alirezarezvani/claude-skills / finance/skills/stock-analysis
stock-analysis
UnverifiedAnalyze, research or value a listed stock or company (US, global, NSE or BSE) with verified data, ratios, scores and valuation. Use whenever asked if a stock is a buy, what it is worth, or how a company looks.
references/sectors/mortgage-reit-specialty-finance.md
View on GitHub# Mortgage REITs, BDCs, equipment lessors and specialty finance — sector playbook
Use this when: the company's economic engine is a **leveraged portfolio of financial assets or leased hardware**, not an operating business — US agency and hybrid mortgage REITs (mREITs), commercial mortgage REITs, business development companies and listed private-credit vehicles, listed credit funds and CLO equity vehicles, aircraft/container/railcar/energy-equipment lessors, and Indian equipment-rental, leasing and asset-financing companies that sit outside the conventional NBFC lending frame.
These businesses are routinely mis-routed, and the mis-route is catastrophic rather than merely imprecise. A mortgage REIT has the word "REIT" in its name and owns **no buildings** — occupancy, WALE, same-store NOI, AFFO, cap-rate spread and rent reversion are not "less relevant" here, they are undefined. It is a levered agency/non-agency MBS book funded overnight in the repo market. A BDC looks like an asset manager but is a levered senior-loan portfolio with a mark-to-model NAV. An equipment lessor looks like an industrial but earns a spread between funding cost and lease yield, plus a residual-value bet.
Three facts govern everything below. **Borrowings are the raw material**, so every enterprise-value construct is meaningless. **Reported earnings are dominated by marks and hedges**, so P/E is noise and book value is the anchor. And **the dividend is usually the reason the stock is owned**, so dividend coverage by a properly defined economic earnings measure — not by GAAP EPS — is the central analytical question.
**Routing:** if the entity originates and holds whole loans on its own balance sheet as its main business (NBFC, HFC, gold-loan, MFI, consumer lender), use `references/sectors/nbfc.md`. If it takes deposits, use `references/sectors/banks.md`. If it owns and leases **real property** and collects rent, use `references/sectors/realestate-reit.md`. If it manages third-party capital for fees rather than deploying its own balance sheet, use `references/sectors/holdco-assetmgr.md`.
## Contents
- [Why the generic ratio set fails here](#why-the-generic-ratio-set-fails-here)
- [The metrics that actually matter](#the-metrics-that-actually-matter)
- [How to value companies in this sector](#how-to-value-companies-in-this-sector)
- [Peer set construction](#peer-set-construction)
- [Sector-specific red flags](#sector-specific-red-flags)
- [Cycle and structural context](#cycle-and-structural-context)
- [India vs global notes](#india-vs-global-notes)
- [Checklist](#checklist)
## Why the generic ratio set fails here
Do not compute these. If a screener hands them to you, report them as inapplicable rather than reporting them with a caveat — a printed number gets used.
**Every equity-REIT metric, for a mortgage REIT — undefined, and the most damaging error in this playbook's scope.** Occupancy, WALE / WAULT, same-store NOI, rent per square foot, releasing spreads, cap rate, NAV-per-square-foot, development pipeline, AFFO and even FFO have no referent: there is no property, no tenant, no lease and no depreciation of a building. FFO exists to add back depreciation on appreciating real assets; an mREIT holds securities carried at fair value with no meaningful depreciation, so an "AFFO" figure computed for one is a fabricated number. A commercial mortgage REIT lends *against* buildings — the buildings are collateral inside someone else's balance sheet, and cap-rate movement matters only through loan-to-value and loss severity, never as a valuation multiple applied to this company.
**EV, EV/EBITDA, EV/Sales, net debt, Debt/EBITDA — undefined and never used.** Enterprise value assumes debt is financing you can strip out to see the operating asset underneath. Here the borrowings *buy the assets that generate the return*: repo funds the MBS, the credit facility funds the loan portfolio, the ABS funds the fleet. Subtracting "net cash" removes the liquidity buffer that determines survival; adding debt to market cap double-counts the asset side. For lessors specifically, EV/EBITDA is superficially computable and therefore especially dangerous — EBITDA excludes both depreciation (the actual consumption of the leased asset) and interest (the actual cost of the raw material), which between them are most of the cost base. A lessor's "EBITDA margin" of 80–90% is an accounting artefact, not a moat.
**D/E, gearing and interest coverage — mis-signed.** Agency mREITs run 6–9x debt/equity **by design**, and a well-run one deleveraging to 4x may be signalling stress, not prudence. Hybrid and credit mREITs run 1–4x; BDCs are capped near 2:1 by statute in the US; lessors run 2–4x. A generic "D/E above 2 is risky" screen rejects the entire sector's healthiest names and passes the impaired ones. Interest coverage is nearly meaningless when interest expense is the cost of goods sold — though for lessors and BDCs a fixed-charge coverage or asset-coverage ratio *does* bind contractually, so read the covenant, not the textbook.
**OPM, gross margin, EBITDA margin, ROCE, asset turnover — meaningless.** There is no COGS and no unit of output. "Capital employed" would include every repo line. Asset turnover collapses to portfolio yield, which is measured directly and better. For a BDC, "revenue" is total investment income, which is just yield × assets — a margin on it tells you only the fee and leverage structure.
