Investing in Listed BDCs

On November 5, 2025, Blue Owl announced that its listed BDC, Blue Owl Capital Corporation, would absorb its non-traded sibling, Blue Owl Capital Corporation II. The non-traded fund sold its shares at net asset value; the listed fund traded that week at roughly 20% below its own. The terms, as the Financial Times reported on November 16, would have handed OBDC II’s investors shares worth about 0.80 times what their statements said, with no chance to redeem first.

Fourteen days later, on November 19, the two boards terminated the deal, citing market conditions. By February 2026 the non-traded fund had ended its quarterly tender offers, the windows in which a non-traded fund buys back a slice of its own shares, and was selling loans, about $600 million of them at 99.7% of par, to raise cash.

Nothing about the loans changed in those fourteen days. What changed was which price got to be the truth. Two vehicles run by the same manager, holding the same kind of senior loans, carried two prices at once: one written by the manager every quarter, one by the market every second. The gap was 20 cents on the dollar, and the market’s number was the one the manager had to use when it wanted to combine the two.

That gap is the subject of this guide. A listed BDC is the one part of the private-credit market you can buy from a brokerage account for the price of a share and sell tomorrow at 9:31, and the one part where you can see, every day, what strangers with money at risk think the loan book is worth. The wider market, its managers, the redemption queues at the perpetual funds and the case for the asset class as a whole are the ground of our hub flagship, Investing in Private Credit; this guide goes deeper on the listed vehicles alone. Three guides on the same hub take strategies where no daily price exists at all — Investing in Litigation Finance, Investing in Music Royalties and Investing in Pre-IPO Shares — and each of them is worth reading against this one, because the discount a listed BDC prints every day is the number those three markets do not have.

Four statutory rules give a BDC its leverage, its payout and its tax bill, and most of what follows in this guide is a consequence of them. Congress created the business development company in 1980, through the Small Business Investment Incentive Act, as an amendment to the Investment Company Act of 1940: a publicly traded closed-end fund that would lend to small and mid-sized American companies the banks and the bond market did not serve. For twenty-five years it stayed a niche; the banks’ retreat from leveraged lending after 2008 made it the public face of a trillion-dollar market.

The first rule is the 70% test under Section 55 of the 1940 Act: at least 70% of assets must be “qualifying assets,” in practice securities of private US companies or of public companies under $250 million of market capitalisation, plus cash and Treasuries. The other 30% can be anything, which is where CLO equity (the first-loss slice of a loan securitisation), joint ventures and foreign loans sit.

The second is the leverage cap under Section 61. Until 2018 a BDC needed 200% asset coverage, one dollar of debt per dollar of equity; the Small Business Credit Availability Act, signed March 23, 2018, allowed 150% coverage, two dollars of debt per dollar of equity, after a shareholder vote or a board approval and a one-year wait, with ongoing disclosure. Nearly every large BDC took it; in 2026 most ran 1.0 to 1.3 turns of debt to equity, and the gap between 0.7 turns and 1.2 is worth about a point of dividend yield on the waterfall in the record section.

The third rule is tax. A BDC elects to be a regulated investment company under Subchapter M of the Internal Revenue Code: no corporate tax provided it distributes at least 90% of its investment company taxable income, and a 4% excise tax on what it holds back past the calendar-year threshold. The income arrives untaxed at the fund and fully taxed in your hands. The fourth rule is disclosure. A BDC files a 10-K and 10-Qs like any operating company, and its schedule of investments lists every loan by borrower with cost, fair value, rate and non-accrual status, meaning whether the fund has stopped booking interest on it because collection is in doubt; no other private-credit wrapper shows you that much.

The universe

The Small Business Investor Alliance’s most recent census, on Q4 2024 data, counted 50 publicly traded BDCs with $159 billion of assets, 47 non-traded BDCs with $205 billion and 59 private BDCs with $69 billion, about $433 billion in all.

The listed share has fallen since Blackstone launched the perpetual non-traded model in January 2021, but the listed 50 are the ones with a price, and the largest, Ares Capital at more than $25 billion of assets, is bigger than most of the banks it competes with. The VanEck BDC Income ETF, ticker BIZD, holds the listed universe by market capitalisation and had about $1.6 billion of net assets in the summer of 2026 ($1.57 billion on September 2, per Investing.com); it is what our tape tracks.

Every listed BDC carries two prices, and the gap between them, the discount to NAV, is the number the rest of this guide returns to. Net asset value per share is the fund’s own estimate: the fair value of every loan, set quarterly by the manager and its board with a valuation firm’s help, minus the fund’s debt, divided by shares. For loans that never trade, fair value is a model of spreads, borrower leverage and coverage, plus a judgment about whether principal will come back; it moves slowly by design, and a manager paid on assets has no reason to hurry a markdown.

