitment fees can be engaged in a U.S. trade or business, and if it is, its non-U.S. investors have effectively connected income and the fund must withhold under §1446. The own-account trading safe harbor of §864(b)(2)(A)(ii) is unavailable to a dealer, and the Tax Court in YA Global Investments held that a Cayman fund investing through a U.S. manager in convertible instruments of U.S. companies crossed that line.¹⁴ Whether Westbrook is on the safe side depends on how it originates, what its manager does, and what fees it takes, questions that require their own memorandum. The usual mitigants (an originating vehicle that seasons and sells, treaty-based structures, corporate blockers for non-U.S. and tax-exempt investors) each carry conditions and costs of their own. None is automatic.

The investor: after-tax return and who funds the tax. The instrument Northline issues determines how each of Westbrook’s investors is taxed, and in one case determines who has to write a check. A coupon multiplied by one minus a tax rate is not this transaction’s return: the note’s tax yield is 12.49%, not 12%; the warrants have their own basis and character; the annual tax must be funded while the security balance compounds; and the corporate holder’s result depends on the dividend-equivalent slice of the disqualified discount.

Swipe sideways for the other columns.

Exhibit 10. Same money, five holders, two drafts: the treatment. Hypothetical. Treatment stated at the level of the rule; each row depends on the holder’s facts, Westbrook’s own activities and borrowing, and the instrument’s final terms.

The non-U.S. rows are the reason Harbor’s deal team should care which instrument Parent issues. Portfolio interest on Option A is exempt from U.S. withholding if the core conditions are met: the note is in registered form, the holder gives the required beneficial-owner statement, the holder is not a 10-percent shareholder (tested at the partner level through Westbrook, with attribution, and counting voting power for a corporate issuer, so Westbrook’s warrants and any voting rights matter), the interest is not contingent on Northline’s receipts, income, property values, or distributions, and the bank-loan and related-CFC exclusions do not apply.¹⁵ A cash sweep that accelerates payment of a fixed amount is a timing provision; a coupon whose amount moves with cash flow is contingent interest, and only the contingent portion loses the exemption. Mere accrual of discount does not by itself trigger withholding; the payment and redemption rules, and an anti-avoidance rule for sales arranged to avoid them, determine when it is collected.¹⁶ On Option B, by contrast, the $11 million year-one PIK distribution is governed by §305. On the full-E&P sensitivity, with 40% attributable to non-U.S. investors subject to 30% and no treaty reduction, the withholding is $1.32 million, and it is due before a dollar of cash exists.

Exhibit 11. Two ways to fund withholding on a PIK dividend. Hypothetical. The regulation requires the withholding agent on an in-kind distribution to liquidate property or obtain the tax from another source (Treas. Reg. §1.1441-3(e)(1)); which agent, and whose source, is a documentation question the term sheet must answer.

Does the gross-up make the note cheaper? Not on these numbers. If Parent funds the full gross-up each year on the compounding preferred, the cash cost is $1.89 million in year one, then $2.09 million, $2.32 million, and $2.58 million, $8.88 million in all; carried to the 2030 exit at an illustrative 12%, about $10.46 million. That narrows the preferred’s $15.2 million gross advantage to roughly $4.8 million, before other differences, on assumptions that leave the 11% entitlement unchanged, treat both the PIK and the gross-up as dividends, and charge the cash drain to Harbor. On these terms, preferred still leaves Harbor more residual after a modeled gross-up. We would still take the note if the preferred cannot be sold at 11% to a book that is 40% non-U.S., or if the unitranche will not let the gross-up cash leave the LLC. Placement and cash access decide it. The tax rate does not.

