After the 2008 crisis, tightening bank capital rules (Basel III's risk-weighting of leveraged corporate exposures, the same regime M3.10 built for bank valuation) made it steadily more expensive for banks to hold non-investment-grade, leveraged middle-market loans on their own balance sheets. Banks retreated; non-bank capital stepped in. In the US that capital organizes mostly as Business Development Companies (BDCs), closed-end investment vehicles regulated under the Investment Company Act of 1940, which must keep the large majority of assets in "eligible portfolio companies" (generally private or thinly-traded US operating businesses) and can be publicly listed or sold non-traded to accredited/retail investors. In India, private credit organizes mostly as Category II Alternative Investment Funds under the SEBI (Alternative Investment Funds) Regulations, 2012, closed-end pools that lend directly to promoters and mid-market companies, often via structured debentures rather than plain bank-style loans.
How does direct lending differ from syndicated loans?
Place direct lending on the leveraged-finance spectrum and the whole asset class explains itself. At one end sit broadly syndicated loans (BSL) (liquid, traded among dozens of institutional holders, usually "cov-lite" (covenants tested only when the borrower takes an action) an incurrence test). At the other end sit high-yield bonds: public, liquid, rated, incurrence-covenanted. Direct/private loans sit in between on liquidity but, tellingly, not on covenant protection: because a direct lender cannot sell out of a deteriorating credit the way a BSL holder can, it retains maintenance covenants (tested every quarter regardless of borrower activity) as a substitute early-warning system for the exit option it has given up. That trade, illiquidity for control and a first look at trouble, is precisely why direct lenders have historically been paid an illiquidity premium over an equivalent-rated liquid loan, commonly cited in the 100–300bps range against broadly syndicated benchmarks (compare a private-credit index such as the Cliffwater Direct Lending Index against the LSTA/S&P leveraged loan index, check the current spread; it compresses and widens with the capital cycle). US direct loans typically float off SOFR plus a spread; Indian private credit more often uses a fixed coupon on a structured debenture, though floating structures referencing T-bill or MIBOR benchmarks exist in the larger deals.
Advertisement
Capital structure layering & unitranche/FOLO structuring
A sponsor-backed borrower's capital structure layers by seniority and by leverage multiple of EBITDA, roughly: a senior secured revolver (drawn only for working capital, 0x), then a unitranche facility (blending what would once have been separate senior and junior debt into one instrument, typically 3–6x EBITDA), then mezzanine/subordinated debt (often 6–7x, frequently carrying warrants), then preferred and common equity below that. Everything above the sponsor's own equity check is somebody else's money at risk before the sponsor loses a rupee or a dollar, which is exactly why lenders police an equity cushion requirement (the covenants section).
The unitranche's defining trick is that the borrower sees one loan: one interest rate, one covenant package, one amortization schedule, one signature. Behind that single facility, the lenders split the economics via a private side letter called the Agreement Among Lenders (AAL), which the borrower never signs and often never sees, dividing the facility into a First-Out (FO) tranche (repaid first in a workout, therefore priced at a lower spread) and a Last-Out (LO) tranche (subordinated to the FO within the same nominal facility, therefore priced at a materially higher spread to compensate). Worked Example 1 builds this exact split with real numbers. Sponsors favor unitranche precisely because "one lender relationship instead of two negotiations" buys speed and certainty of execution in a competitive acquisition process, the same illiquidity-premium logic from the comparison with syndicated loans, now visible from the borrower's side of the table.
Covenants, OID, PIK & call protection: the yield mechanics
Maintenance covenants (leverage and interest-coverage tests run every quarter, whether or not the company has done anything) are direct lending's signature feature, and the comparison with syndicated loans already told you why: illiquidity forces the lender to build its own tripwire, because it cannot simply sell. A typical package layers a leverage covenant (total or net debt/EBITDA, with a maximum that steps down over the life of the loan as the business is expected to delever), an interest-coverage covenant (EBITDA/cash interest, with a stated minimum), an equity cushion requirement (a minimum sponsor equity contribution, commonly cited around 35–40% of total capitalization, so the sponsor's capital absorbs losses before the lender's does), and a most-favored-nation (MFN) clause protecting existing lenders' pricing if the borrower raises a similar incremental facility at a wider spread within a defined window.
What moves a lender's real yield away from the coupon?
Two mechanisms move a lender's real yield away from the headline coupon, and confusing them is one of the most common errors in this discipline (flagged again in Common Mistakes). Original issue discount (OID) is a one-time funding discount: the lender advances, say, 99 for every 100 of face value, and is repaid the full 100 at maturity or refinancing, a pure yield pickup with no effect on the stated coupon rate and no compounding. Payment-in-kind (PIK), by contrast, is a borrower option to pay some or all of the interest by adding it to principal instead of paying cash, it preserves the borrower's liquidity in a downturn at the direct cost of compounding the lender's outstanding exposure, dollar for dollar, year after year. Finally, call protection (a prepayment-premium schedule such as non-call for the first year, then 102, then 101, then par) compensates the lender for reinvestment risk if a strong-performing borrower refinances the loan away early. Worked Example 1 computes an all-in yield combining a cash coupon with OID, and a call premium under an early-refinance scenario; the mezzanine sub-example computes a PIK balance compounding over a multi-year hold.
Lender economics: default, recovery, and fund-level mechanics
A covenant breach is not a payment default. In practice, tripping a maintenance covenant typically triggers a negotiated waiver or amendment (often priced with a fee and/or a repricing of terms) rather than automatic acceleration. The lender's real leverage is the ability to accelerate, which is what forces the borrower to the negotiating table; actually exercising that right is usually the last resort, not the first move. If a credit does progress all the way to a payment default or formal insolvency, you do not need a new waterfall: it runs through the same machinery M3.10 already built, the IBC/CIRP process and Section 53 priority in India, Chapter 11 and the absolute-priority rule in the US. A direct lender is simply a secured financial creditor (senior/unitranche) or an unsecured/subordinated financial creditor (mezzanine) inside that same queue. This is exactly why the security package and the maintenance-covenant early warning matter in practice: first-lien direct loans have historically recovered materially more in a default than unsecured bonds, commonly cited ballparks run roughly 65–80% for first-lien secured debt against 30–50% for unsecured (Moody's and PitchBook/LCD recovery studies are the primary sources, check the current cycle's figures, which move with collateral values and covenant quality).
How is a direct lending fund set up and paid?
At the fund level, US BDCs may elect a reduced 150% asset-coverage ratio (the Small Business Credit Availability Act of 2018 amended the 1940 Act to allow this, up from the prior 200%, effectively permitting up to roughly 2:1 debt-to-equity at the fund, versus 1:1 before; check the current election status of any specific vehicle) and, to keep pass-through tax status as a Regulated Investment Company (RIC) under Internal Revenue Code Subchapter M, must distribute at least 90% of taxable income to shareholders. India's Category II AIFs carry no equivalent forced-distribution rule, but in practice distribute income promptly given a closed-end, current-yield-seeking LP base. Contrast this explicitly with PW1.02's fund mechanics: a PE/VC fund runs a capital-appreciation waterfall, capital is called, the J-curve bites, and carry is earned on realized gains at exit, years later. A direct-lending fund runs a current-income waterfall instead, a smaller management fee (often 1.0–1.5% versus PE's more typical ~2%), an incentive fee with a hurdle, but both computed against net investment income earned this period, not a terminal capital gain. Same GP–LP alignment grammar (management fee, hurdle, incentive fee, sometimes a clawback); a fundamentally different cash-flow engine underneath it.