**Free cash flow and FCF yield — noise, and frequently inverted.** Cash flow is dominated by portfolio purchases, paydowns, repo rollovers and fleet capex. A growing lessor buying equipment shows deeply negative FCF; a shrinking one liquidating its fleet shows strongly positive FCF at the exact moment its earning base is disappearing. Operating and financing flows cannot be cleanly separated, so no FCFF/WACC model can be built.
**P/E on reported EPS — noise, and often perversely signed.** GAAP net income for an mREIT includes unrealised marks on securities and on hedges, which frequently move in opposite directions and can swamp the actual spread income; a quarter with a large derivative gain can show enormous "EPS" while book value fell. BDC GAAP EPS mixes net investment income with unrealised depreciation. Use net interest spread / EAD / distributable earnings for mREITs, NII for BDCs, and always cross-read against book value change.
**Current ratio, quick ratio, working capital, inventory and receivable days, cash conversion cycle — undefined.** These balance sheets are portfolios, not classified current/non-current in any economically meaningful way. Liquidity here means unencumbered assets, undrawn facilities and days of margin-call capacity.
**Dividend yield read as a quality signal — inverted.** A 15%+ headline yield in this sector is a market forecast of a dividend cut and/or book-value erosion, not a bargain. Yield is an output of price falling, and price falls here almost always because book value fell or coverage broke.
**"Growth in AUM/portfolio" as an unqualified positive — inverted.** Portfolio growth funded by issuing equity **below** book value or NAV is per-share value destruction dressed as scale. Always convert to per-share terms.
## The metrics that actually matter
Ranges are **indicative only**. They vary by market, sub-sector, rate and credit cycle, and by period, and any of them can be wrong for a specific company in a specific year. Comparison against a tight peer set and against the company's own 5–10 year history overrides every absolute band below. Metrics marked (M) are mortgage-REIT, (B) BDC/private credit, (L) lessor; unmarked ones apply across all three.
| Metric | Definition / how to compute | Indicative healthy range | Why it matters |
|---|---|---|---|
| **Book value per share (BVPS) / NAV per share** | (M) Shareholders' equity less preferred liquidation preference, ÷ common shares. (B) Net assets ÷ shares, struck quarterly by the board using ASC 820 fair value. (L) Book equity per share, plus a separate estimate of fleet market value vs carrying value. | Not a level — a *trajectory*. Flat-to-rising BVPS across a full rate cycle is the mark of a well-hedged mREIT; a 30–50% multi-year decline is common among the weak ones. | This is the entire foundation. It is the denominator of the primary multiple, the base for economic return, and the thing that is silently destroyed while the dividend keeps being paid. Always adjust for share issuance and buybacks — aggregate equity growth from serial ATM issuance conceals falling BVPS. |
| **Economic return on book value** | (Change in BVPS over the period + dividends declared per share) ÷ opening BVPS, annualised. The sector's single most important number. | Roughly 8–14% through a cycle for an agency mREIT; a negative economic return in any year where credit did not blow up means the hedge or leverage decision failed. | Dividends paid out of returned capital are not income. A company can pay a 14% yield while book falls 18% — the shareholder's economic return is negative and the "yield" was a partial liquidation. This metric is the only honest way to score management, and it is the correct thing to rank on, never dividend yield. |
| **Net interest spread and net interest margin (M)** | Weighted average asset yield − weighted average cost of funds (including the effect of interest-rate swaps). NIM = net interest income ÷ average interest-earning assets. | Agency spread roughly 1.0–2.0%; NIM on levered equity is spread × leverage plus unlevered yield. Direction and *composition* matter more than level. | The engine. A widening spread driven by cheaper repo is durable; one driven by buying lower-coupon, longer-duration paper is a duration bet in disguise. Always check whether swap costs are inside the reported spread or presented below it — presentation differs across companies and is a common comparability trap. |
| **Leverage: debt/equity and economic leverage (M, B, L)** | (M) Repo + other borrowings ÷ total equity; *economic* leverage adds the notional of TBA dollar-roll positions, which are off-balance-sheet long exposure. (B) Debt ÷ net assets, against the statutory cap. (L) Net debt ÷ tangible equity, plus off-balance-sheet lease commitments. | Agency mREIT 6–9x; hybrid/credit mREIT 1–4x; BDC 0.9–1.25x typical against a 2.0x US regulatory ceiling; lessors 2–4x. | Leverage is the business model, but it is also the mechanism of ruin. Compare *economic* leverage, not stated: a company at 6x stated and 8.5x economic via TBAs is not comparable to a peer at 6x with no TBA book. A BDC operating at 1.5x+ has little room before a mark-driven covenant breach forces asset sales into a falling market. |
| **Funding profile: repo mix, tenor and haircuts (M)** | % of borrowings in repo, weighted average days to maturity, number and concentration of counterparties, haircut by collateral type (agency vs non-agency vs whole loan), and unencumbered assets as % of equity. | Agency haircuts typically 3–5%; non-agency and credit collateral 15–30%+. Weighted average repo tenor of 30–90 days is normal; concentration in the top 3 counterparties well under half the book. | This is how mREITs die — not from credit, but overnight. Repo is short-dated and haircuts widen exactly when marks fall, so a price decline triggers margin calls, forced sales, further price declines. Every mREIT failure in March 2020 followed this sequence, and the ones that survived were the ones with unencumbered liquidity and agency-only collateral. |