The share price is set by everyone else, institutional credit funds, BDC-specialist hedge funds, the retail investor who owns the stock for its yield and the ETFs that hold the sector by weight, and it embeds their view of three things the NAV does not: how much of the marked value will be collected, how long the dividend can be maintained, and what the manager’s fee will take on the way through.

Below NAV, the market is saying one of the three is worse than the filing implies; above NAV, as Main Street Capital has traded for most of its listed life, it is paying for rising NAV per share and a fee structure that leaves more income in the fund.

The spread of premiums and discounts at one moment is therefore a ranking. On the dates shown, the market paid 70–75% over book for Main Street, about 15% over for Sixth Street Specialty Lending, about 3% under for Ares Capital, roughly 20% under for Blue Owl Capital Corporation and about a third under for FS KKR, which was buying back its own shares at an average of $10.73 against an $18.30 NAV with $300 million of sponsor support behind it. Sector-wide, the median listed BDC traded at about 0.74 times forward NAV in late March 2026, a 26% discount and the widest since October 2020, per Octus.

Two mechanics of the discount matter to a buyer. A BDC below NAV cannot issue new shares without shareholder approval, because selling stock at 90 cents to buy loans worth a dollar dilutes existing holders; a fund below book stops growing, and its manager’s fee stops growing with it, which is the strongest incentive a manager has to close the gap through buybacks.

A persistent discount invites the cold open’s arbitrage in reverse: a listed fund at 80 cents is the cheaper claim on the same loans than a non-traded sibling redeeming at 100, and the historical remedy, merging the non-traded fund into the listed one, lands at the listed price. If you own the listed share, that catalyst works for you.

Invest Alternative’s own price series for the listed sector shows what daily pricing did to a book of loans whose reported values barely moved. Our private-credit series is the daily close of the VanEck BDC Income ETF, collected from Yahoo Finance since August 30, 2021, with the live file holding the last 500 sessions, from $16.17 on September 9, 2024 to $13.33 on September 4, 2026.

It is a price series, so it understates total return by BIZD’s quarterly distributions, which the next section adds back. Our Private Credit sub-index, built on the same series and rebased to 100 on September 2, 2025, stood at 83.157 on September 8, 2026, down 16.79% over one year, at a 9% weight in our composite.

The series is three separate falls, each with a cause. From the February 19, 2025 high of $17.82 the sector lost 21.7% in seven weeks to $13.96 on April 8, as the tariff announcements priced a recession into floating-rate borrowers; the worst session in the file, April 4, 2025, was −7.4%, and the best, April 9, was +7.3%, one trading day apart.

From $16.03 on September 2, 2025 it lost 15.8% in five weeks to $13.49 on October 10, the weeks of the Tricolor and First Brands bankruptcies. From $14.18 at the end of 2025 it ground down 14.6% to $12.11 on April 1, 2026, the quarter in which the non-traded funds’ redemption queues and the software markdowns arrived together. Peak to trough, 32.0%. Annualised daily volatility across the window is about 20%, which is equity volatility for an asset sold as a substitute for bonds.

The funds inside the tape reported something else over the same months. Ares Capital’s NAV per share fell $0.24 in Q2 2026, Blue Owl’s $0.15, Golub’s $0.10, while Main Street’s rose and Hercules’s rose 2.1%. A sector whose own books moved two or three percent traded through a 32% drawdown. One of those is the real volatility of a levered loan portfolio, and it is not the one in the quarterly letter.

The number on the quote screen and the number you keep are different numbers, and confusing them is the most common mistake made with this asset class. A 10% dividend yield is the yield on the loans, multiplied by leverage and inflated by the discount; the return you earn is that income minus fees, interest and credit losses, plus or minus whatever the discount does while you hold.

Start with the loans. The Cliffwater Direct Lending Index, which measures the loans themselves, unlevered and before any fund fee, across the BDC universe’s own SEC filings, returned 9.5% a year over the twenty years to 2025, with one negative year, 2008, and an average annual credit loss of 1.01% for 2004–2024, per its March 31, 2026 release. Whether that is adequate for the risk is contested in the literature the flagship covers; the narrower point is that a BDC shareholder does not own the index. They own it after a fund’s costs, on borrowed money, at a market price.

The arithmetic, for $100 of your equity in a typical externally managed fund. Borrow $100 at 6.0% and hold $200 of loans yielding 10.5%: gross income $21.00. Interest costs $6.00. A base fee of 1.5% on gross assets takes $3.00; operating costs at 0.5% of assets take $1.00. Pre-incentive income is $11.00, which clears a 7% hurdle with catch-up (above the hurdle the manager takes all the income until its share reaches 20% of the whole), so a 20% incentive fee takes $2.20.

