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/aviation-hotels.md
View on GitHub# Airlines, hotels, travel, restaurants and OTAs — sector playbook
Use this when: the company sells a perishable, time-bound unit of capacity or brokers someone else's — scheduled airlines and their MRO/cargo/loyalty arms, hotel owners and operators, restaurant and QSR chains, online travel agencies and travel aggregators, tour operators, cruise lines, airport and ground-handling concessions, and multiplex/leisure operators that share the same fixed-cost-plus-footfall economics.
Four structural facts govern everything below. **The unit of production expires** — an unsold seat or room-night at departure or midnight is gone forever, so pricing is dynamic, marginal cost is near zero and operating leverage is extreme in both directions. **Lease accounting has destroyed margin and leverage comparability** since IFRS 16 / Ind-AS 116, splitting the peer group into owners and lessees that no longer share a margin definition. **The dominant cost line is exogenous** — fuel for airlines, the RevPAR cycle for hotels — so margin is mostly not a skill signal. And **customers pay before they consume**, which inverts the working-capital and liquidity ratios in the generic checklist. Replace the ratio set; do not caveat it.
## 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
**OPM / EBITDA margin — not comparable across companies, across periods, or against itself.** Since Ind-AS 116 (India, FY20 onward) and IFRS 16, operating lease rentals moved out of opex into depreciation plus finance cost. For a lessee-heavy carrier running a sale-and-leaseback fleet this mechanically lifted reported EBITDA margin by roughly 10–18 percentage points overnight with zero change in economics, while loading a large lease liability onto the balance sheet. So the margin series breaks pre- vs post-FY20; an owned-fleet legacy carrier is not comparable to a sale-and-leaseback LCC; and a hotel *owner* is not comparable to an operator running the same assets on long leases. The only margin that compares them is **EBITDAR** — before rent.
**OPM is also not a skill signal even within one company.** Fuel is roughly 30–45% of an airline's operating cost, worse in India where ATF carries state VAT of ~1–30% on top of central duties. A quarter in which OPM doubles almost always means crude fell or the currency strengthened, not that management improved. Hotels sit at the opposite pole: 55–70% of cost is fixed (payroll, energy, property tax, depreciation), so margin is close to a pure function of where RevPAR sits in the cycle. Ranking companies on OPM alone — the classic single-metric screen failure — will systematically rank the luckiest, not the best.
**D/E — inflated, undefined, or actively inverted.** Capitalised leases inflate the numerator for lessees. The denominator is frequently *negative*: carriers with accumulated losses and asset-light operators after years of buybacks both run negative net worth, and a negative denominator produces a negative D/E that screens as "no debt". This is the single most dangerous artefact in the sector. Use **adjusted net debt / EBITDAR** and nothing else.
**P/E and EPS — a cyclical trap sitting on top of non-operating noise.** These are high-operating-leverage cyclicals: earnings are negative for years at a time (the entire industry, FY21–FY22), so P/E is undefined; and at the cycle peak trailing P/E looks cheapest precisely when the stock should be sold. Worse, reported PAT routinely contains large non-cash and non-recurring items — mark-to-market on USD-denominated lease liabilities and maintenance provisions (a single currency move can swing a large carrier's PAT by hundreds of crore with no cash effect), sale-and-leaseback gains, OEM compensation for grounded aircraft, asset-sale gains at hotel companies, and deferred-tax-asset recognition on accumulated losses. P/E, ROE and EPS growth built on that PAT are noise, not signal.
**ROCE / ROE / P/B — distorted in both directions at once.** Hotel real estate sits at historical cost, often land acquired decades ago, so owned-portfolio book value can be a fraction of market value: P/B looks expensive and ROCE looks poor when the true asset value is a multiple of book. At the other extreme, asset-light operators and buyback-heavy carriers have small or negative equity, making ROE and ROCE infinite, negative or absurd. OTAs have essentially no operating assets, so ROCE is a meaningless number.
**Current ratio — inverted.** Airlines, hotels, OTAs and tour operators collect cash before delivering service, so unearned ticket revenue, advance deposits, vouchers and loyalty deferrals sit in current liabilities by design. A structurally healthy carrier runs a current ratio well below 1.0 — roughly 0.6–0.9 in good years. Applying the textbook ">1.5" rule flags the strongest operator in the peer set as distressed.
**FCF and cash conversion — flattered on the way up, collapsing on the way down.** The same negative working-capital cycle means a rapidly growing airline or OTA shows superb operating cash flow purely from float on advance bookings, even while destroying economic value; a shrinking one bleeds cash as that float unwinds, so FCF deteriorates fastest exactly when bookings decline. Add lumpy pre-delivery payments on aircraft orders, sale-and-leaseback proceeds parked in investing cash flow, and hotel renovation capex that is genuinely maintenance but is presented as growth, and headline FCF is close to uninterpretable without adjustment. Strip working-capital movement out of OCF before calling any of it free.
**P/S — misleading for OTAs specifically.** Platforms report *net* revenue (take rate) or *gross* revenue depending on principal-versus-agent conclusions under Ind-AS 115 / IFRS 15 / ASC 606. A 5x EV/Sales on net revenue can be 0.4x on gross bookings. Reconstruct GBV and take rate before any cross-company sales multiple.
**Any 3–5 year CAGR spanning FY20–FY22 is arithmetic garbage** — the pandemic zeroed the base, so a two-year revenue CAGR off FY21 can exceed 100% and means nothing. Anchor every growth series to FY19 or FY20 as the pre-COVID comparator and say so explicitly in the output.