The tax-exempt investor and the NAV facility. One more thing can change the investor’s answer after closing: Westbrook’s own borrowing. Interest and dividends are excluded from a tax-exempt investor’s unrelated business taxable income, but income from debt-financed property is brought back in proportion to the property’s acquisition indebtedness. If Westbrook draws $200 million on a NAV facility and the proceeds are traced to a new $200 million note, that note is debt-financed property and its discount is subject to the debt-financed-income fraction for the year: average acquisition indebtedness over average adjusted basis, which the accruing discount itself raises (on a full year at par, roughly 94%, or about $22.6 million of a $24 million accrual before connected deductions). If the proceeds fund distributions or follow-ons and the facility is found to be acquisition indebtedness of the portfolio as a whole, under the statute’s “but for” and “reasonably foreseeable” tests, then a share of all of Westbrook’s interest, dividends, and gains becomes taxable to its exempt investors, on a ratio of average debt to average adjusted basis that is computed, not assumed, and that uses a highest-indebtedness rule for gains on disposition.¹⁷ Whether a facility drawn years after the investments were made, and not described in the offering documents, is acquisition indebtedness is a question of causal connection, timing, and purpose. A reference in the private placement memorandum is evidence one way; its absence is not proof the other.

05 Terms that change the answer

Six drafting choices, each of which moves at least one of the three seats. Each is stated with Northline’s number where there is one.

Swipe sideways for the other columns.

Exhibit 12. Six terms, three seats. Authorities in the appendix. The amendment row is stated conditionally: the payment-deferral safe harbor in Treas. Reg. §1.1001-3(e)(3) covers a specific kind of change with conditions, and other changes in the same amendment can independently be significant.

The amendment row deserves a sentence more, because if the 2028 maturity calendar holds, a good deal of today’s hybrid paper will be amended before it is repaid. A significant modification is a deemed exchange of the old note for a new one. Whether the exchange produces cancellation-of-debt income to Parent depends on the new note’s issue price under the applicable rules, which turn on whether either instrument is publicly traded (with a small-issue exception measured on outstanding stated principal at the time, which Northline’s accreting note will have exceeded by 2028) and, if not, on the stated principal and adequate stated interest. Any COD is income to Parent, subject to the insolvency and other exclusions of §108 and their attribute reduction; whether Westbrook has a recognized loss, and its character, depends on Westbrook’s basis, which is its cost plus accrued discount, and on whether the exchange is a reorganization. None of that is a reason to avoid amending. All of it is a reason to model the amendment before the maturity forces it.

The characterization rule. Decide what the instrument is for tax on day one, write every term to match, and record the repayment case that supports it. An instrument that is equity for the rating agency and debt for tax is common and lawful. An instrument that is debt for tax and never had a plausible path to repayment is an invitation to recharacterize, and the invitation is accepted at the worst possible time.

06 When we would decline

Independent judgment means being able to say no, and no is cheaper before signing. We would advise against a hybrid financing, or require it to be restructured, in four situations.

No credible operating and exit case supports repayment. Exhibit 8 is the test. An accepted downside, one the parties have priced and can absorb, is a different thing from the absence of a repayment case. We would decline if no credible operating and exit case supports repayment of the accreted claim on top of the senior debt, or if the downside exposure exceeds the parties’ agreed risk tolerance. In the first situation the instrument is not financing a recovery; it is transferring the equity to the noteholder in slow motion, and the sponsor should sell now or bring fresh common equity.

The proceeds fund a distribution without a credible repayment plan. A holdco note whose only purpose is a distribution, on a company whose cash flow cannot service the senior debt and whose exit depends on multiple expansion, is a distribution before the loss becomes visible. The investors who receive the $98 million may be glad of it. The investors who are still in the fund four years later will not be.

Cash cannot reach the issuer. A catch-up clause, a withholding gross-up, or a mandatory redemption that the operating company’s credit agreement will not let it fund is not a term; it is a default with a date. Every cash obligation of the holdco instrument should be traced to a permitted payment under the senior documents before signing.

Tax and gross-up leakage make the security uneconomic. If the only investors who can hold the instrument efficiently are not the investors the fund has, the price will move after signing or the deal will not close. Exhibit 10 should be run on the actual investor base before the term sheet is issued, not after.