Net investment income, NII from here on, is $8.80, the number the dividend is paid from. Subtract the index’s long-run credit loss, 1% of assets or $2.00 on $200 of loans, and the total return on NAV is 6.8%. At a 2% loss rate, the kind of year 2008 was, it is 4.8%. Price the shares at 85 cents on the NAV dollar and the $8.80 dividend reads as a 10.4% yield, which is what the screen shows, on a book earning 6.8% before the discount moves.

Our tape says the same thing in prices. BIZD’s 12-month distribution yield was 13.59% on July 31, 2026, per VanEck, about $1.81 a share on a $13.33 price. Add two years of that, roughly $3.60, to the $2.84 price fall from $16.17 to $13.33 over our window, and the estimated total return over two years is about +4.8%, or 2.4% a year, against a 37% rise in the S&P 500’s monthly average level between September 2024 and August 2026 on the Shiller series we hold.

Over the single year to September 4, 2026 the estimate is about −5.7%: the price fell 16.9% and the distributions covered two-thirds of it. It is an estimate, but a 13% yield bought a negative year.

The reason to hold the asset class anyway is the long record of the loans and the fact that the discount mean-reverts: it widened past 20% in 2008, in early 2016 and in March 2020 and narrowed each time within a year or two. The mistake is to buy the yield. The correct object is the total return on NAV, high single digits through a cycle for a well-run fund, bought at a discount that gives you a margin for the losses the marks have not admitted yet.

Seven names on the same quarter show the range across the sector. Figures are as of June 30, 2026 unless noted and will be stale when you read them, which is why the next section teaches you to refresh them. Between them the seven have every kind of manager: two internal (Main Street, Hercules), five external, one of which needed its sponsor’s balance sheet in 2026.

Ares Capital (ARCC)

The largest listed BDC ended Q2 2026 with NAV of $19.35, down $0.24; core earnings of $0.47 and GAAP net investment income of $0.50 against a $0.48 dividend; and non-accruals of 2.4% of cost and 1.4% of fair value, up from 2.1% at cost and, in management’s words, below long-term averages. The stock traded at $18.76 on August 7, about 3% below NAV. Its software book, more than $6.6 billion across 135-plus borrowers on the Founderpath tally as of March 2026, is the sector’s largest in dollars and among the smallest as a share of the portfolio.

Blue Owl Capital Corporation (OBDC)

Blue Owl reported NAV of $14.26, down from $14.41, “primarily reflecting markdowns on a small number of names,” with non-accruals of 2.8% of cost and 0.8% of fair value, up from 2.0% and 1.0%; adjusted earnings of $0.34 covered $0.33 of dividends including a $0.02 supplemental. It traded about 20% below NAV on September 9, 2026, the widest of the seven after FS KKR, and the manager said in 2026 it would reduce software exposure; the discount is the market’s memory of the 0.80× merger as much as a view on the loans.

FS KKR Capital (FSK)

FS KKR is the stressed one. NAV fell 2.8% to $18.30 on valuation losses; non-accruals were 7.1% of cost and 3.8% of fair value, down from 8.1% and 4.2% but three times Ares’s. NII of $0.44, an annualised 9.6% on NAV, covered a $0.44 Q3 distribution; the payout had been cut twice between 2025 and early 2026 per PitchBook, the second time from $0.48 to $0.42 with the Q1 2026 results.

In the same quarter a KKR subsidiary bought $150 million of stock by tender and $150 million of new convertible preferred, and the fund began a $300 million buyback, purchasing about 3.7 million shares for $40 million at an average of $10.73, some 40% below NAV, by August 5; on September 9, 2026 the stock stood about 33% below NAV per Seeking Alpha, the widest discount of the seven. Management’s stated goal is “a smaller, higher-quality fund.”

Golub Capital BDC (GBDC)

Golub reported fiscal Q3 2026 NAV of $14.25, down $0.10; adjusted NII of $0.34, a 9.5% return on equity; non-accruals of 2.9% of cost and 1.9% of fair value across twenty positions; 87% of the book in its top two internal ratings; and debt-to-equity of 1.23×. Golub charges the lowest external fee of the five, which is why its NII competes at a lower portfolio yield.

Main Street Capital (MAIN)

Internally managed, Main Street is the premium name. Preliminary Q2 2026 NAV was $33.88 to $33.96, up $0.42 to $0.50 on portfolio appreciation and accretive issuance, the sale of new shares above NAV, which lifts NAV per share for everyone already in; distributable NII of $1.02 to $1.06; non-accruals of 1.1% of fair value and 4.0% of cost, the gap being a lower-middle-market book with equity stakes that gets marked hard when it goes wrong. The market paid 70–75% over book for it in 2026, down from higher levels without any change in the numbers; at that premium every dollar buys 58 cents of loans, and the rest is the manager and the record.