| **Duration gap and convexity (M)** | Asset duration − liability-plus-hedge duration, in years. Read with the disclosed rate-shock table (BVPS change for ±25/50/100 bps). | Duration gap typically within ±1.0 year; a ±100 bp shock moving book value more than 8–12% signals an unhedged rate bet. | Tells you how much of the equity is a bet on rates rather than on spread capture. Negative convexity is the specific curse of MBS: when rates fall, borrowers refinance and the asset shortens; when rates rise, prepayments stop and the asset extends — the portfolio lengthens exactly when duration is most painful. |
| **Hedge ratio and hedge composition (M)** | Notional of interest-rate hedges ÷ repo and other rate-sensitive borrowings; split into swaps, futures, swaptions, TBA shorts. | 70–100% is typical for a fully hedged agency book. Sudden moves in this ratio are management taking a view. | A falling hedge ratio into a hiking cycle is the classic setup for book-value destruction. Also check *what* is hedged: hedging the funding leg does nothing about spread widening, which is a separate and often larger risk. Post-LIBOR, most books are SOFR-based — confirm the transition is complete and basis risk is disclosed. |
| **Prepayment speed (CPR) and extension risk (M)** | Constant prepayment rate, annualised % of principal prepaying. Read with weighted average coupon, loan age, and the premium/discount to par at which assets are carried. | Highly regime-dependent: single-digit CPR in a high-rate, out-of-the-money regime; 20–40%+ in a refi wave. What matters is CPR *versus what the carrying premium assumes*. | Premium-priced MBS amortise faster than expected when CPR spikes, destroying yield and book value; discount-priced MBS suffer when CPR collapses and the expected pull-to-par is deferred. Specified pools and loan-balance stories exist precisely to buy prepayment protection — check whether the company is paying up for it and whether the protection is actually delivering. |
| **Agency vs credit exposure (M)** | % of portfolio in agency (Fannie/Freddie/Ginnie-guaranteed) vs non-agency RMBS/CMBS, credit risk transfer, whole loans, mezzanine and bridge loans. For commercial mREITs: loan-to-value distribution, watchlist/risk-rated 4–5 loans, CECL reserve as % of loans, and office concentration. | Agency-only books have essentially no credit risk and pure rate/spread/prepay risk. Credit books need loss estimates. Watchlist loans <5% of the book and not rising. | These are two entirely different businesses sharing a label. An agency book's risk is duration and repo; a credit book's risk is default and severity, with far lower leverage and far wider funding haircuts. Never place them in the same peer scatter. |
| **Dividend coverage by economic earnings** | (M) Earnings available for distribution (EAD) / distributable earnings / "core EPS" ÷ dividend per share — *and* recompute it yourself, since these are non-GAAP measures the company defines. (B) Net investment income per share ÷ dividend per share. | Coverage ≥1.0x sustained; 0.9–1.0x is a warning; below 0.9x for two or more quarters usually precedes a cut. | The dividend is the thesis for most holders of these securities, and it is the last thing management cuts. Scrutinise the non-GAAP bridge: excluding "realised losses on terminated swaps" or capitalising origination costs can turn uncovered into covered. Check whether the payout is being funded by return of capital — the tax character disclosed on the US Form 1099-DIV or in the annual distribution statement tells you directly. |
| **Non-accruals (B)** | Investments on non-accrual, expressed both at **cost** and at **fair value**, as % of the portfolio. Always take both. | Under 1–2% at fair value is clean; above 3–4% at cost, or a widening cost-vs-fair-value gap, indicates real deterioration. | The core asset-quality metric for private credit. The gap between the two measures is itself the signal: non-accruals at 5% of cost but 2% of fair value means the marks have already absorbed roughly 60% severity. A company quoting only the fair-value number is presenting the flattering half. |
| **PIK income as % of total investment income (B)** | Payment-in-kind interest (accrued, not received in cash) ÷ total investment income. Track the trend and split "structural" PIK (agreed at origination) from "amendment" PIK (granted to a struggling borrower). | Below 5–8% is normal; above 10–15% and rising is a serious quality tell. | The highest-signal quality metric in private credit. PIK is income you have booked but not collected; a borrower converted to PIK is a borrower that cannot pay cash. It flatters NII and therefore dividend coverage while the underlying credit degrades, and it typically appears one to three quarters before non-accruals rise. |
| **Portfolio yield vs cost of debt, and first-lien mix (B)** | Weighted average yield on debt investments at cost and at fair value; weighted average interest rate on borrowings; % first-lien senior secured; % unitranche; % equity and junior positions; average EBITDA of portfolio companies; weighted average interest coverage of borrowers. | Spread of 400–700 bps over cost of debt typical. First lien >70–80% for a conservatively positioned BDC. Borrower interest coverage below ~1.5x across a large share of the book is a stress signal. | Yield without a corresponding funding cost tells you nothing about earnings power. A rising portfolio yield that comes from moving down the capital structure or into smaller borrowers is rented, not earned. Borrower-level interest coverage is the closest thing to a leading indicator of the next non-accrual wave. |