07 Before signature

Six decisions, each with an owner and the evidence that closes it.

Swipe sideways for the other columns.

Exhibit 13. The signing table. La Presa Partners.

The financing decision is not finished when the company has the cash. It is finished when the parties know how the instrument can be repaid, which deductions can actually be used, and who funds any tax due before repayment.

What we would watch next

Treasury’s conforming amendment to Treas. Reg. §1.163(j)-1 after Rev. Proc. 2026-17 and FS-2026-14; the regulation’s text still carries the 2022 sunset.

The appellate posture of YA Global, checked against the docket on the date any origination-style fund relies on it.

The Federal Reserve’s October and December meetings, which decide whether the coverage arithmetic in Section 02 holds through 2027.

The 2028 refinancing calendar, and the first cohort of 2024-vintage holdco PIK notes passing their fifth anniversaries in 2029, with the first statutory AHYDO testing period ending after that anniversary (in 2030, under the annual-anniversary convention used here).

FASB ASU 2026-01 on measuring PIK dividends at the stated rate, effective for annual periods beginning after December 15, 2026, and applicable to in-scope equity-classified preferred, including existing instruments depending on transition.

Our take

The space between senior debt and common equity is a permanent feature of the capital structure, and its width in 2026 is a fact about interest rates, hold periods, and lender caution that no one should mistake for a law of nature. Flexible capital will be used across the cycle, for growth, for bridges, and for rescues, and it will be priced according to which of those it is. The instruments are not new. What is new is their scale, the sophistication of the buyers, and the number of investors, on both sides, who hold them through funds whose tax profiles are not their own.

For tax the point is simple and less comfortable. These instruments are hybrid for the market and, component by component, one thing or the other for the Code. The same term (a redemption date, a PIK payable in shares, a warrant in the unit, a coupon that moves with cash flow, a facility drawn at the fund) can move the issuer’s deduction, the foreign investor’s withholding, and the exempt investor’s taxable income at once. On Northline’s facts the deduction the sponsor is buying is delayed to the year of payment and uncertain in use after closing, because the §163(j) cap already binds; the withholding the fund is avoiding is cash due before any cash exists. On Northline’s terms, preferred leaves Harbor more residual even after the modeled gross-up; we would take the note only if the preferred cannot be placed at 11% with a 40% non-U.S. book or the unitranche will not release the gross-up cash. Placement and cash access decide it, and that is where the negotiation should sit. Hybrid capital can buy valuable time for a business with a credible path forward. It can also let a distribution arrive before the loss becomes visible. The term sheet should make that distinction easier to see, and a financing that works for the issuer but fails for its investors is not yet a solution.

The market is asking what capital solutions is, and the question has been well answered. The question we would put to a client on either side of the term sheet is what the instrument will be in 2028 when it is amended and in 2030 when it is repaid, and whether the answer will still be the one everyone agreed on the day it was signed. Function over form is a sound principle for structuring a deal. It is also the principle the examiner brings.

Technical appendix

Schedules, assumptions, and authorities supporting Sections 03 through 05. A reader who needs the conclusion does not need this; a reader who needs to check it does.