Hercules Capital (HTGC)

Also internally managed, Hercules lends to venture-backed technology and life-science companies, a different risk from buyouts. Q2 2026 was a record on income, $149.1 million total and $92.9 million net, covering the base distribution by 125%; NAV rose 2.1% to $12.15; non-accruals were two loans, about $16 million of cost marked to $5.5 million, under 0.5% of cost. Venture lending carries warrants, and its losses arrive in clusters when the venture cycle turns.

Sixth Street Specialty Lending (TSLX)

Sixth Street held NAV flat at $16.24, earned NII of $0.43, and carried three non-accrual positions at 1.3% of fair value with none added in the quarter. It has traded at a premium for most of its listed life, about 15.5% in August 2026, for a NAV per share that has not fallen over a cycle and a habit of returning excess income as special dividends.

Read this chart with the discount chart and the logic is plain: a premium for rising NAV and low non-accruals, a discount for software and for non-accruals. The 20% gap between Blue Owl’s price and its book, and the 33% gap at FS KKR, are either the best value on the list or the most honest numbers on it; the filings are how you decide which.

A BDC’s 10-Q or 10-K runs to hundreds of pages, and the decision lives in ten numbers, each compared with the same fund’s prior eight quarters. Read them in this order.

1. Net investment income per share, from the statement of operations: income after every expense, before gains and losses. Against the distribution declared for the quarter, below 1.0× means you are being paid from something other than earnings.
2. NII with PIK removed. The notes disclose interest paid in kind. Subtract it and recompute coverage; above 1.0× on this second calculation is the test that matters, and the release does not do it for you.
3. Non-accruals at cost and at fair value. Cost says how much lending went wrong; fair value says how much has already been written off. A wide gap, as at FS KKR (7.1% against 3.8%), means the writedown has been taken; a narrow one means it has not.
4. Weighted average yield on debt investments at cost and at fair value. A fair-value yield rising well above the cost yield means the book is being marked down while coupons stay the same, stress arriving before non-accruals.
5. Debt-to-equity, gross and net of cash. The statutory ceiling is 2.0×; the practical ceiling, set by the rating agencies and the fund’s note covenants, is about 1.25×. Above 1.3× there is no room to absorb markdowns without selling loans.
6. Internal risk ratings. The share in the bottom two categories, and its change over eight quarters, is the watch list. Golub’s 87% in its top two and KBRA’s sector figure of 93.7% at or above expectations show what normal looks like.
7. First-lien share at fair value, a first lien being the debt repaid first when a borrower fails. Above 80% is a senior book; below 60%, the yield is coming from subordinated paper and equity, which is where the 2008-vintage BDCs died.
8. PIK as a share of total investment income, and whether the fund separates loans that were cash-pay at origination from those amended to PIK. The second disclosure is rare and the best single sign of a candid manager.
9. Industry concentration. More than a quarter of fair value in software is the 2025–2026 risk whether or not non-accruals show it yet.
10. NAV per share over eight quarters against dividends paid. A fund whose NAV has fallen by more than a third of its cumulative dividends has been returning your capital and calling it income.

Two footnotes. Non-accrual is a manager’s decision, and the incentive to delay it is direct, since it cuts the distribution and the incentive fee at once; and fair value on a loan amended three times is an estimate of a security nobody has bid on. The share price is the only number in the filing the manager did not write.

Sector-wide, non-accruals are rising from a low base and are concentrated by industry, and the data below places any fund on the curve. PitchBook’s count of borrowers with at least one loan on non-accrual across the BDC universe reached 4.69% in Q1 2026, from 4.26% a year earlier and 3.69% in Q1 2023, about forty basis points a year for three years. At the ten largest BDCs, non-accrual tranches were 3.95% of debt at cost in Q2 2026, up 20 basis points on the quarter, and borrowers with any non-accrual tranche were $5.0 billion and 5.95% of debt across 101 companies.

KBRA’s rated universe, which skews to higher quality, showed a median of 2.5% of cost and 1.3% of fair value in its Q3 2025 compendium; in its Q2 2026 compendium, across 35 rated funds, the median for non-perpetual BDCs rose to 2.75% of cost at June 30, 2026 from 1.81% a quarter earlier.

Two features matter more than the level. Non-accruals understate stress by construction, because the modern remedy for a troubled loan is amendment rather than default: a maturity extension, a covenant reset, a switch to PIK, a sponsor equity injection, each of which keeps the loan accruing. About $522 million of cash interest across 213 BDCs in Q1 2026 was attributable to loans on non-accrual, roughly 138 basis points of $38 billion per PitchBook; the loans amended to stay current appear nowhere in that figure.