| **Fee structure and external management drag (B, M)** | Base management fee as % of gross (not net) assets, incentive fee rate and hurdle, whether there is a total-return hurdle with a lookback, and fee waivers with expiry dates. Express total fees as % of average net assets and as % of gross investment income. | Base 1.0–1.5% of gross assets, incentive 15–20% over a ~6–8% hurdle, is the common shape. Total fee load above ~3% of net assets is a heavy drag. | Externally managed vehicles have a structural conflict: a fee on *gross* assets rewards raising and levering capital regardless of per-share returns. This is the mechanism behind dilutive issuance below NAV. A fee waiver supporting current dividend coverage is a temporary subsidy with a known expiry — model the cliff. |
| **Spillover / undistributed taxable income (B, M)** | Taxable income earned but not yet distributed, per share. Disclosed by US RICs and REITs. | One to two quarters of dividends of spillover is a healthy cushion. Falling spillover with flat dividends means the buffer is being spent. | The shock absorber that lets a vehicle maintain its dividend through a weak quarter. Its depletion is an early, quantitative warning of a cut — visible before coverage formally breaks. |
| **Fleet age, utilisation and lease yield (L)** | Average fleet age vs economic life; **time utilisation** (units on hire ÷ units available) and **dollar utilisation** (rental revenue ÷ original equipment cost); lease yield = lease revenue ÷ net book value of equipment on lease; average remaining lease term. | Time utilisation 85–95% strong for aircraft/containers, 65–80% typical for general equipment rental; dollar utilisation trends matter more than the absolute. Fleet age well inside economic life. | Time utilisation tells you whether the assets are working; dollar utilisation tells you whether they are working *at a price*. The two diverge exactly when pricing is deteriorating but management keeps units on hire by cutting rates — high time utilisation with falling dollar utilisation is the classic pre-downturn configuration. An ageing fleet is deferred capex and a future residual-value problem. |
| **Residual value realisation and gain/loss on disposal (L)** | Proceeds on sale of off-lease equipment vs net book value, as a %, tracked over 3–5 years; disposal gains as % of pre-tax profit; the depreciation policy's assumed residual and useful life vs peers. | Disposal gains a modest and stable share of profit — persistently above ~20–25% of pre-tax profit means earnings depend on selling the fleet. | The lessor's central hidden bet. Optimistic residual assumptions understate depreciation and overstate current profit for years before the truth arrives at disposal. Consistent gains on sale mean the fleet is conservatively depreciated (a hidden reserve); losses mean the book is overstated and impairments are coming. Changes to useful-life or residual assumptions are a first-order earnings-quality event — always find the note. |
| **Funding cost, mix and maturity ladder (L, B)** | Weighted average cost of debt; split secured/ABS vs unsecured bonds vs bank revolver; fixed vs floating; debt maturities by year against lease cash inflows; % of assets encumbered. | Debt maturity profile matched to lease term; a large unencumbered asset pool; no single year holding an outsized share of maturities. | Borrowing short to lease long is the sector's recurring failure mode. An asset–liability maturity mismatch converts a funding-market closure into an insolvency. Rising encumbrance means the unsecured lenders are being structurally subordinated — usually the first visible sign of funding stress. |
| **Credit losses on lessees / receivables (L)** | Provision for doubtful lease receivables ÷ average lease receivables; ageing of receivables; concentration of the top 5–10 lessees; exposure to lessees in a single sector or country. | Loss rates in the low tens of basis points for investment-grade lessee books; higher for SME rental. Top-10 lessee concentration well under half of revenue. | A lessor is a lender whose collateral it already owns — which helps on severity but not on the earnings hole from an unexpected off-lease unit. Lessee concentration turns one airline or one shipping-line failure into a capital event, and repossession across jurisdictions is slow and expensive. |
## How to value companies in this sector
Value the **equity directly, against the asset base it owns**. Every enterprise-value method is structurally inapplicable, and every earnings multiple applied to mark-driven profit is noise.
**Primary for mREITs and BDCs — price-to-book / price-to-NAV, regressed against sustainable ROE.** These are portfolios of financial assets carried close to fair value, so book value is a genuine estimate of what the equity owns, not a historic-cost relic. The multiple is not a free parameter — it follows the same warranted-multiple logic used for banks:
> **Justified P/B = (sustainable ROE − g) ÷ (COE − g)**
In practice, run a cross-sectional scatter of P/B against economic return on equity (mREITs) or against NII return on NAV net of fees (BDCs) across a tight peer set, and explain the residual with hedge discipline, funding quality, non-accrual history, fee load and management track record. A persistent discount to book is usually the market pricing (a) future book erosion, (b) the fee drag of external management, or (c) disbelief in the marks — decide which before calling it cheap.