## The metrics that actually matter
Ranges are **indicative only**. They vary by market, sub-sector, cycle stage, fuel and rate environment, and 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 multi-year history overrides every absolute band below. Tags: **[A]** airlines, **[H]** hotels, **[O]** OTAs and platforms, **[All]** cross-sector.
| Metric | Definition / how to compute | Indicative healthy range | Why it matters |
|---|---|---|---|
| **RASK, CASK and CASK ex-fuel [A]** | Revenue and cost per Available Seat Kilometre — the unit revenue and unit cost of flying one seat one km. CASK ex-fuel strips the exogenous line. The **RASK-minus-CASK spread** is the profit engine. | Positive spread through a full cycle. Efficient Indian narrowbody LCCs have run RASK around ₹4.3–5.0 and CASK ₹4.0–4.6 per ASK in normal years, CASK ex-fuel ~₹2.6–3.2. US majors: RASM ~15–18 US cents, CASM-ex ~10–12 cents. The global ultra-low-cost floor is around EUR 0.035 CASK ex-fuel. | The only structural, cycle-neutral measure of airline competitiveness — airlines compete on cost per seat, not on margin. Track CASK ex-fuel over 8–12 quarters: falling with scale (denser configuration, higher utilisation, single fleet type) means genuine compounding advantage; drifting up means the advantage is eroding regardless of what OPM shows in a cheap-fuel quarter. |
| **Passenger load factor and yield, read together [A]** | PLF = RPK ÷ ASK (share of available seats sold). Yield = passenger revenue ÷ RPK (average fare per km). Their product drives passenger RASK. | Indian domestic LCC 82–88% sustained; full-service and long-haul 78–84%; US/EU majors 82–86%. Below ~75% for a narrowbody LCC signals overcapacity or route errors; above ~92% usually means fares were left on the table. | PLF alone is a vanity metric — any airline can fill seats by cutting fares. The diagnostic is the trade-off: rising PLF with falling yield means the carrier is buying traffic; rising PLF with flat-to-rising yield is genuine demand. Watch PLF against industry ASK growth — when capacity growth outruns passenger growth, **yields break before load factors do**. |
| **Aircraft utilisation and fleet composition [A]** | Revenue block hours per aircraft per day, plus fleet age, number of aircraft types, owned-vs-leased split, and grounded/AOG count. | Indian narrowbody LCC 11–13 block hours/day is strong; below 10 indicates groundings, MRO bottlenecks or network problems. Global LCC 11–12; long-haul widebody 13–16. One to two fleet types is optimal; four or more is a structural cost penalty. | Utilisation is the single biggest lever on CASK — a grounded aircraft carries lease rent, insurance and crew cost with zero ASK. Engine-reliability episodes that ground a large share of a fleet are the mechanism by which viable Indian carriers have become insolvent. Always reconcile balance-sheet fleet count against aircraft actually in revenue service in the investor deck. |
| **Ancillary revenue per passenger and mix [A]** | Non-ticket revenue per departing passenger — seat selection, baggage, cargo/belly freight, onboard sales, change fees, co-brand card and loyalty monetisation — and its share of total revenue. | Global LCC leaders USD 20–45 per passenger, 25–35%+ of revenue. Indian carriers materially lower, roughly ₹400–900 per passenger and ~12–18% of revenue, capped by unbundling rules and price sensitivity; cargo is a bigger swing factor in India. | High-margin, far less price-elastic than base fares, and structurally rising — the main reason two carriers with identical yields have very different profitability, and the most durable earnings stream through a fare war. In developed markets, loyalty cash flows (co-brand card sales to banks) can be worth more than the flying operation itself. |
| **Fuel exposure: fuel cost per ASK, hedge ratio, pass-through lag [A]** | Fuel cost ÷ ASK; percentage of next 12 months' consumption hedged and at what structure; historical lag between a crude/ATF move and the fare response; sensitivity of PAT to a USD 10/bbl and ₹1/USD move. | Indian carriers hedge little or not at all (no deep ATF hedging market) — assume near-full spot exposure. Global carriers hedge 0–60%. Fare pass-through typically lags 1–2 quarters, longer where capacity is oversupplied. | Isolates the exogenous component of the margin so the remainder can be attributed to management. Also the correct way to build the mid-cycle case: normalise fuel and currency to a long-run level, then ask whether the airline still clears its cost of capital. A carrier whose profitability exists only in a cheap-fuel window is not a compounder. |
| **EBITDAR and EBITDAR margin [A][H]** | Earnings before interest, tax, depreciation, amortisation **and rent/lease expense**. Rebuild by adding lease-related depreciation and lease finance cost back to reported EBIT, or by using pre-Ind-AS-116 rentals where disclosed. | Airlines 18–25% is strong for an efficient LCC in a normal fuel environment; sustained below ~10% through a cycle indicates structural unprofitability. Hotels 30–40% at corporate level for Indian branded owners in an up-cycle; asset-light fee businesses 60–75%. | The **only** margin that compares a company leasing its assets with one owning them — the central comparability problem in this sector. It is also what lessors, rating agencies, credit committees and M&A buyers actually underwrite against. Use EBITDAR margin, never OPM, for cross-company ranking. |
| **Adjusted net debt / EBITDAR and liquidity months [A][H]** | (Gross debt + capitalised lease liability + maintenance/redelivery provisions + unfunded pension − cash and liquid investments) ÷ EBITDAR. Where leases are not capitalised, proxy the liability at 7–8x annual rent. Liquidity months = (cash + undrawn lines) ÷ average monthly cash operating cost. | Airlines: below ~3.5x is investment-grade-like, 4–5x stretched, above ~6x pre-distress. Cash at least 15–25% of trailing-twelve-month revenue, or 4–6 months of operating cost. Hotels: net debt/EBITDA below 3x for owners, below 1x for asset-light operators. | Replaces D/E entirely, which is undefined or perverse here. It is also the ratio that actually predicts failure: every large airline collapse of the last two decades showed deteriorating adjusted leverage and collapsing liquidity months well before the reported loss made headlines. An airline with under two months of liquidity is a going-concern question, not a valuation question. |