The stress is concentrated by industry: software was 26% of distressed investments at fair value across the twelve largest listed BDCs in Q1 2026, from 19% at the end of 2025, its marks falling faster than the rest of the book, against a median software share of about a quarter of portfolios. The 2025–2026 cycle is a software cycle.

The honest comparison is with previous cycles: double-digit non-accruals at the large BDCs of 2008–2009, and above 10% at Fifth Street before its 2015–2016 collapse. A sector median under 3% of cost in mid-2026 is elevated against 2021 and low against every real downturn, and the 20% to 33% discounts at the software-heavy names are priced for the second comparison.

Payment-in-kind interest is where dividend coverage is most often flattered and where the data moved most in 2024–2026. A PIK loan pays interest by adding to principal rather than sending cash. The lender books the income, pays fees and incentive fees on it, and distributes it out of other cash while the borrower’s debt grows. Some PIK is designed at origination, in recapitalisations and growth companies; a loan that was cash-pay when written and has since switched to PIK has been restructured, and the second kind is the one that matters.

PitchBook’s series for the fifteen largest listed BDCs puts PIK at 9.0% of interest income in Q1 2023, 8.6% at the end of 2025 and 8.2% in Q1 2026, the lowest since Q4 2023 and the third quarterly decline in a row; the share has sat in a band of roughly 8% to 9% throughout, while the dollar amount grew with the portfolios before falling to $229 million in Q1 2026, a two-year low.

The listed funds, marked daily by a market that punishes PIK, have been working it down; at the nine largest non-traded BDCs PIK income rose 42% year over year in the same quarter, to $285.4 million from $200.7 million, per AltsWire. Penn Mutual Asset Management’s June 18, 2026 note across 32 BDCs found the share still growing and often arriving through amendment. Loans with a PIK component were about 16% of investments, roughly $20 billion, across the fifteen largest listed BDCs at June 30, 2025, with five funds holding about 76% of it, per PitchBook.

The arithmetic is short. Take the $8.80 of net investment income from the waterfall and suppose 10% of gross income, $2.10, was PIK. Cash NII is $6.70 against an $8.80 dividend: coverage of 0.76× on cash, reported as 1.0× on GAAP, with the difference borrowed, issued or sold. Every large BDC discloses PIK income somewhere in the 10-Q; the better ones show it as a share of total investment income and split the designed from the amended. Our rule: PIK above 10% of total investment income and rising is a reason to stop adding, and above 15%, or rising alongside non-accruals, the distribution is a fiction maintained for the share price.

Who runs a BDC and how they are paid is the structural feature that most determines its long-run return. Most listed BDCs are externally managed: the fund has no employees and pays an affiliated adviser a base fee on assets plus an incentive fee on income. A minority are internally managed, with staff paid in salary and stock as an operating expense. The cost difference is roughly a factor of two, and it compounds.

The external terms, as published. Ares Capital charges 1.5% a year on total assets, stepping down to 1.0% on assets financed above 1.0× debt-to-equity, plus a 20% income incentive fee above a 1.75% quarterly hurdle (7% annualised) with full catch-up, plus 20% of net realised gains. Blue Owl Capital Corporation and Sixth Street Specialty Lending use the same base with 17.5% over 6% hurdles; FS KKR charges 1.5%, 1.0% on assets financed above 1.0× leverage, and 17.5% over 7%, cut from 20% when its advisory agreement was last amended; Golub charges 1.0% and 15% over 8%.

Because the base fee is on gross assets, 1.5% at 1.0× leverage is 3% of your equity before the incentive fee, and the incentive fee is charged on income that includes PIK. With interest and other costs, the typical external BDC runs an expense ratio of about 4% to 4.5% of total assets, or 30% to 40% of revenue, per the Motley Fool’s undated explainer on externally managed BDCs.

The internal terms. Main Street Capital reports operating expenses of about 1.5% of assets against 3.1% for the average externally managed BDC, per Simply Safe Dividends’ undated reading of its filings; Hercules runs similarly. No incentive fee, no fee on borrowed money, and compensation in the stock, which is much of why both trade at premiums and why Main Street’s NAV per share has risen through cycles. The conflict is different, not absent: staff paid to grow the fund find that accretive issuance above NAV becomes a business in itself.

Put the Golub terms into the waterfall from the record section: at 1.0% and 15% over 8%, the same $200 of loans yields $10.20 of NII instead of $8.80, and 8.2% on NAV after average losses instead of 6.8%. Put the internal structure in, 1.5% of assets as operating cost and no incentive fee, and it is $12.00 of NII and 10% after losses. Over ten years the gap between the first and the third is the difference between doubling your money and not. The external fee also explains the sector’s two worst habits: a manager paid on assets wants a bigger fund, and a manager paid on booked income wants PIK to count.