**Adjusted book value.** For credit mREITs and BDCs, reported book contains marks you may not accept. Re-strike NAV by: haircutting level-3 assets to observable proxies where they exist; applying your own severity to non-accruals and watchlist loans; deducting the capitalised value of any fee-waiver subsidy; and deducting preferred stock at liquidation preference, not carrying value. Two vehicles at 0.85x reported NAV can be at 0.80x and 1.05x adjusted.
**Mark observability is a valuation input, not a footnote.** Agency MBS are level 1/2 with deep, screen-quoted markets — the book value is close to a fact. Non-agency, whole loans, bridge loans and BDC private credit are level 3, valued by model or by a third-party valuation firm engaged by the manager. Report the level-1/2/3 split explicitly and apply a wider valuation range to level-3-heavy books. A vehicle whose marks barely moved through a quarter when public credit spreads widened 100 bps is telling you the marks are stale, not that the portfolio is resilient.
**Dividend discount models — appropriate here, with care.** These are pass-through vehicles by construction (US REITs and RICs must distribute the large majority of taxable income), so a DDM is one of the few cash-flow models that genuinely fits. Use a *sustainable* dividend — coverage-adjusted, with the fee waiver removed and PIK income excluded — not the trailing declared one.
**Lessors — NAV / replacement-value based.** Build a fleet NAV: independent appraised or market value of the equipment by type and vintage, less net debt, less lease-liability and maintenance obligations, plus the value of the contracted lease book. Cross-check against price-to-tangible book, and against the private-market transaction values for comparable assets (aircraft and container sale-leaseback pricing is observable). Replacement cost matters because new-build cost sets the ceiling on lease rates: when equipment prices rise, an existing fleet is worth more and lease rates follow. For lessors with a genuine operating-services business bolted on, value that separately on operating metrics and sum the parts.
**Cross-checks.** Economic return on book over 3, 5 and 10 years versus the sector — the cleanest management scorecard available. Dividend coverage by EAD or NII. Implied credit loss: solve for the cumulative portfolio loss rate the current price implies and test it against the vehicle's own worst historical experience. For lessors, implied residual value: solve for the disposal proceeds assumption embedded in today's price.
**Do not use:** EV/EBITDA, EV/Sales, EV/EBIT, net debt or Debt/EBITDA, FCF yield, any WACC-discounted enterprise cash-flow model, P/E on reported GAAP EPS, PEG, and — for mortgage REITs — FFO, AFFO, cap rates, NOI multiples or any per-square-foot measure.
## Peer set construction
A valid comparable shares the **asset class, funding model, leverage regime and management structure**. The label "REIT", "credit fund" or "leasing company" is not a peer set.
**Splits that must never be mixed:**
- **Mortgage REITs vs equity REITs.** Different assets, different risks, different metrics, different multiples. This is the primary mis-route this playbook exists to prevent. They share only a tax regime.
- **Agency mREITs vs credit/hybrid mREITs vs commercial mREITs.** Agency: 6–9x leverage, no credit risk, rate and prepayment risk. Credit/hybrid: 1–4x, real default risk, wider haircuts. Commercial: originating balance-sheet loans against buildings, CECL reserving, office and construction concentration. Three distinct businesses; a common P/B scatter across them is meaningless.
- **BDCs vs equity/venture asset managers vs banks.** A BDC deploys its own balance sheet; a manager earns fees on other people's capital and belongs in `holdco-assetmgr.md`.
- **Internally vs externally managed vehicles.** The fee drag is worth a large and persistent multiple gap. Never explain that gap away as "cheapness".
- **Upper-middle-market vs lower-middle-market BDCs.** Borrower EBITDA size drives loss severity, covenant quality and yield; a $100m-EBITDA-borrower book and a $15m one have different loss distributions at the same stated yield.
- **First-lien senior-secured BDCs vs junior/mezzanine/CLO-equity vehicles.** Same wrapper, opposite position in the waterfall.
- **Long-lived asset lessors (aircraft, rail, containers, marine) vs short-cycle rental (general equipment, construction, industrial tools).** Long-lived: 10–25 year assets, contracted lease terms, residual bets over decades, capital-markets funded. Short-cycle: 3–8 year assets, day/week rate exposure, utilisation-driven, priced off construction activity. Different volatility, different multiples.
- **Operating lessors vs finance lessors.** An operating lessor holds residual risk on its balance sheet; a finance lessor has effectively made a loan and belongs closer to `nbfc.md`. The accounting differs (Ind-AS 116 / IFRS 16 / ASC 842) and so does the risk.
- **Listed BDC vs non-traded/perpetual private BDC.** Non-traded vehicles have different fee structures, redemption gates and NAV-striking processes; their marks are not disciplined by a traded price.
**Also align:** rate and credit regime (a 2021 comparison against a 2024 one compares different worlds); accounting regime (US GAAP with CECL vs IFRS 9 vs Ind-AS); fiscal year end; leverage policy as *stated by management*, not just as observed; hedging philosophy; and size, since scale drives repo access, unsecured bond market access and funding cost directly.