| **RevPAR, decomposed into ARR and occupancy [H]** | Revenue per Available Room = Average Room Rate × Occupancy, computed on **all** available room-nights including unsold ones. Always demand like-for-like/same-store. | Indian branded occupancy of 65–72% is a healthy market; luxury ARR roughly ₹12,000–20,000+, upscale ₹6,000–9,000, mid-scale/economy ₹3,000–5,000. US upper-upscale occupancy 70–75%. Market occupancy above ~75% signals ARR has room to run. | The sector's fundamental unit metric, but the **decomposition** is the analysis. RevPAR growth driven by ARR flows to EBITDA at 60–80% incremental margin because the room is already staffed and lit; growth driven by occupancy flows at only 40–55% because of variable cost per occupied room. Identical RevPAR growth, very different earnings outcomes. Reported growth including newly opened or acquired keys tells you nothing about pricing power. |
| **RevPAR index: RGI, MPI, ARI vs competitive set [H]** | Market-share indices (STR/HotStats): RGI = own RevPAR ÷ comp-set RevPAR × 100; MPI is the occupancy version, ARI the rate version. | RGI above 100 means more than fair share; 110–125 is a genuinely strong brand or asset. Sustained sub-100 for a luxury asset indicates brand or product decay. | Separates cycle from skill. A post-shock demand-supply squeeze lifts every operator's RevPAR and flatters all of them equally; RGI strips the cycle and shows actual outperformance. It is the hotel equivalent of alpha versus beta, and it is what owners use when deciding whether to renew or fire a management company. |
| **Demand growth vs forward supply pipeline, at micro-market level [A][H]** | Forward branded room supply CAGR against expected demand CAGR **by city/micro-market**, not nationally; and the company's own pipeline as a percentage of operating keys. For airlines, the equivalent is industry order book and ASK growth against passenger growth. | India has run branded supply at ~6–8% CAGR against demand at ~9–11% — a favourable window. Any market where supply growth exceeds demand growth for two consecutive years is entering a pricing downcycle. Company pipeline of 25–50% of operating keys is healthy; above ~80% is execution risk. | Profitability here is dominated by the industry supply-demand balance, not by company execution — a well-run hotel in an oversupplied market earns less than a mediocre one in a tight market. Because a hotel takes 3–5 years to build and aircraft order books are visible years ahead, **supply is one of the few genuinely forecastable variables in equity analysis**, and it appears nowhere on a generic checklist. |
| **GOPPAR, GOP margin and incremental flow-through [H]** | Gross Operating Profit per Available Room — property-level profit after departmental and undistributed costs, before rent, property tax, insurance, D&A and management fees. Flow-through = incremental EBITDA ÷ incremental revenue. | GOP margin 38–48% at property level is strong for Indian owned hotels; 30–35% average. Flow-through of 55–70% on incremental revenue in an up-cycle is the benchmark; below ~45% means cost inflation is eating the cycle. | GOPPAR is the true measure of operating skill because it isolates what the *operator* controls (rate, mix, labour productivity, energy) from what the *owner* controls (capital structure, rent, land cost). Flow-through is the operating-leverage test: converting under half of incremental revenue into EBITDA in a rising market means a cost or wage problem that will be brutally exposed when RevPAR turns. |
| **Asset-light mix: managed/franchised keys and fee revenue share [H]** | Share of total keys managed or franchised rather than owned/leased; management and franchise fee income as a % of revenue **and** of EBITDA. Fees are typically a base fee of 2–4% of revenue plus an incentive fee of 6–10% of GOP. | Global brand operators are 95%+ asset-light. Indian players are mid-transition: fee income at 15–25% of revenue but 25–40% of EBITDA is typical, with managed keys becoming the majority of the pipeline. | Determines the multiple the market will pay, and explains why consolidated ROCE fails here. Fee income needs near-zero capital, is contractually sticky (10–30 year contracts) and is far less cyclical than owned EBITDA — it deserves a 25–35x EBITDA multiple where owned-hotel EBITDA deserves 12–16x. A company with flat consolidated ROCE can be radically improving its economics if the mix is shifting. **Analyse owned and managed segments separately or you will misprice the business.** |
| **Non-room revenue mix: F&B, banqueting, weddings, MICE [H]** | Share of hotel revenue from food and beverage, banqueting, weddings and other non-room sources; ideally also per available room. | Indian hotels: 40–50% of revenue from F&B and banqueting is normal and structural. Global upscale: 25–30%. Indian wedding and MICE concentration makes Oct–Dec and Jan–Mar seasonally dominant. | A genuine India-versus-developed-market divergence that breaks cross-border comparison. Indian hotel companies will always look lower-margin than US peers because F&B carries 20–30% margins against 70%+ for rooms — but that F&B business is real-estate-light, brand-driven and annuity-like. Judging an Indian owner-operator on a global brand manager's margin structure is a category error. |
| **Deferred revenue / forward bookings and the negative working-capital cycle [All]** | Unearned ticket revenue, advance room deposits, loyalty deferrals and travel credits/vouchers held as current liabilities — tracked as days of revenue, and as a trend **against the cash balance**. | A healthy growing airline runs materially negative working capital with forward sales of 30–60 days of revenue and a current ratio of 0.6–0.9. Critical test: cash grows at least in line with deferred revenue. | The sector's most misread line. Rising unearned revenue is normally a leading demand indicator — but rising unearned revenue *with falling cash* means customer float is funding operating losses, the precise failure signature of every large airline and tour-operator collapse. It is also why FCF must be adjusted. In a downturn the float unwinds while refunds come due simultaneously, turning a P&L problem into insolvency inside one quarter. |
| **OTA economics: gross bookings, take rate, contribution margin after marketing [O]** | Gross Booking Value through the platform; take rate = net revenue ÷ GBV; contribution margin after direct marketing and customer incentives. Plus repeat-customer rate, direct/app traffic share vs paid, and CAC payback. | Take rates 10–16% for hotel-led models, 4–8% for air-heavy models (air commissions are thin). Adjusted EBITDA margin on net revenue of 20–30% is good for an Indian OTA; global leaders run 35%+. Direct/organic traffic above ~60% is a durable moat; below ~40% means demand is being rented from search engines. | P/S on reported net revenue is meaningless because revenue definitions differ by company under principal-versus-agent rules. GBV plus take rate makes them comparable and reveals whether growth is volume or mix. The marketing line is the real test: an OTA growing GBV 30% while marketing grows 45% has no business model, only a subsidy — and discount-led share wars have repeatedly destroyed Indian OTA profitability. |