Prospect Capital shows how the conflict resolves. An externally managed fund that has paid a monthly dividend for two decades, it cut the payout about 25% in 2024 and again to $0.035 a month in May 2026 after its fiscal Q3 2026 report; NAV per share fell from about $7.25 at the end of 2024 to $6.21 at December 31, 2025 and $6.05 at March 31, 2026, per its results as reported, with realised losses of $449.8 million across two quarters. The fee, calculated on assets rather than on the shareholder’s result, was paid throughout.

The listed sector has been through four downturns, and the shape of each is the shape the next will take. In every case the price moved first and the NAV a year later, and the funds with the most subordinated paper, leverage and fee-driven growth did not come back.

2008–2010: Allied and American Capital

The two giants of the pre-crisis sector, Allied Capital and American Capital, held large equity and mezzanine books and paid dividends partly from realised gains. Allied had paid a dividend every quarter since 1963 and carried $403 million of undistributed taxable income into 2008; by September 2008 it had breached the net-worth covenant on its bank line on unrealised depreciation, and on March 26, 2010 its shareholders accepted Ares Capital’s offer, valued at $3.47 a share, about 90% below the 2007 peak. American Capital suspended its regular dividend in 2008 and was absorbed by Ares Capital on January 3, 2017, in a deal valued at about $3.4 billion. The survivors were the senior lenders.

2015–2016: Fifth Street and the energy loans

A smaller cycle, driven by energy loans and by the funds that had grown fastest on fees. Fifth Street Finance paid no dividend for February 2015 and cut the monthly payment thereafter by about a third; its non-accruals topped 10%, Fitch cut it to BB+ on February 23, 2015, its stock traded at a 21.3% discount on February 20, 2015, and its NAV fell from $9.00 in September 2015 to about $8.40 by year-end.

Over the prior two years its dividend had fallen 37% and NAV 7% while management fees rose 56%. Oaktree bought the management contracts for $320 million in October 2017. Sector-wide, the listed BDCs traded at about 77% of NAV, a 23% discount, at the February 2016 low, per Baird as reported by Benzinga, and the gap closed within a year.

2020: the fastest drawdown

The fastest drawdown in the sector’s history. Between February 20 and March 23, 2020 the listed BDC index lost roughly half its value on the price record (a figure we could not re-open from this desk), as much as the levered syndicated-loan indices per Morningstar, while March 31 NAVs fell by a tenth or so. Almost no listed BDC breached a covenant, the discount had closed by spring 2021, and the BDC index outran the loan indices later that year. A first-lien BDC at 1.0× leverage took a 50% price move and a 10% NAV move and paid its dividend through both.

2025–2026: the software cycle

The software cycle, still open in September 2026. It has none of the leverage of 2008 and none of the energy concentration of 2015; what it has is a decade of lending to sponsor-backed software companies at six to seven turns on the theory that recurring revenue does not recur less, and a year in which artificial intelligence made that theory a question. Our tape shows the three legs; the listed sector’s version is the discount chart and FS KKR’s sponsor rescue.

The lesson that repeats is about what kills a BDC. A 30% price fall never has; every cycle’s fall reversed within two years for the funds that kept lending senior. What killed Allied and American Capital was subordinated paper marked at prices no buyer would pay, and what killed Fifth Street was a manager paid to grow. The 2026 cases that rhyme are the funds with the most PIK and the widest gap between non-accruals at cost and at fair value.

A 10% yield and a 10% return are the same number only if NAV holds and the discount does not move; three lines turn one into the other. Total return over a holding period is dividends received, plus the change in NAV per share, plus the change in the discount applied to the ending NAV, all over the price paid.

Dividends are the reliable term when coverage is above 1.0× on cash NII; mid-2026 coverage ran from Hercules’s 125% of its base distribution down to Ares’s core earnings of $0.47 against a $0.48 dividend, about 0.98× (1.04× on its GAAP net investment income of $0.50), thin enough that a 50 basis point fall in SOFR, the overnight rate the loans float over, or a one-point rise in non-accruals puts the dividend under the income line.

NAV per share is the term most investors ignore: a fund that pays out all its income and absorbs 1% of assets a year in credit losses at 1.0× leverage loses about 2% of NAV a year unless accretive issuance replaces it. The discount is the wild card: from 20% to zero it adds 25% to your return; the reverse takes the same away.

With the terms filled in, for a purchase at 90 cents on the NAV dollar: a 9.6% dividend on NAV is a 10.7% yield on price; NAV eroding 2% a year subtracts about 2.2 points; the discount unchanged adds nothing; expected total return about 8.5% a year. Widen the discount to 20% over five years and it falls to about 7%; close it and it rises to about 10.5%. The dividend is the least important of the three terms to forecast and the only one the sector advertises.