A. AHYDO computation, Option A note. Face $100,000,000; issue price $97,000,000 after allocation of $3,000,000 to warrants by relative fair market value under §1273(c)(2) and Treas. Reg. §1.1273-2(h); the issuer’s allocation binds holders absent timely disclosure. Stated redemption price at maturity $221,068,000 ($100m × 1.12⁷). Yield to maturity (annual compounding) 12.4884%. Mid-term AFR, September 2026, annual: 4.49% (Rev. Rul. 2026-17, Table 1). Test (A): seven years > five years. Test (B): 12.49% ≥ 4.49% + 5% = 9.49%. Test (C): at the close of the first accrual period ending after the fifth anniversary (year six under the annual-anniversary convention), accrued OID is $99.5 million, interest paid is nil, and issue price times yield is $12.1 million; $99.5m > $12.1m. Disqualified yield: 12.4884% − (4.49% + 6%) = 1.9984%; disqualified portion: 1.9984 ÷ 12.4884 = 16.0%. Total OID over seven years $124,068,000; permanently disallowed approximately $19,854,000. At year five, for comparison, unpaid OID is $77.7 million against the same $12.1 million allowance. In the four-year exit case, OID accrued is $58.3 million, disallowed $9.3 million, and $49.0 million is released on payment; the $2.04 million paid above the $155.3 million adjusted issue price is repurchase premium under Treas. Reg. §1.163-7(c), generally deductible as interest subject to applicable limitations, for a potential $51.0 million before §163(j). The holder’s character on the premium is analyzed separately. A protected note with a mandatory cash-payment provision requires its own schedule and yield. Authorities: §163(e)(5), §163(i); Treas. Reg. §§1.1272-1, 1.1273-2, 1.1275-2(c)(3) (PIK not a payment); Treas. Reg. §1.701-2(f), Example 1 (corporate partners of a partnership issuer).

B. Section 163(j) assumptions. Parent is the sole taxpayer for Parent and the disregarded LLC (Treas. Reg. §301.7701-2(c)(2)(i)). Adjusted taxable income for taxable years beginning after December 31, 2024 adds back depreciation, amortization, and depletion (§163(j)(8)(A)(v), as amended by Pub. L. No. 119-21, §70303(a)); for taxable years beginning after December 31, 2025, subpart F, GILTI, and §78 amounts are excluded (§163(j)(8)(A)(vi), added by §70342) and the limitation applies before capitalization (§163(j)(10), added by §70341). ATI is not book EBITDA; Northline’s 2026 ATI of $52 million and 2030 ATI of $66 million are assumptions. Ordering: current-year business interest is taken into account before disallowed business interest carryforwards (Treas. Reg. §1.163(j)-5(b)). Guidance: Rev. Proc. 2026-17 (elections after the OBBBA); IRS Fact Sheet FS-2026-14 (August 2026). The regulation at Treas. Reg. §1.163(j)-1(b)(1)(i)(D)–(F) had not been conformed as of the date of this alert. The exit is a sale of Parent stock with no §338 or §336(e) election; Parent survives with its business, and the note is retired from sale proceeds at closing. §382 applies to Parent’s carryforwards only if the sale is an ownership change (§382(g)), with disallowed business interest treated as a pre-change loss under §382(d)(3); the value of the carryforward to a buyer depends on post-acquisition ATI, debt levels, and the negotiated attribute economics.

C. Debt-equity characterization. No single bright-line test; §385(a)–(b) supplies the statutory framework and factors and contemplates part-debt, part-stock treatment; the §385 regulations survive only as to related-party distribution and funding transactions (Treas. Reg. §§1.385-3, -4; T.D. 9880 removed the documentation rules in 2019). Multi-factor analysis by circuit: Fin Hay Realty Co. v. United States, 398 F.2d 694 (3d Cir. 1968); Estate of Mixon v. United States, 464 F.2d 394 (5th Cir. 1972); Roth Steel Tube Co. v. Commissioner, 800 F.2d 625 (6th Cir. 1986); Hewlett-Packard Co. v. Commissioner, 875 F.3d 494 (9th Cir. 2017) (names given to the certificates are among the factors, at pp. 11–12; put to ABN AMRO, at p. 6 and n.3). Notice 94-47, 1994-19 I.R.B. 9 (instruments designed as debt for tax and equity for regulatory, rating, or accounting purposes will be scrutinized; terms of 50 years or more examined closely). Treas. Reg. §1.1275-1(d) (OID rules do not determine indebtedness). A modification that changes debt into non-debt is a significant modification (Treas. Reg. §1.1001-3(e)(5)); in testing whether a modified instrument remains debt, deterioration in the obligor’s financial condition is not taken into account (§1.1001-3(f)(7)).