| **Lease-adjusted ROIC vs WACC, cycle-averaged [All]** | NOPAT ÷ (net fixed assets + capitalised lease liability + working capital + capitalised maintenance reserves), averaged across **at least 7–10 years spanning one full up- and down-cycle**. | The global airline industry has historically averaged ROIC of ~4–7% against a WACC of 8–9% — structurally value-destroying, with only a handful of ultra-low-cost operators clearing the bar. Indian branded hotel owners earn 12–18% ROCE on owned assets in an up-cycle; asset-light fee businesses 40%+. Demand 10%+ mid-cycle lease-adjusted ROIC. | Single-year ROCE here tells you where you are in the cycle, not whether the business creates value — peak 25% and trough −10% average to capital destruction. Cycle-averaged, lease-adjusted ROIC is the honest test and explains why this sector has consumed enormous capital for negligible aggregate shareholder return. For Indian hotel owners, also present ROCE on **revalued** property, since historic-cost land overstates the reported figure relative to a new entrant's economics. |
### Restaurants and QSR chains — the additional metric set
Restaurant chains share the fixed-cost-plus-footfall structure but need their own KPIs; do not analyse them with hotel or airline metrics.
| Metric | Definition / how to compute | Indicative healthy range | Why it matters |
|---|---|---|---|
| **Same-store sales growth (SSSG) and its decomposition** | Sales growth for stores open through both comparable periods, split into **transactions (footfall) vs average ticket/price**. | Low-to-mid single digits is normal in a mature system; India expects mid-to-high single digits in a growth phase. Negative transaction growth masked by price increases is the warning state. | Total revenue growth in a chain is mostly new-store arithmetic and says nothing about brand health. SSSG driven by price is borrowed from the future — it eventually breaks footfall. This is the sector's like-for-like discipline, exactly parallel to same-store RevPAR. |
| **Average daily sales (ADS) per store and sales per square foot** | Gross store revenue ÷ operating days ÷ stores; and revenue per sq ft of trading area. | New-format stores should reach mature ADS within 12–18 months. Declining system ADS while store count grows is cannibalisation or format fatigue. | The direct productivity measure of the asset, and the input to store-level payback. Rising store count with falling ADS is value-destroying expansion. |
| **Restaurant operating margin (store-level EBITDA, pre-Ind-AS-116)** | Store revenue less COGS, store payroll, rent (on a pre-lease-capitalisation basis), utilities and store overheads — before corporate costs and D&A. | 15–22% is healthy for Indian QSR at store level; below ~12% leaves nothing after corporate cost. Report pre- and post-Ind-AS-116 to stay comparable. | Corporate EBITDA blends store economics with head-office leverage and, post-Ind-AS-116, with rent capitalisation. Store-level margin is the only clean read on unit economics — and rent as a % of sales (ideally 6–10%) is the line that kills chains in a weak-demand year. |
| **Capex per store and cash payback period** | Fit-out plus equipment plus pre-opening cost per new store, divided by annual store-level cash profit at maturity. | 2.5–4 years payback is good for Indian QSR; beyond ~5 years the growth is value-neutral at best. | Store rollout is the entire capital allocation decision in this sector. A chain expanding at a payback longer than its lease term or refurbishment cycle is compounding backwards, however fast revenue grows. |
| **Delivery / dine-in / takeaway mix and aggregator dependence** | Channel split of revenue, plus commission paid to delivery aggregators as a % of delivery revenue and own-app share of delivery orders. | Aggregator commissions of 18–25% of order value are typical in India. Own-app share above ~30% of delivery is a meaningful margin and data advantage. | Delivery revenue carries structurally lower margin and rents a customer relationship from a platform. A chain whose growth is entirely aggregator-led has weaker pricing power and a third party sitting between it and its customer — the restaurant analogue of an OTA's paid-traffic dependence. |
| **Store closures, churn and net new additions** | Gross openings, gross closures, and the closure rate as a % of the base; plus the age profile of the estate. | Closure rates above ~3–5% p.a. in a growth chain signal poor site selection. Watch for closures netted silently against openings. | Chains headline net additions. Gross closures reveal underwriting quality and the true survival curve of a cohort — and pre-empt the write-off cycle. |
| **Raw-material basket and menu pricing power** | COGS as a % of sales and the specific commodity basket (wheat, dairy/cheese, chicken, edible oil, coffee); the lag and magnitude of menu price increases. | Indian QSR gross margins run roughly 65–70%; a 200–300 bps COGS swing on a dairy or protein spike is normal. | The pass-through test: a brand that can raise menu prices without losing transactions has real equity; one that absorbs input inflation to protect footfall does not. This is measurable directly from SSSG decomposition. |
## How to value companies in this sector
**Airlines and asset-heavy hotels — EV/EBITDAR is the primary cross-company multiple.** Define enterprise value as market cap + gross debt + capitalised lease liability (or 7–8x annual rent where leases sit off balance sheet) + maintenance/redelivery provisions + unfunded pension − cash and liquid investments, over EBITDAR. This is the only construction that treats an owned aircraft and a leased aircraft identically, and it is what lessors, credit committees and M&A buyers use. Indicative bands: global LCCs 4–6x, legacy carriers 3–5x, best-in-class LCCs with structural cost advantage and dominant share have traded 6–9x. **P/E must never be the anchor multiple for an airline.**
**Hotels — run three methods and triangulate.**
1. *EV/EBITDA, computed separately on the owned/leased portfolio and on the fee business.* Indian branded owners have traded roughly 15–25x forward EV/EBITDA in an up-cycle against 10–14x in normal times; fee/management income justifiably carries 25–35x because it is capital-free and contract-protected. Blending the two into one multiple systematically misvalues any company mid-transition to asset-light — this is the core valuation argument in most Indian hospitality names today.