Tax is the largest cost in the worked example, and it is the same for every listed BDC. A regulated investment company passes income through with its character preserved. Interest, nearly all of a lending BDC’s income, arrives as ordinary dividends, taxed at your marginal federal rate, up to 37% in 2026, plus state tax and, above $200,000 of modified adjusted gross income single or $250,000 joint, the 3.8% net investment income tax under IRC §1411, thresholds unindexed since 2013.

It is not a qualified dividend, because the fund earned it as interest; a BDC with equity holdings can designate the small part that came from portfolio companies’ qualified dividends. Long-term gains realised by the fund can be designated as capital-gain dividends, taxed at 20% at the top; return-of-capital distributions, which appear when a fund pays more than it earns, are untaxed on receipt and reduce your basis. The Form 1099-DIV and the fund’s annual tax-character notice, usually in January or February, split the year’s distributions among these.

The comparison that costs BDC holders money is with REITs. Since 2018, IRC §199A has allowed a 20% deduction on qualified REIT dividends, which the One Big Beautiful Bill Act of July 4, 2025 made permanent; the House-passed version of that bill extended the deduction to “qualified BDC interest dividends,” and the Senate dropped it. The Small Business Investor Tax Parity Act (S.2962, 119th Congress, referred to the Senate Finance Committee on October 1, 2025, with an identical House bill, H.R. 652) would restore it for tax years beginning after December 31, 2026, and as of September 2026 it sits in committee.

Until it moves, a BDC dividend and a REIT dividend of the same size are taxed at 37% and 29.6% respectively in the top bracket, and the 2026 §199A thresholds of $201,750 single and $403,500 joint per Rev. Proc. 2025-32 are irrelevant to BDC holders. Watch that bill; it is worth about 7 points of tax rate on every dollar of BDC income to a top-bracket holder.

Placement follows from the character. A BDC’s dividend is ordinary income in a taxable account and is not unrelated business taxable income in a retirement account, because a RIC’s dividends are excluded from UBTI, unlike a partnership’s; the listed BDC is therefore one of the best uses of IRA and Roth space that exists, converting the worst-taxed income in the market into untaxed or deferred income at no cost. Foreign holders get a smaller version of the same relief: a BDC can designate interest-related dividends exempt from the 30% US withholding tax under IRC §871(k), made permanent in 2015.

One position run through five years in dollars sets the fee, the discount, the tax and the account type against each other. The position is $50,000 in a listed BDC bought at $18.00 a share against a $20.00 NAV, a 10% discount, 2,778 shares; the fund pays $1.92 a share a year, 9.6% on NAV and 10.7% on your price, close to mid-2026 payouts; the holder is in the 24% federal bracket with a 5% state tax, below the NIIT thresholds; sale at the end of year five.

Path A: nothing happens

NAV stays at $20.00, the dividend holds, the discount stays at 10%. Dividends: 5 × $1.92 × 2,778 = $26,667. Sale at $18.00: $50,000. Pre-tax $76,667, an 8.9% annual return. Tax on the dividends at 29%: $7,733, paid year by year. Net $68,933, or 6.6% a year. In a Roth IRA the same path nets the full $76,667; the wrapper is worth $7,733, more than any fee in the table. At 37% federal with NIIT and the same state tax, the taxable version nets $64,453, or 5.2% a year.

Path B: a slow leak

NAV erodes 2% a year, the dividend is cut 15% in year three, the discount widens to 20%. Dividends $24,250. Sale at $14.46: $40,174, a capital loss of $9,826 worth about $1,965 against other gains. Net $59,357, or 3.5% a year. This is what a 2% annual NAV leak, the fee waterfall’s own arithmetic, does to a five-year hold: you kept your money and were paid a Treasury-bill return for loan risk.

Path C: the discount closes

NAV grows 1% a year on accretive issuance, the dividend rises to $2.00 by year five, the discount closes to zero. Dividends $27,111; sale at $21.02, $58,389; gain $8,389 taxed at 20% federal-plus-state, $1,678. Net $75,960, or 8.7% a year taxable, and $85,501, or 11.3%, in a Roth. This is what buyers of Ares at $18.76 in August 2026 were underwriting.

Path D: the fund goes wrong

NAV falls 5% a year, the dividend is cut 30% in year two, the discount widens to 30%. Dividends $20,222; sale at $10.83, $30,091. Pre-tax $50,314, five years for nothing; after tax and the loss benefit, $48,431, a slightly negative return, with tax on $20,222 of dividends paid along the way. This is the Fifth Street path.

Three things fall out of the table. The account decision is worth $7,733 on Path A, more than the whole difference between the single name and the ETF ($991, the ETF’s 0.42% fee compounding plus its drag on the sale). The discount is worth more than the dividend: Path C differs from Path A by $8,834 pre-tax, and $8,389 of that is the discount closing.