D. Corporate preferred. Distributions with respect to preferred are taxable under §305(b)(4); redemption premiums and accreting liquidation preferences are constructive distributions under §305(c) and Treas. Reg. §1.305-5(b), accrued on constant-yield principles where redemption is mandatory or at the holder’s option (unless subject to a remote contingency) or, for issuer calls, more likely than not to occur. Dividend treatment to the extent of E&P (§§301, 316); the valuation of distributed shares for tax is a separate question from the accounting measurement under ASU 2026-01. In-kind distribution withholding: Treas. Reg. §1.1441-3(e)(1). Individual qualified-dividend holding period: more than 60 days in the 121-day window, or more than 90 days in the 181-day window for preferred dividends attributable to periods exceeding 366 days (§1(h)(11)(B)(iii), cross-referencing §246(c)). Corporate dividends-received deduction: §243 (50%; 65% for 20%-or-more owners); holding period §246(c) (more than 45 days in a 91-day window; more than 90 days in a 181-day window for preference dividends attributable to periods exceeding 366 days), reduced for diminished risk of loss (Treas. Reg. §1.246-5); §246A (debt-financed portfolio stock); §1059 (extraordinary dividends; 5% threshold for preferred; basis reduction if held two years or less). §306 stock: preferred received as a nontaxable stock dividend, in a reorganization, or in a §351 exchange in the circumstances the section specifies. Nonqualified preferred stock: §351(g)(2) (boot in a §351 exchange); §354(a)(2)(C) (reorganization exchanges). Affiliation: §1504(a)(4) excludes non-voting, non-participating, non-convertible preferred with no unreasonable redemption or liquidation premium from the 80% test; the practical consequence for Harbor is nil, since a fund is not a common parent, but it matters where a corporate sponsor consolidates the portfolio company.

E. Partnership issuers. Where the issuer is a partnership, a fixed preferred return determined without regard to income is a guaranteed payment for the use of capital under §707(c) and Treas. Reg. §1.707-1(c); the 2015 proposed regulations (REG-115452-14) would treat the entire minimum as a guaranteed payment and remain proposed. The 2020 final §163(j) regulations removed guaranteed payments for the use of capital from the automatic definition of interest, but the anti-avoidance rule of Treas. Reg. §1.163(j)-1(b)(22)(iv) can treat them as interest (T.D. 9905). Classification and deductibility require separate analysis; a §707(c) payment may also be subject to capitalization. Holding a bona fide note of a partnership does not make the creditor a partner; holding preferred equity in a partnership whose assets are an operating business gives a tax-exempt holder a share of operating income under §512(c) and a non-U.S. holder effectively connected income with withholding under §1446(a), and on a transfer, withholding under §1446(f) generally at 10% of the amount realized, which can include the transferor’s share of liabilities, subject to the regulations’ exceptions and certifications.

F. Non-U.S. holders. Portfolio interest: §871(h), §881(c); registered form; beneficial-owner statement; not a 10-percent shareholder (§871(h)(3); partner-level test with attribution under Treas. Reg. §1.871-14(g); voting power for corporate issuers, capital or profits for partnership issuers); not contingent interest (§871(h)(4)); bank-loan and related-CFC exclusions (§881(c)(3)); not effectively connected. Withholding timing on OID: Treas. Reg. §1.1441-2(a)(6), (b)(3), (e). U.S. trade or business: §864(b)(2)(A) (dealer exclusion); Treas. Reg. §1.864-4(c)(5) (banking, financing, or similar business); YA Global Investments, LP v. Commissioner, 161 T.C. No. 11 (2023). FIRPTA: Treas. Reg. §1.897-1(d) distinguishes an interest solely as a creditor from an interest other than solely as a creditor; a participating loan is a USRPI in its entirety under §1.897-1(d)(1), but a separately issued warrant or conversion right is tested on its own under §1.897-1(d)(3) subject to the aggregation and anti-avoidance rules of §1.897-1(d)(4); for an interest in a corporation the rules apply if the corporation is a U.S. real property holding corporation under §897(c)(2) at any time during the §897(c)(1)(A)(ii) testing period, subject to the statutory exceptions; direct real-property interests and interests in other entities require their own analysis. §267A reaches related-party hybrid arrangements and, under Treas. Reg. §1.267A-2(f), structured arrangements between unrelated parties as defined in §1.267A-5(a)(20); imported mismatches (§1.267A-4) and §245A(e) hybrid dividends require separate analysis.