2. *EV per key versus replacement cost.* Value the portfolio at market development cost per key by segment — roughly ₹2.0–4.0 crore per key for luxury in India, ₹0.8–1.5 crore upscale, ₹0.4–0.7 crore mid-scale/economy; USD 300k–1m per key for US upper-upscale. If the market values the company below the cost of building the same portfolio, downside is anchored; well above, the market is capitalising the cycle. This is the sector's P/B substitute and it works precisely where P/B fails, because it uses replacement value rather than depreciated historic cost.
3. *SOTP / NAV.* Apply a cap rate to stabilised property NOI, then add separately valued fee income, land bank, commercial real-estate annuity and listed subsidiary stakes. Cap rates of 7–9% are typical for Indian hospitality assets, 5–7% in developed markets. This is the correct primary method for any owner whose book value bears no relationship to asset value. India has no hotel REIT regime comparable to the US, so FFO/AFFO is generally not used for Indian operators; for **US hotel REITs**, FFO/AFFO, EBITDA per key and implied cap rate remain the primary lens.
**OTAs and asset-light platforms** — use EV/gross bookings (typically 0.3–1.5x), EV/net revenue, and EV/adjusted EBITDA once profitable, cross-checked against a DCF driven by take rate, contribution margin and repeat rate. Never use ROCE or P/B. Because reporting differs under principal-versus-agent rules, the **GBV-plus-take-rate reconstruction is mandatory** before any comparison.
**Restaurant chains** — EV/EBITDA on a pre-Ind-AS-116 basis (or EV/EBITDAR) for comparability across owned, leased and franchised estates, cross-checked against a store-economics build: mature ADS × store-level margin × store count, less corporate cost, with new-store NPV at the observed payback. P/E is usable only for a mature, low-growth, franchised system.
**Apply mid-cycle normalisation to everything.** Because operating leverage is extreme, value on **mid-cycle EBITDAR** — an average over a full cycle, or an explicitly modelled normal-fuel, normal-RevPAR year — never on trailing twelve months. Then invert the classic reading of the multiple: a low trailing P/E or low EV/EBITDA at peak margins is a sell signal; a high or negative one at trough margins is often the entry point. Buying at 6x trailing EV/EBITDA on peak RevPAR or a peak fuel spread is the single most common way investors lose money in this sector.
**Value hidden assets separately.** Airline loyalty programmes, cargo divisions, MRO subsidiaries, ground-handling arms and owned airport slots are frequently worth more than the flying operation and are invisible in a consolidated multiple. Slot portfolios at capacity-constrained hubs, co-brand card contracts, and land under hotel properties all merit standalone valuation. Finally, sanity-check equity value against liquidity: for any airline with under six months of liquidity, the correct framework is a **probability-weighted scenario including dilution or restructuring**, not a multiple.
## Peer set construction
A valid comparable shares business model, asset ownership mode, lease treatment, cycle position and regulatory market. Get any one of these wrong and the table is worse than no table.
**Splits that must never be mixed:**
- **Airlines: LCC vs full-service vs regional vs charter/cargo.** Different cost bases, distribution, ancillary potential and yield curves. Also split **short-haul narrowbody vs long-haul widebody** — utilisation, turn times and load-factor norms differ structurally.
- **Airlines: owned-fleet vs sale-and-leaseback.** Even on EBITDAR these differ in residual-value risk and capital intensity; note the split explicitly alongside the multiple.
- **Hotels: owner vs lessee vs manager/franchisor.** Three different businesses — a property investor, a leveraged operating-lease bet, and a royalty stream. Never rank them on the same margin or ROCE. Companies mid-transition must be segment-split before comparison.
- **Hotels: by chain scale.** Luxury, upper-upscale, upscale, mid-scale, economy and long-stay have different ARR, occupancy stability, F&B share and cap rates. A luxury-heavy portfolio's lower occupancy is not a weakness.
