Hospital business model

Now switch machines entirely. A hospital sells a bed-day plus clinical capability. Economically it's a real-estate-plus-labour capacity business, and its whole story comes down to three numbers: how much each occupied bed earns, how full the beds are, and how a new hospital climbs from cash sink to cash machine over four to six years.

Master the last one, the J-curve, and you'll out-analyze most of the market. The market reads hospitals on blended margins and gets the J-curve backwards.

4.8.1 The revenue identity and the KPI panel

Every hospital's revenue reduces to:

Revenue ≈ operational beds × occupancy × ARPOB × days

where ARPOB = Average Revenue Per Occupied Bed, per day, the sector's headline intensity metric. Run this panel:

How do you read the main hospital KPIs?

A few of these deserve a closer look.

ARPOB is a mix metric, and you shouldn't read it as a price metric. It rises for three reasons: real case-mix escalation (more cardiac surgery, fewer fevers), payer-mix improvement (more cash/insurance, less government scheme), or plain medical inflation. Only the first two count as quality. A hospital whose ARPOB climbs purely on price, with the same case mix, is running to stand still. One whose ARPOB climbs because it added an oncology block or a cath lab is upgrading its economics. So always decompose it.

ALOS is the throughput lever, and it has a trap in it. A falling Average Length of Stay is clinically excellent (better protocols, fewer infections, day-care surgery). It also raises throughput, since the same bed serves more patients a year. But shorter stays cut bed-days, so falling ALOS mechanically pressures occupancy unless new admissions refill the freed capacity. Read the three together. Falling ALOS with steady occupancy and rising ARPOB is the ideal. You get more patients and more intense procedures, each staying a shorter, more profitable time. Falling ALOS with falling occupancy just means empty beds.

Payer mix and case mix are the same lever seen twice. The realization on a coronary bypass paid in cash is a large multiple of the same bed-day reimbursed under a government scheme (Ayushman Bharat/PMJAY, CGHS, ESI). Those schemes are all priced at low package rates, sometimes below a private hospital's cost. A chain that fills beds with scheme volume buys occupancy at the price of ARPOB. That can be the right call for a new hospital climbing the J-curve and the wrong call for a mature one. You can't judge occupancy without knowing what it was bought with.

4.8.2 The maturation J-curve: the one idea that separates the analysts

A hospital doesn't open profitable. Its entire cost base arrives on day one: depreciation on the building and equipment, doctor retainers, nursing, power. Revenue ramps for years as occupancy builds through referral networks, doctor reputation, insurance empanelment, and case-mix escalation. The trajectory is a J:

Why does a blended EBITDA margin mislead for hospital chains?

Here's the trap, and your edge. A chain's blended EBITDA margin is a weighted average of mature cash-cow beds and loss-making new beds. That average sets up a pair of mistakes, and the market makes both:

1. A chain adding beds aggressively looks like a margin disaster on blended numbers, even if every mature hospital is superb, because the new beds drag the average into the trough. Value it on the blend and you undervalue the mature estate. You also miss that the "margin decline" is an investment ramp and that nothing is deteriorating.
2. A chain that has stopped adding beds shows a beautiful blended margin, because everything is mature, but it has no runway. The market may pay a growth multiple for what is really a no-growth annuity.

The fix is segment analysis applied to beds. Demand the mature-versus-new split and compute EBITDA/bed for each cohort. Value the mature beds at mature economics, and treat the ramping beds as an investment, or a call option, with an explicit maturation assumption. That assumption is the judgment that carries every hospital thesis. Will these beds fill on schedule? It also sets up the sector's hospital red flag, debt-funded bed additions crushing blended margin while the ramp slips. If new beds are financed with debt and occupancy doesn't climb on plan, the J-curve losses plus interest overwhelm the mature cash flow. Then the nightmare arrives, and the beds simply never mature.

Capital is destroyed out of sight, under a falling reported margin that management calls "investment."

4.8.3 Greenfield vs brownfield: two ways to add a bed

• Greenfield. You build new, from the land up, at a capex of ₹1–2.5 cr per bed for a metro tertiary facility. You get a long J-curve (4–6 years) and full control of design and case mix, with the highest execution-and-ramp risk.
• Brownfield. You acquire an operating hospital, or add a tower to an existing campus. That buys existing occupancy and ramp, so the J-curve is much shorter and EBITDA comes faster. In return you pay an acquisition premium and inherit legacy issues. Adding capacity to an already-humming campus is often the best risk-adjusted move of all. It uses an established brand and referral base, so the new beds ramp far faster than true greenfield.

This is question 3 of the spine, capital allocation. A management team densifying proven markets brownfield takes far less risk than one greenfielding trophy hospitals in unproven cities. Before you believe the next expansion, audit the last one. Did those beds ramp on schedule? The ramp track record is the best predictor you have of whether the announced growth creates or destroys value.

4.8.4 Valuation: EV/EBITDA cross-checked by EV/bed

The primary lens is EV/EBITDA, because the P&L captures the capacity business's operating economics. P/E is a poor choice here. Depreciation on the property is huge and policy-sensitive, and leverage varies, the same reasons P/E was a poor fit for telecom. Blended EBITDA is J-curve-contaminated, though, so it needs a cross-check:

• EV/bed (EV per operational bed) is a capacity and replacement-value sanity check that the J-curve can't touch. Suppose EV/bed is far above the ₹1–2.5 cr greenfield build cost while much of the estate is still immature. Then the market is paying full price for beds that haven't earned yet, which means either deep faith in the ramp or overvaluation. If EV/bed is near or below build cost for a mature estate, that's the value signal.
• Sum-of-the-parts is the sophisticated version. It values the mature estate on mature EBITDA at a full multiple, and the ramping estate separately (at cost, or on a haircut DCF of its expected maturation). That's how you avoid both J-curve errors at once.

What do hospital chains trade at, and who are the benchmarks?

Some illustrative multiples. Mature Indian hospital chains traded at ~18–30× EV/EBITDA through the compounder era, at an EV/bed of ₹1.5–4 cr. The scaled US operator HCA trades far lower, at ~7–9× EV/EBITDA, on a different payer system (commercial insurance + Medicare/Medicaid), heavy leverage, and relentless buybacks. Same business, different financial animal. Never compare the multiples without the context.

Anchors (illustrative, ~FY24/FY25 scale; pull the filings). Apollo Hospitals is India's tertiary pioneer. The listed entity blends hospitals with pharmacy and diagnostics/digital, so isolate the hospital segment's ARPOB, occupancy and mature-vs-new EBITDA/bed before valuing it. Max Healthcare runs a dense metro-cluster model, with among the highest ARPOBs and mature-estate margins in India. Fortis is a turnaround and rebuild story after governance trauma, and a governance case in its own right. Narayana Hrudayalaya is the deliberately low-ARPOB, high-throughput cardiac model: cut price, run volume, still profit. It proves that a high ARPOB is a strategy you choose, never a virtue in itself. HCA Healthcare (US) is the scaled for-profit benchmark for how the capacity business runs at maturity in a commercial-payer system.

4.8.5 Hospital red flags