G. Tax-exempt holders and fund-level debt. Unrelated debt-financed income: §512(b)(4), §514(a)(1) (average acquisition indebtedness over average adjusted basis), §514(b)(1), §514(c)(1) (“but for” and “reasonably foreseeable”), §514(c)(7) (highest indebtedness in the preceding 12 months for gains on disposition); directly connected deductions, including interest, are allowed in the same proportion. §4968 endowment excise tax (as amended by Pub. L. No. 119-21, §70415: 1.4%, 4%, and 8% tiers for taxable years beginning after December 31, 2025, for institutions meeting the student-adjusted endowment and enrollment tests) uses rules similar to §4940(c), which exclude amounts taken into account under §511, so the same income is not taxed under both regimes. Partnership liabilities: nonrecourse liabilities are allocated under the three tiers of Treas. Reg. §1.752-3(a) (minimum gain, §704(c) minimum gain, then excess nonrecourse by profits or another permitted method); a partner guarantee shifts the liability only to the extent of the recognized economic risk of loss under Treas. Reg. §1.752-2, including the bottom-dollar rules. Distributions of borrowed money are tax-free to basis under §731(a)(1); disguised-sale analysis under §707(a)(2)(B) and Treas. Reg. §1.707-5 applies where a partner contributed property and receives a debt-financed distribution. Debt-financed income is computed on average acquisition indebtedness over average adjusted basis for the year (Treas. Reg. §1.514(a)-1(a)(2)–(3)), so a fully debt-funded purchase does not fix the fraction at 100% once basis accrues or debt is repaid. Interest tracing for individual partners follows use, not collateral: Treas. Reg. §1.163-8T and Notice 89-35 (allocation of debt-financed distributions by the recipients’ use), preserved by T.D. 9943’s preamble.

H. Modifications. Treas. Reg. §1.1001-3(e)(2) (yield change greater of 25 bp or 5% of annual yield); §1.1001-3(e)(3) (deferral of scheduled payments within the lesser of five years or 50% of the original term, provided payments are unconditionally payable by the end of the safe-harbor period; other changes are tested separately); §1.1001-3(e)(4) (change in obligor or security: for recourse debt a change in security or priority is significant only if it results in a change in payment expectations, and the obligor rules carry their own exceptions); §1.1001-3(e)(5) (change in nature). Issue price of the new instrument: §1.1273-2(b)–(f) (publicly traded property; the small-issue exception applies where outstanding stated principal does not exceed $100 million at the determination time); §1274 where not publicly traded. COD: §108(e)(10) (debt-for-debt), §108(e)(8) (debt-for-equity, including partnership interests), §108(d)(6) (partnership-level determination, partner-level exclusions), §108(b) attribute reduction. Northline’s note will have accreted above $100 million of stated principal by 2028 and must be tested on its actual adjusted issue price and holder basis.