- **Hotels: by micro-market.** Business-district, airport, leisure/resort and pilgrimage assets have different seasonality, weekday/weekend curves and supply pipelines. Aggregate national RevPAR conceals it.
- **OTAs: air-led vs hotel-led vs experiences vs corporate travel.** Take rates differ by 2–3x, so a blended peer table on take rate or EV/sales is meaningless. Split also by **principal vs agent** revenue recognition.
- **Restaurants: QSR vs casual dining vs cafés vs cloud kitchens**, and **franchisee-operator vs brand-owner**. An Indian master-franchisee pays royalties and cannot set the brand's strategy; it is not comparable to the brand owner it licenses from.
- **Cruise, tour operators and travel-adjacent** — do not fold into hotels or airlines; capital intensity and deposit float structures differ.
**Additional gates:** same lease-accounting regime and the same post-Ind-AS-116/IFRS-16 period; same cycle stage (never compare a company at trough RevPAR to a peer at peak); comparable state-tax and route-rights environment for airlines; and consistent consolidation treatment of JVs, associates and managed-hotel gross revenue. State exclusions and the reason for each in the output.
## Sector-specific red flags
- **Sale-and-leaseback gains dressed as operating profit.** SLB gains booked in other income, or netted against costs, can convert an operating loss into reported profit. Test: strip other income entirely and recompute EBITDAR margin. If profitability depends on selling aircraft, the airline is monetising its balance sheet, not earning. Under Ind-AS 116 only the portion of the gain relating to rights transferred is recognisable — aggressive interpretation here is a recurring audit issue.
- **OEM and manufacturer credits netted against costs.** Compensation for grounded aircraft, credit memos on new orders and delivery incentives are frequently netted into maintenance cost or other income rather than disclosed separately. They flatter CASK and margin, are non-recurring, and vanish when the dispute settles. Always ask what CASK ex-fuel looks like *before* credits.
- **Payables stretching and statutory dues in arrears.** In India the earliest and most reliable distress signal is not the P&L — it is delayed TDS, PF and GST payments and unpaid airport/AAI/oil-company dues, alongside ballooning trade payable days. Every major Indian airline failure showed this first. Read the auditor's report, the **CARO annexure on statutory dues**, and any emphasis-of-matter or going-concern paragraph before opening the investor presentation.
- **Unearned revenue rising while cash falls.** Customer float funding operating losses. This is what converts a troubled carrier or tour operator into an insolvent one within a quarter: bookings slow, the float unwinds, and refunds fall due simultaneously. Also size unredeemed vouchers and travel credits — unfunded liabilities that crystallise at the worst possible moment.
- **Under-provisioned maintenance and aircraft redelivery obligations.** Leased aircraft must be returned in a contractually specified condition, requiring costly engine and airframe checks. Accrual conservatism varies enormously. A young leased fleet with unusually low maintenance provisions is deferring a large cash cost — compare provision per aircraft against peers. The hotel analogue is the **FF&E reserve**: a genuine reserve is 4–5% of revenue, and its absence means deferred renovation.
- **RevPAR or ARR growth that is mix, not pricing.** ARR rises automatically when a company opens luxury keys, closes economy keys or acquires upscale assets. Insist on like-for-like RevPAR and check RGI against the comp set. Equally, "system-wide revenue" or "gross revenue of managed hotels" is **not the company's revenue** — operators headline it and it can be 4–8x consolidated revenue.
- **Renovation capex misclassified as growth capex.** Hotels need soft refurbishment every 6–8 years and hard refurbishment every 12–15. Booking it as expansion overstates maintenance-adjusted FCF and understates capital intensity. Cross-check: near-zero capex per existing key while the portfolio ages means a large deferred bill, and it shows up as RGI erosion before it shows up in the accounts.
- **Custom, non-GAAP "adjusted EBITDA" definitions.** Watch for EBITDA presented pre-Ind-AS-116, before ESOP cost, before one-offs, or with lease rentals selectively added back — and for the definition changing between quarters. Rebuild EBITDAR yourself from the statutory statements. Any company whose adjusted metric flatters results by more than ~15% with an unstable definition deserves a discount.
- **Forex MTM and deferred tax assets inflating PAT.** Carriers with USD-denominated lease liabilities and maintenance provisions report large non-cash forex gains when the rupee strengthens; separately, recognising a deferred tax asset on accumulated losses creates a one-time PAT boost with no cash counterpart. Both make trailing P/E and EPS growth meaningless. Model cash profit before FX and DTA.
- **Negative net worth presented as capital efficiency.** For a buyback-driven asset-light operator this can be benign; for an airline it is a solvency issue that D/E actively conceals by turning negative. Compute adjusted net debt/EBITDAR instead, and establish whether the negative equity came from buybacks or from accumulated losses — the distinction is everything.
- **Aggressive loyalty breakage assumptions.** Loyalty programmes recognise revenue using estimated redemption rates. Raising assumed breakage releases deferred revenue straight to the P&L with no economic event. Check the accounting policy note year on year and be sceptical of margin improvement concentrated in the loyalty segment.
- **Capacity expansion into a deteriorating supply-demand balance.** Order books and hotel pipelines are visible years ahead. A carrier inducting aircraft into an oversupplied market, or an operator opening keys where supply growth exceeds demand growth, is committing capital to a yield decline. Watch aggregate industry ASK growth against passenger growth and city-level room pipelines — the equity market usually reacts about a year late.
- **Grounded aircraft not reconciled to reported fleet.** Companies report fleet size; what matters is aircraft in revenue service. A widening gap means ownership cost with no ASK, crushing CASK while headline fleet growth looks healthy. Cross-check regulator (DGCA) in-service counts against company disclosure.