Aim for 5–8 peers. State the basis explicitly, and benchmark every metric twice — against peers *and* against the company's own record through at least one full rate or credit cycle.
## Sector-specific red flags
- **Dividend maintained while book value per share erodes.** The defining pathology of the sector. A stable dividend and a falling BVPS means capital is being returned and called income. Compute economic return on book: if it is below the dividend yield, the payout is partially a liquidation.
- **Equity issued below book value or NAV.** Immediately and permanently dilutive to BVPS, and the standard behaviour of externally managed vehicles paid on gross assets. Check every ATM programme and follow-on against the prevailing P/B. US BDCs generally require shareholder approval to issue below NAV — watch for that proposal appearing on the proxy, it is a statement about management's intent.
- **Rising PIK share of investment income (BDC).** Booked but uncollected. Separate amendment PIK from structural PIK; a jump in amendment PIK is borrowers being restructured quietly. It flatters NII and dividend coverage precisely while credit deteriorates.
- **Non-accruals disclosed only at fair value, or the definition changing.** Demand both cost and fair-value measures. A widening gap is severity already taken; a narrowing gap with rising cost-basis non-accruals is marks that have not caught up.
- **Restructurings that reset the clock.** Amend-and-extend, covenant holidays, converting cash interest to PIK, or moving a loan into a joint venture or unconsolidated vehicle so it leaves the non-accrual line. The private-credit analogue of bank evergreening.
- **Non-GAAP earnings definitions that change, or that exclude recurring items.** "Distributable earnings", "core EPS" and "earnings available for distribution" are company-defined. If the definition changed in the period where coverage would otherwise have broken, that is the finding. Rebuild it yourself from the income statement every time.
- **Falling hedge ratio, lengthening duration gap, or a hedge book concentrated in short-dated instruments rolled quarterly.** Management taking an undisclosed rate view with shareholder book value. The disclosed rate-shock table is the check.
- **Repo concentration, shortening tenor, or rising haircuts.** Fewer counterparties, more overnight funding, more encumbered assets, shrinking unrestricted cash. This is the pre-margin-call configuration; it develops over weeks, not years.
- **Leverage rising into a spread-widening environment.** Buying more of an asset whose price is falling, funded by borrowings whose haircuts are rising, is how forced sellers are created.
- **Level-3 marks that do not move when observable markets do.** Compare the reported mark trajectory against public credit indices or agency MBS spreads. Stale marks are a solvency question, not a presentation quibble.
- **Third-party valuation coverage narrowed.** BDCs disclose what proportion of the portfolio was reviewed by an independent valuation firm. A falling proportion, or a change of valuation provider, deserves an explanation.
- **Fee waivers supporting dividend coverage, with a stated expiry.** Model the post-waiver quarter explicitly. Also watch incentive fees accrued on unrealised gains that later reverse without clawback.
- **Realised losses persistently exceeding unrealised write-ups (BDC).** The marks were optimistic and are being validated downward at exit. Track cumulative realised vs unrealised over 5 years — this is the single best test of whether a manager's NAV is honest.
- **Lessor: depreciation policy changes, extended useful lives, raised residual assumptions.** Boosts current profit and defers a loss to disposal. Also watch for impairments concentrated in one "clean-up" year.
- **Lessor: disposal gains carrying an outsized share of pre-tax profit,** or a fleet being sold faster than it is replaced. Earnings from shrinking the earning base.
- **Lessor: high time utilisation with falling dollar utilisation,** or lease renewals repeatedly at rates below expiring ones. Volume held by price concession.
- **Concentration.** Top-5 lessee, top-5 borrower or top-3 repo counterparty concentration. In this sector one counterparty failure is a capital event, not a bad quarter.
- **Related-party and affiliate transactions.** Assets bought from or sold to an affiliate of the external manager, co-investments alongside manager-affiliated funds, or servicing paid to a related entity. Read the related-party note in full, every year.
- **A structurally wide discount to book that never closes.** Often correct: the market is saying it does not believe the NAV, or that the fee load consumes the return. Do not treat persistent 0.6–0.7x P/B as automatic upside without identifying what would change.
## Cycle and structural context
**Rates drive mREITs; credit drives BDCs; capex and asset prices drive lessors.** Get the dominant driver right before scoring anything else.
**The rate cycle.** Rising short rates lift repo cost immediately while asset yields reprice slowly, compressing spread; the hedge book is what determines whether book value survives. Curve shape matters directly — an inverted curve is structurally hostile to a borrow-short-lend-long model, and steepening is the sector's classic recovery trade. Falling rates trigger refinancing waves: premium MBS prepay away and reinvestment happens at lower yields, so a rate cut is not unambiguously good news. Spread widening independent of rates (2020, 2022) hits book value even in a fully rate-hedged agency book, which is why spread duration must be tracked separately from rate duration.
**The credit cycle.** Private credit deteriorates with a long lag and in a specific sequence: borrower interest coverage falls, then amendment PIK rises, then watchlist ratings migrate, then non-accruals appear, then realised losses print. By the time non-accruals are visible the underwriting decision was made two to four years earlier. Floating-rate BDC books gain yield when rates rise but simultaneously squeeze borrower coverage — the yield benefit and the credit damage arrive from the same event, and analysts routinely score only the first.