Selected research and primary authorities

1. Mayer Brown, “What Is Capital Solutions? Bridging the Gap Between Debt and Equity,” Michael Fiddy, Electra Callan, Mark Evans, and Christopher Street (September 2026).
2. Bain & Company, Global Private Equity Report 2026 (February 23, 2026), and Private Equity Midyear Report 2026 (June 8, 2026).
3. Lincoln International, Valuations & Opinions Group, Private Market Perspectives, Q1 2026.
4. PitchBook LCD, “2026 US distressed credit outlook: bifurcation, maturity wall promise busy year” (December 2025); “Leveraged loan issuers lean in to amend-and-extend deals” (July 2026). Fitch Ratings, global leveraged finance maturity profile (January 2026).
5. Federal Reserve, FOMC statement and Summary of Economic Projections (September 16, 2026); FEDS Note, “Private Credit and Leveraged Loan Markets: Similarities, Differences, and Substitution” (August 11, 2026). Federal Reserve Bank of New York, SOFR (September 21, 2026).
6. Blackstone, Capital Opportunities Fund V closing release (April 7, 2026). Apollo Global Management, Hybrid Value Fund III close (May 5, 2026); Bayer capital solution closing release (September 16, 2026); Hybrid Value strategy page. PitchBook, “17Capital’s $7.5B NAV fund” (March 2026). QXO, Form 8-K Exhibit 99.1 (January 12, 2026).
7. ILPA, Guidance on NAV-Based Facilities (July 2024). FASB, Accounting Standards Update 2026-01.
8. Internal Revenue Code §§1(h)(11), 108, 163(e)(5), (i), (j), (l), 243, 246, 246A, 267A, 301, 305, 306, 316, 351(g), 354, 382, 385, 512, 514, 707, 731, 752, 864(b), 871(h), 881(c), 897, 1001, 1059, 1273, 1274, 1441, 1446, 1504(a)(4), 4940(c), 4968. Treasury Regulations §§1.163-7, 1.163-8T, 1.163(j)-1, 1.163(j)-5, 1.246-5, 1.267A-2, 1.267A-4, 1.267A-5, 1.305-5, 1.385-3, 1.514(a)-1, 1.701-2, 1.707-1, 1.707-5, 1.752-2, 1.752-3, 1.864-4, 1.871-14, 1.897-1, 1.1001-3, 1.1272-1, 1.1273-2, 1.1275-1, 1.1275-2, 1.1441-2, 1.1441-3, 301.7701-2. Pub. L. No. 119-21 §§70303, 70341, 70342, 70415. T.D. 9880; T.D. 9905; T.D. 9943; REG-115452-14. Rev. Proc. 2026-17, 2026-15 I.R.B.; IRS Fact Sheet FS-2026-14 (August 2026); Rev. Rul. 2026-17 (September 2026 AFRs); Notice 94-47, 1994-19 I.R.B. 9; Notice 89-35.
9. Hewlett-Packard Co. v. Commissioner, 875 F.3d 494 (9th Cir. 2017); Fin Hay Realty Co. v. United States, 398 F.2d 694 (3d Cir. 1968); Estate of Mixon v. United States, 464 F.2d 394 (5th Cir. 1972); Roth Steel Tube Co. v. Commissioner, 800 F.2d 625 (6th Cir. 1986); YA Global Investments, LP v. Commissioner, 161 T.C. No. 11 (2023).

Important notes

This alert addresses U.S. federal income tax only. State and local, non-U.S., and treaty consequences differ and are noted only where they change the analysis. Northline Services, Harbor Fund IV, Westbrook Capital Solutions, and their figures are hypothetical and rounded; they illustrate mechanics and are not a model of any transaction. Market figures are quoted with their publisher, date, and basis; where publishers differ, we name each rather than average them. The appellate status of any case cited should be confirmed on the date of reliance.

This client alert is for general information only. It is not tax, legal, investment, or accounting advice. Consult your own advisers about your particular circumstances before acting on any matter discussed here. Questions: Carlos Schmidt, La Presa Partners LLC, carlos@lapresallc.com, (917) 558-6393.

Download the alert (PDF, 361 KB)

This alert is for general information only and is not tax, legal, investment, or accounting advice. Consult your own advisers before acting. Carlos A. Schmidt, MBT, MBA, CPA · Managing Member · carlos@lapresallc.com · (917) 558-6393.