The tax is regressive against outcomes: on Path D you paid $5,864 of tax on dividends from a position that finished $1,569 below your cost after tax, because the income and the loss are taxed in different years at different rates. That is the strongest case for the Roth, where the two net inside the wrapper and nothing is owed on either.

The listed BDC is the one private-credit vehicle where the whole process is a brokerage order. There are three routes in, each with a published cost, and then a sequence. The non-traded and perpetual BDCs, interval funds and tender-offer funds are the ground of the flagship and of the sister guide Investing in Interval Funds and Non-Traded BDCs; the discount section already says why a listed share below NAV is the cheaper claim on the same loans than a non-traded sibling redeeming at 100.

A single listed BDC

Any brokerage account, no accreditation, no minimum beyond one share, commission-free at the large brokers, a spread of a cent or two on the large names. You choose the manager, the fee structure and the discount; it is the route for anyone willing to read the ten lines once a quarter.

The sector ETF

VanEck’s BIZD holds the listed universe by market capitalisation, so Ares Capital and the other giants dominate it, and charges a management fee of 0.40% plus 0.02% of other expenses; the 9.69% total expense ratio in its prospectus adds 9.27% of “acquired fund fees,” the SEC-mandated look-through of the underlying BDCs’ own expenses, which are already inside their share prices and are not a second charge.

Putnam’s PBDC is the actively managed alternative at a 0.75% management fee. The ETF gives you the sector’s average discount and average fee problem with no way to avoid the worst fund in it, in exchange for one line and no single-name risk; its 30-day SEC yield was 9.60% and its trailing 12-month distribution yield 13.59% on July 31, 2026, per VanEck.

A closed-end fund of BDCs

Several exist, at fees of 1% to 2% plus their own leverage and discounts; they add a layer of each to an asset that already has both, and we do not see the case.

The sequence

Six steps, in order; only the third takes real time.

1. Decide the size first. A listed BDC is an equity with a loan book behind it and 20% annualised volatility on our tape; size it in the equity budget, not the bond budget, at a weight you can watch fall 30% without selling.
2. Open the room in a Roth or traditional IRA if you have it. The tax section’s arithmetic is worth more than any selection skill you will bring.
3. Pull the last 10-K and two 10-Qs for three candidates and fill in the ten lines. This takes an evening per fund.
4. Compute the discount to the latest reported NAV on the day you buy and write it down with the date. It is your margin of safety and your exit signal.
5. Buy in thirds over a quarter; the sector’s worst and best days on our tape were one session apart, and the discount moves five points in a week.
6. Each quarter, on the day the 10-Q lands, refresh the ten lines and the discount. Sell on the coverage rule or on a premium above 25% at an external manager. Otherwise hold; the loans pay quarterly and the discount does the rest.

Eight readings would change our view. Each carries a threshold and an as-of date, so a reader in 2027 can check it against the filings of the day.

• The sector discount. FS KKR about 33% below NAV, Blue Owl Capital Corporation about 20% below and Ares Capital about 3% below in August–September 2026. A sector-wide narrowing to under 10% says the market believes the marks; a sector median past 30%, or Ares or Blue Owl past 30%, says it expects the marks to come to it.
• Non-accruals at cost. PitchBook’s universe borrower rate 4.69% in Q1 2026, rising about 40 basis points a year; top-10 tranches 3.95% of cost in Q2 2026. Above 6% at the universe level, or above 5% at any of the seven other than FS KKR, is the 2015–2016 comparison arriving.
• PIK share. 8.2% of interest income at the largest listed BDCs in Q1 2026, falling for three quarters. A reversal above 10%, or any fund above 15%, is our stop-adding signal.
• Coverage at the largest fund. Ares Capital’s core earnings of $0.47 (GAAP net investment income $0.50) against a $0.48 dividend in Q2 2026. Two quarters below 0.95× at Ares would mean the sector’s dividends reprice, because Ares is the one the rest are marked against.
• SOFR. About 3.68% in early September 2026 with the funds target at 3.50–3.75%. The loans float; each 100 basis points off SOFR takes roughly a point off NII yields across the sector before any floor.
• Software marks. 26% of distressed investments at the twelve largest listed BDCs in Q1 2026, from 19% at end-2025. Above a third, the cycle has a name.
• The tax bill. S.2962 in committee as of September 2026. Passage would be worth about 7 points of after-tax yield to a top-bracket taxable holder and would likely close part of the sector discount on its own.
• Our tape. BIZD at $13.33 on September 4, 2026, 25.2% below the February 19, 2025 high of $17.82 and 10.1% above the April 1, 2026 low of $12.11; the Private Credit sub-index at 83.157 on September 8, 2026. A close above $15.00 without a matching fall in non-accruals is the market front-running the marks; a close below $12.11 is a new leg.

Every figure in this guide should be checked against the filing of the day before you act on it.