**The equipment cycle.** Utilisation and lease rates are pro-cyclical and track construction, freight, air traffic or industrial capex. New-build supply is the swing variable and has long lead times, so gluts persist. Residual values are correlated with new equipment prices, so an inflationary period lifts used values and flatters disposal gains — a transient benefit that reverses. Sale-leaseback volume is a useful read on how tight financing conditions are for the lessee industry.
**Funding-market access is the survival variable in all three.** These vehicles cannot function without continuous access to repo, ABS, revolvers or unsecured bonds. Every crisis in this sector — 2008, March 2020, the 2022 UK LDI episode's second-order effects — was a funding event that became a solvency event through forced selling. Score unencumbered liquidity, undrawn committed facilities and the maturity ladder as risk metrics of first order.
**Structural shifts to weigh explicitly.** Private credit's decade-long expansion has compressed spreads and loosened documentation (covenant-lite is now the norm in the upper middle market), so today's yields are being earned on weaker structures than a decade ago — the loss experience of the last cycle is not a good guide. Non-traded perpetual BDCs have absorbed enormous retail inflows, changing both competition for deals and redemption dynamics. Bank retrenchment under Basel III endgame keeps pushing assets toward non-bank balance sheets. For mREITs, the Fed's balance-sheet trajectory sets agency MBS spreads directly. For lessors, fleet electrification and emissions rules create a new class of technological obsolescence risk in residual values.
## India vs global notes
The mortgage-REIT and BDC structures are essentially **absent from India**; the lessor and specialty-finance families are present but sit inside a different regulatory frame. Do not assume an Indian analogue exists — check the actual structure first.
| Dimension | India | US / global |
|---|---|---|
| Mortgage REIT equivalent | No listed mREIT structure. Indian REIT regulations (SEBI REIT Regulations, 2014) require at least 80% of value in **completed, rent-generating property** — a mortgage-securities vehicle cannot qualify. The closest analogues are HFCs, NBFC-MFIs and securitisation vehicles under RBI rules, which route to `nbfc.md`. RMBS is nascent; NHB refinance and direct assignment / PTC pools dominate | US mREITs elect REIT status under the Internal Revenue Code, distribute ≥90% of taxable income, file 10-K/10-Q with the SEC, and disclose repo, haircuts, duration gap and CPR in quantitative detail |
| BDC equivalent | No BDC regime. Private credit reaches borrowers through Category II AIFs, NBFC-ICCs, ARCs and special-situation funds — mostly unlisted, so NAV, non-accruals and PIK disclosure are far thinner. Where a listed vehicle exists, treat it as an NBFC with a concentrated book | BDCs are regulated under the Investment Company Act of 1940, taxed as RICs, statutorily capped near 2:1 leverage (150% asset coverage after the 2018 change), must value the portfolio at fair value quarterly under board oversight, and file detailed schedules of investments naming every position |
| Lessor / rental regime | Equipment financing largely through NBFC-AFCs (asset finance companies) under RBI scale-based regulation; growing construction-equipment and industrial rental, and a large aircraft-leasing push through **GIFT City IFSC** with IFSCA as regulator and tax incentives. Ind-AS 116 governs lease accounting | Aircraft, container, railcar and general-equipment lessors under IFRS 16 / ASC 842; Ireland and Singapore are the traditional aircraft-leasing hubs; appraised fleet values from recognised appraisers are a market standard |
| Accounting | Ind-AS (IFRS-converged): Ind-AS 109 for ECL on financial assets, Ind-AS 116 for leases, Ind-AS 113 fair-value hierarchy. RBI provisioning norms run alongside Ind-AS for NBFCs, and the higher of the two applies via an impairment reserve | US GAAP with CECL for loans; ASC 820 fair-value hierarchy; fair-value option common in mREITs; IFRS 9 ECL elsewhere |
| Regulators | RBI (NBFCs, securitisation, ARC), SEBI (REITs/InvITs, AIFs, listing and disclosure), IFSCA (GIFT City leasing and finance) | SEC (registration, 10-K, proxy, closed-end fund rules), Fed/FHFA (agency MBS market structure), FINRA for distribution |
| Units and reporting | ₹ crore and lakh; fiscal year April–March; quarterly results plus investor presentation and concall — the concall is often the only source for utilisation, lease-rate and residual commentary | $ millions; calendar year common; 10-K/10-Q plus supplemental packages containing CPR, duration, haircuts, schedule of investments, non-accrual and PIK detail |
| Ownership | Promoter holding, pledged-share disclosure and group-company exposure are central governance checks; related-party lending within a promoter group is a recurring failure mode | Widely held; the external manager relationship (and its ownership) is the equivalent governance focus |
| Distribution rules | InvITs/REITs must distribute ≥90% of net distributable cash flows (SEBI); NBFCs have no such mandate and dividends are subject to RBI's dividend-payout criteria linked to CRAR and NPA levels | US REITs ≥90% of taxable income; RICs (including BDCs) ≥90% to avoid entity-level tax, with a separate 4% excise-tax calculation that drives spillover management |
**India-specific checks:** for an Indian leasing or asset-finance company, confirm whether it is an NBFC-AFC under RBI scale-based regulation (and therefore subject to RBI provisioning norms, CRAR requirements and dividend-payout eligibility criteria), or a pure operating-lease company outside it — the constraint set differs entirely. Check the Ind-AS 109 impairment reserve reconciliation. For GIFT City aircraft-leasing entities, understand which tax exemption is being relied on and its sunset date, because a large part of the reported return may be a time-limited fiscal benefit rather than operating economics.