- **Related-party management contracts, promoter pledging and asset transfers.** In Indian hospitality, hotels are often owned by promoter-affiliated entities and managed by the listed company, or the reverse, with fee terms set off-market. Scrutinise the related-party note for management fees, lease rentals and asset purchases. High promoter pledging in a cyclical, leveraged business compounds risk — a downcycle triggers margin calls exactly when the equity is cheapest.
- **Interest capitalisation during long hotel construction.** A hotel takes 3–5 years to build, with interest capitalised into the asset rather than expensed — inflating profit and assets during the build and depressing returns afterwards. A large capital-work-in-progress balance relative to gross block means profit reported today will be paid for in depreciation and interest later. Compute interest coverage on **total**, not expensed, interest.
- **OTA customer-acquisition subsidy disguised as revenue growth.** Incentives and discounts may be recorded as a reduction of revenue (correct under Ind-AS 115) or as marketing expense, and companies switch presentation. Compare GBV growth against marketing plus incentives combined; if the latter grows faster, the growth is purchased. Falling direct/app traffic share alongside rising paid-search spend is the same disease in a different metric.
## Cycle and structural context
**Position in the cycle is the dominant variable, and it must be stated explicitly.** Airline profitability is a function of the industry capacity cycle interacting with fuel; hotel profitability is a function of the room-supply cycle interacting with GDP and corporate travel budgets. Both are amplified by operating leverage, so the swing from trough to peak margin is far wider than the swing in demand. Never present a valuation without naming where in the cycle the trailing numbers sit, and never extrapolate peak flow-through.
**The supply side is forecastable and the demand side is not.** Aircraft order books, delivery schedules, engine availability and city-level hotel pipelines are published years ahead. Build the supply picture first; it constrains pricing regardless of demand narrative. Aircraft delivery delays and engine reliability problems have recently acted as an *involuntary* supply constraint that propped up yields — a benefit that reverses when the backlog clears, so do not capitalise it.
**Structural threats to model explicitly.** Video conferencing has permanently reduced some corporate travel demand, hitting full-service carriers and business-district hotels hardest. Alternative accommodation platforms compress the low end of hotel pricing power in leisure markets. Aggregators and metasearch continue to disintermediate direct booking, raising customer-acquisition cost across airlines, hotels and restaurants alike. Decarbonisation is a rising cost: EU ETS and CORSIA obligations, SAF blending mandates in Europe and elsewhere, and potential carbon pricing add cost per ASK that cannot be avoided by efficiency alone. Airport charges and slot scarcity at constrained hubs are a rising rent extracted from carriers.
**Consolidation is the sector's only reliable margin mechanism.** Airline returns improve durably only when capacity leaves — through failures, mergers or capacity discipline. A market with a distressed marginal player pricing below cost is a market where nobody earns their cost of capital. Conversely, a market that has just consolidated into two or three disciplined players can sustain returns for years. Assess the competitive structure before assessing the company.
**Regulation shapes the economics directly.** Airlines face route-rights and bilateral regimes, slot allocation, foreign-ownership caps, safety oversight, and consumer-compensation regimes (EU261 in Europe, DGCA passenger-rights rules in India) that create real liabilities. Hotels face licensing, liquor permits, fire and environmental clearance, and coastal-zone rules that determine how fast supply can be added. Restaurants face food-safety regulation and, in India, the GST rate structure on restaurant services and the input-tax-credit position, which materially changes effective margin.
## India vs global notes
| Dimension | India | US / global |
|---|---|---|
| Regulator / oversight | DGCA (safety, slots, fleet in-service counts), MoCA policy, AAI and private airport operators for charges, BCAS security; hotel classification via the Ministry of Tourism; FSSAI for restaurants | FAA/DOT (US), EASA/national CAAs (EU); slot coordination by IATA-designated coordinators; consumer regimes such as EU261 and DOT rules create quantified liabilities |
| Accounting | Ind-AS 116 from FY20 — **breaks the pre-FY20 margin and leverage series**; Ind-AS 115 for principal-vs-agent at OTAs; Ind-AS 21 forces large non-cash FX MTM on USD lease liabilities | IFRS 16 from 2019, ASC 842 in the US (operating leases stay in opex under US GAAP — so a US carrier's EBITDA is *not* directly comparable to an IFRS carrier's; EBITDAR reconciles them) |
| Fuel and taxes | ATF carries state VAT of ~1–30% on top of central duties, so identical routes have different fuel cost by departure state; almost no hedging market, so near-full spot exposure | Jet fuel taxed more uniformly; deep hedging markets available; hedge ratios of 0–60% are a real policy choice and must be disclosed |
| Cost and currency exposure | Revenue largely in ₹, but leases, maintenance, fuel and aircraft purchases largely USD-linked — a structural short-USD mismatch that dominates reported PAT | Better natural hedging for US carriers; European carriers carry USD fuel/aircraft exposure against EUR revenue and typically hedge it |
| Units and reporting | ₹ crore/lakh; fiscal year April–March; ARR and RevPAR quoted per room-night in ₹; ASK/RPK in millions; **quarterly seasonality: Oct–Mar strong for hotels (weddings, MICE, leisure), Q1 monsoon-weak for both hotels and airlines** | USD millions; calendar year typical; RevPAR and ADR in USD; RASM/CASM in US cents per ASM (miles, not kilometres) — **never mix ASK and ASM without converting** |
| Disclosure venue | Quarterly investor presentation plus the **concall** — the primary source for CASK ex-fuel, fleet in service, same-store RevPAR, pipeline keys and store-level economics; annual report notes for lease liabilities and maintenance provisions; CARO 2020 on statutory dues and defaults | 10-K/10-Q on EDGAR plus monthly traffic releases and an investor supplemental; hotel REITs publish property-level and same-store detail; airlines publish monthly capacity and load-factor data |
| Ownership and governance | Promoter holding is central — check pledging, promoter-owned hotel-owning entities managed by the listed company, and related-party fee terms; SEBI related-party approval rules apply | Widely held; governance axis is board independence and management incentive design; franchisor-franchisee conflicts are the analogous related-party issue |
| Hotel structure | Owner-operator model still dominant, transitioning to managed/franchised; F&B and banqueting 40–50% of revenue; no hotel REIT regime, so SOTP/NAV is the standard lens | Brand-manager model dominant with owners separated out; F&B 25–30%; deep hotel REIT market (FFO/AFFO, EBITDA per key, implied cap rate) and observable transaction cap rates |
| Ancillary and unbundling | Unbundling constrained by regulation and high price sensitivity — ancillary per passenger materially lower; cargo/belly freight a bigger contributor; loyalty monetisation still immature | Ancillary and loyalty highly developed; co-brand card economics can exceed the value of the flying operation; unbundling largely unrestricted |
| Restaurants | GST rate structure on restaurant services and the input-tax-credit position materially change effective margin; aggregator commissions 18–25%; master-franchisee model common on NSE/BSE | Franchisor royalty models dominate; unit economics disclosed as AUV and restaurant-level margin; delivery aggregator penetration and commissions vary by market |
## Checklist
- [ ] Classify first: airline (LCC/FSC), hotel owner, hotel operator/franchisor, OTA, restaurant chain or hybrid — and split hybrids into segments before any multiple.
- [ ] Delete OPM, D/E, P/E, ROE, P/B, current ratio and headline FCF from the analysis; state in the output why each is inapplicable rather than reporting it with a caveat.
- [ ] Rebuild **EBITDAR** from the statutory statements and use EBITDAR margin for every cross-company margin comparison.
- [ ] Compute **adjusted net debt / EBITDAR** (debt + capitalised leases + maintenance/redelivery provisions + unfunded pension − cash) and **liquidity months**; treat under two months of liquidity as a going-concern question.
- [ ] Confirm the lease-accounting regime and period for every peer; never compare pre-FY20 to post-FY20 margins, or IFRS 16 to US GAAP ASC 842 EBITDA, without reconciling to EBITDAR.
- [ ] Anchor all growth series to FY19/FY20; discard any CAGR whose base falls in FY21–FY22 and say so.
- [ ] Airline: build the RASK–CASK spread and track **CASK ex-fuel over 8–12 quarters**; ask what it looks like before OEM credits.
- [ ] Airline: read PLF against yield, and PLF against industry ASK growth — rising loads with falling yield is bought traffic.
- [ ] Airline: reconcile balance-sheet fleet against aircraft **in revenue service**; check block hours/day, fleet age and number of types.
- [ ] Airline: quantify ancillary revenue per passenger and its trend; value loyalty and cargo separately in SOTP.
- [ ] Airline: normalise fuel and currency to mid-cycle and re-test whether the business clears its cost of capital.
- [ ] Hotel: demand **like-for-like RevPAR**, decompose it into ARR vs occupancy, and compute incremental flow-through.
- [ ] Hotel: pull RGI/MPI/ARI against the competitive set to separate cycle from skill.
- [ ] Hotel: split owned/leased EBITDA from fee EBITDA and value each on its own multiple; never blend them.
- [ ] Hotel: check GOP margin and GOPPAR; verify an FF&E reserve of ~4–5% of revenue and capex per existing key.
- [ ] Hotel: build the **micro-market supply pipeline vs demand growth** for the company's key cities; flag any market with two consecutive years of supply outrunning demand.
- [ ] Hotel: confirm whether headline revenue is consolidated or "system-wide" gross revenue of managed hotels.
- [ ] OTA: reconstruct **GBV and take rate**; compare GBV growth against marketing plus incentives; check direct/app traffic share and repeat rate.
- [ ] Restaurant: decompose SSSG into transactions vs price; check ADS, store-level margin pre-Ind-AS-116, capex payback, gross closures and aggregator dependence.
- [ ] Track unearned revenue and vouchers as days of revenue **against the cash balance** — rising float with falling cash is the pre-insolvency signature.
- [ ] Strip working-capital movement, SLB proceeds and PDPs out of cash flow before calling anything free cash flow.
- [ ] Strip other income entirely and recompute EBITDAR; identify SLB gains, OEM credits, asset-sale gains, FX MTM and DTA recognition in reported PAT.
- [ ] India: read the auditor's report, CARO on statutory dues, going-concern and emphasis-of-matter paragraphs, and the related-party note before the investor presentation.
- [ ] India: check promoter pledging and promoter-affiliated asset-owning or management entities and their fee terms.
- [ ] Compute **cycle-averaged, lease-adjusted ROIC vs WACC** over 7–10 years; for hotel owners also present ROCE on revalued property.
- [ ] Value on **mid-cycle** EBITDAR, and invert the multiple reading: cheap at peak margins is a sell signal, expensive at trough is often the entry.
- [ ] Cross-check hotel EV against **EV per key vs replacement cost**, and hotel NAV against transaction cap rates.
- [ ] For any airline under six months of liquidity, replace the multiple with a probability-weighted scenario including dilution or restructuring.
- [ ] State the cycle position, the supply pipeline, the regulatory/decarbonisation cost trajectory, and the competitive structure of the market — then the peer set with exclusions named.
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 |