## Checklist
- [ ] Confirm the family: mortgage REIT, BDC/private credit, or lessor — and confirm it is *not* an equity REIT, a deposit-taker or a fee-based asset manager. Route away if so.
- [ ] For any mortgage REIT, state explicitly in the report that occupancy, WALE, same-store NOI, cap rates, FFO and AFFO are inapplicable, and why.
- [ ] Delete EV/EBITDA, net debt, D/E-as-risk-score, ROCE, FCF yield, current ratio and reported P/E from the analysis; say why.
- [ ] Compute economic return on book: (change in BVPS + dividends per share) ÷ opening BVPS, for each of the last 5–10 years.
- [ ] Track BVPS/NAV per share, not aggregate equity; check every equity issuance against the prevailing P/B or P/NAV.
- [ ] Rebuild dividend coverage yourself: EAD or core EPS (mREIT) / NII per share (BDC) ÷ dividend per share. Compare to the company's own non-GAAP definition and note any change to that definition.
- [ ] Check the tax character of distributions — how much is return of capital?
- [ ] mREIT: measure stated *and* economic leverage (include TBA notional); read the disclosed ±100 bp rate-shock table; compute the duration gap and hedge ratio.
- [ ] mREIT: map repo tenor, counterparty concentration, haircuts by collateral type, and unencumbered assets as % of equity.
- [ ] mREIT: check CPR against the portfolio's carrying premium/discount; assess extension risk and prepayment protection actually delivered.
- [ ] mREIT: split agency vs credit exposure and never compare across that line.
- [ ] BDC: take non-accruals at **both** cost and fair value; track the gap and the trend.
- [ ] BDC: compute PIK as % of total investment income; split structural from amendment PIK.
- [ ] BDC: check first-lien share, borrower EBITDA size, borrower interest coverage, portfolio yield vs cost of debt, and leverage against the statutory cap.
- [ ] BDC: compare cumulative realised losses against cumulative unrealised write-ups over 5 years.
- [ ] BDC/mREIT: quantify the fee load as % of net assets; identify any fee waiver and model its expiry; check internal vs external management.
- [ ] Check spillover / undistributed taxable income and whether it is being depleted.
- [ ] Report the level-1/2/3 fair-value split and test whether level-3 marks moved with observable markets.
- [ ] Lessor: fleet size, age vs economic life, time utilisation *and* dollar utilisation, lease yield, average remaining lease term.
- [ ] Lessor: 3–5 year history of disposal proceeds vs net book value; disposal gains as % of pre-tax profit; any change to useful life or residual assumptions.
- [ ] Lessor: split maintenance capex from growth capex; confirm the fleet is being replaced, not run down.
- [ ] Lessor: operating vs finance lease mix and the accounting treatment; lessee concentration and credit loss experience.
- [ ] All: map the debt maturity ladder against asset cash flows; check encumbrance, undrawn committed facilities and unsecured market access.
- [ ] Value on P/B vs sustainable ROE (mREIT, BDC) or fleet NAV / replacement value (lessor); re-strike an adjusted book with your own marks.
- [ ] Run a DDM on a *sustainable* dividend, and a reverse test: what loss rate or residual value does today's price imply?
- [ ] State where in the rate, credit and equipment cycle this sits, and what funding-market conditions are assumed.
- [ ] Peer set: same asset family, leverage regime, capital-structure seniority, management structure and accounting regime — stated explicitly.
- [ ] India: confirm the regulatory wrapper (NBFC-AFC under RBI, AIF under SEBI, IFSC entity under IFSCA) and whether any reported return is a time-limited tax benefit.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| add-product-e2e-evalpaperclipai/paperclipAdd or extend a Paperclip full-stack runner E2E workflow, fixture, matcher, or report evidence path for local or Daytona execution. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| add-runner-evalpaperclipai/paperclipAdd or extend a Paperclip Runner protocol evaluation definition, roster, assertion, or report fixture with provenance and narrow validation. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| diagnose-why-work-stoppedpaperclipai/paperclipCompatibility entry point for failed, stalled, looping, or over-recovered Paperclip work. Use fix-failed-paperclip-run for the full diagnosis, regression, repair, real verification, and green-PR workflow; honor an explicitly requested diagnosis-only scope. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| fix-failed-paperclip-runpaperclipai/paperclipDiagnose and fix failed, stuck, slow, or incorrectly blocked Paperclip tasks and runs. Use with a task/run URL or failure report to trace the root cause, prove a regression red then green, check second-order effects, verify the real workflow, and deliver a PR with passing checks. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |