Skip to content

alirezarezvani/claude-skills / finance/skills/stock-analysis

stock-analysis

Unverified

Analyze, 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.

FileKindSize
SKILL.mdskill md32.7 KB
evals/evals.jsonconfig13.2 KB
evals/fixtures/synthetic-ar-excerpt.mdmarkdown5.7 KB
evals/fixtures/synthetic-drhp-excerpt.mdmarkdown3.4 KB
examples/README.mdmarkdown1.9 KB
examples/forensic-analysis-example.mdmarkdown8.8 KB
examples/standard-analysis-example.mdmarkdown15.6 KB
references/01-data-sourcing.mdmarkdown38.5 KB
references/02-core-factors.mdmarkdown49.3 KB
references/03-earnings-quality.mdmarkdown40.1 KB
references/04-balance-sheet-and-cashflow.mdmarkdown56.6 KB
references/05-returns-and-dupont.mdmarkdown49.8 KB
references/06-valuation.mdmarkdown53.7 KB
references/07-forensic-red-flags.mdmarkdown57.5 KB
references/08-governance.mdmarkdown67.0 KB
references/09-risk-and-macro.mdmarkdown52.7 KB
references/10-peer-set.mdmarkdown45.5 KB
references/11-scoring-rubric.mdmarkdown36.8 KB
references/12-report-template.mdmarkdown26.6 KB
references/13-situations.mdmarkdown60.6 KB
references/14-accounting-comparability.mdmarkdown48.6 KB
references/15-document-diligence.mdmarkdown76.1 KB
references/16-market-mechanics-and-tax.mdmarkdown48.6 KB
references/17-process-and-epistemics.mdmarkdown59.6 KB
references/18-forensic-mode.mdmarkdown16.5 KB
references/19-ipo-mode.mdmarkdown15.6 KB
references/20-challenge-pass.mdmarkdown14.1 KB
references/21-data-integrity-tools.mdmarkdown7.2 KB
references/sectors/_index.mdmarkdown67.8 KB
references/sectors/auto.mdmarkdown47.5 KB
references/sectors/aviation-hotels.mdmarkdown49.3 KB
references/sectors/banks.mdmarkdown38.6 KB
references/sectors/biotech-clinical.mdmarkdown47.8 KB
references/sectors/chemicals-cement.mdmarkdown46.2 KB
references/sectors/exchanges-payments.mdmarkdown41.2 KB
references/sectors/fmcg-consumer.mdmarkdown35.7 KB
references/sectors/holdco-assetmgr.mdmarkdown41.4 KB
references/sectors/infra-capitalgoods.mdmarkdown43.0 KB
references/sectors/insurance-brokers-services.mdmarkdown49.9 KB
references/sectors/insurance.mdmarkdown38.0 KB
references/sectors/it-saas.mdmarkdown44.3 KB
references/sectors/metals-mining.mdmarkdown49.4 KB
references/sectors/mortgage-reit-specialty-finance.mdmarkdown43.5 KB
references/sectors/nbfc.mdmarkdown40.3 KB
references/sectors/oil-gas.mdmarkdown43.5 KB
references/sectors/people-businesses.mdmarkdown46.3 KB
references/sectors/pharma-healthcare.mdmarkdown40.6 KB
references/sectors/rail-freight.mdmarkdown66.2 KB
references/sectors/realestate-reit.mdmarkdown44.3 KB
references/sectors/retail-ecommerce.mdmarkdown48.3 KB
references/sectors/semiconductors.mdmarkdown43.2 KB
references/sectors/shipping-logistics.mdmarkdown47.2 KB
references/sectors/telecom-media.mdmarkdown43.1 KB
references/sectors/utilities-power.mdmarkdown44.9 KB
references/sectors/waste-environmental.mdmarkdown46.5 KB
scripts/benchmarks.jsonconfig200.5 KB
scripts/lint_report.pyscript47.8 KB
scripts/ratios.pyscript127.4 KB
scripts/score.pyscript96.8 KB
scripts/valuation.pyscript37.9 KB
scripts/verify_data.pyscript72.5 KB

references/sectors/utilities-power.md

View on GitHub
# Power generation, transmission and regulated utilities — sector playbook

Use this when: the company earns money by generating, transmitting, distributing or trading electricity (or gas/water under an analogous regulated regime) — Indian gencos, transcos, discoms, renewable IPPs, power InvITs, US regulated and multi-utility holdcos, UK/EU RAB-regulated networks, and merchant generators anywhere.

The defining fact is that price is not set by competition. It is set by a regulator's formula or by a long-term PPA, and the asset base itself — not sales, not margin — is the earnings engine. Earnings equal rate base times allowed return; growth equals rate-base growth; risk equals counterparty and regulator behaviour. That inverts or voids most of the generic ratio set, and it means a diversified utility containing regulated networks, contracted renewables and merchant thermal must be pulled apart before any number is computed on 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

A generic OPM / ROCE / D-E / P-E / FCF checklist does not merely add noise here — it mis-ranks the sector almost systematically, and several of its signals point the wrong way. Do not report these without the correction stated alongside.

**OPM / EBITDA margin — an artefact of fuel pass-through accounting, not efficiency.** A distribution utility books the full retail tariff as revenue and power-purchase cost as expense, so 4–10% margins are structurally normal. A regulated transmission company or a solar IPP has almost no variable cost and prints 85–92% EBITDA margins. A coal genco on a fuel-pass-through PPA books fuel in both revenue and cost, so its margin *mechanically falls when coal prices rise* even though profit is unchanged. Margin here measures business mix, not operating quality. The only comparable margin is on **net revenue** (revenue less fuel and purchased power), or better, **EBITDA per unit sold** (₹/kWh, $/MWh).

**Revenue growth — meaningless.** Top line moves with fuel prices, fuel-surcharge true-ups (FPPPA/FAC in India, deferred fuel balances in the US) and pass-through of purchased power. A discom can grow revenue 25% while selling fewer units. The real growth variables are **units sold (MU/TWh), customers, MW commissioned and rate base**.

**ROCE — converges to the allowed return by construction, so it discriminates poorly and often perversely.** Under CERC/SERC norms an Indian regulated genco or transco earns a formula return (15.5% post-tax RoE on 30% notional equity, plus normative O&M, depreciation and interest); in the US an allowed ROE of roughly 9–10.5%, in the UK/EU a WACC-based RAB return, does the same. ROCE far *above* the allowed return usually means old, fully-depreciated assets — a shrinking earnings base, not a moat — or large merchant exposure. ROCE far *below* may simply reflect a big CWIP block that legally earns nothing until commissioning. Always recompute excluding CWIP, and split regulated from merchant.

**D/E — rejects the entire sector.** Regulated networks are financed to 55–65% gearing on RAB *precisely because* the cash flows are contracted; project-financed generation SPVs are built 70:30 or 75:25 debt:equity by regulatory design. A "D/E < 1" filter eliminates essentially every listed transmission utility, most IPPs and every InvIT. Replace it with rating-agency cash-flow ratios (FFO/Debt, FFO interest cover), Debt/RAB, and a recourse-vs-non-recourse split.

**Free cash flow — structurally negative during rate-base growth, and that is a buy signal.** A utility is *supposed* to spend more than depreciation; value is created by earning the allowed return on new capex. Positive FCF often means under-investment and a shrinking future earnings base. Reported FCF also does not separate maintenance from growth capex. Use FFO, FFO less maintenance capex, and capex/depreciation instead.

**P/E — corrupted by non-cash and timing items unique to this sector.** Regulatory deferral account balances (Ind AS 114) and US-style regulatory assets/liabilities; prior-period true-ups and arrears landing in a single quarter; late-payment surcharge income; capitalised interest during construction (IDC); in India, capitalisation of exchange differences on long-term borrowings under the Ind AS 21 D13AA exemption; IFRS 9 mark-to-market on merchant hedges; deferred tax and MAT credits; and book gains on asset drop-downs into sponsored InvITs/YieldCos. Two identical plants can report very different EPS.

**P/B — usable, but only with a translation layer.** The economically relevant capital is the regulator's rate base / RAB, not Ind-AS or IFRS net block. Assets funded by consumer contributions and government grants (sitting as deferred revenue), disallowed capex, and revaluation differences all break the link. P/B remains a core utility tool — but read against *earned* ROE and against RAB, never alone.

**Current ratio, receivable days and working-capital screens — wrong, especially in India.** Receivables from state discoms are a political variable; 90–250 days DSO is common and does not by itself indicate a bad operator. Conversely a sudden fall in receivables may just be a one-off government bailout or securitisation scheme, not an operating improvement.

**Interest cover and asset turnover — structurally punished.** EBIT interest cover is distorted by IDC capitalisation (interest that never hits the P&L). Asset turnover of 0.2–0.4x is normal for the most capital-intensive sector in the market; a generic screen reads it as inefficiency.

**Payout-ratio screens — misfire in the opposite direction.** A 50–80% payout is normal and healthy for a regulated utility. The "low payout = reinvestment quality" heuristic does not apply; here, low payout with weak rate-base growth is simply cash trapped.

**Any consolidated ratio on a diversified utility — misleading by construction.** Regulated networks, contracted renewables and merchant thermal carry different risk, different cost of capital and different multiples. Analyse and value them separately or do not analyse them at all.

## The metrics that actually matter

Ranges are **indicative only**. They vary by market, technology, regulatory regime, control period and point in the cycle, and any of them can be wrong for a specific asset in a specific year. Comparison against a tight peer set, against the *regulator's own normative level*, and against the company's own 5–10 year history overrides every absolute band below.

| Metric | Definition / how to compute | Indicative healthy range | Why it matters |
|---|---|---|---|
| **Regulated Asset Base (RAB) / rate base growth** | The regulator-approved capital on which a return is allowed, and its CAGR. India: practical proxy is approved gross block plus CWIP capitalisation admitted in the current control period (CERC/SERC tariff orders). UK/EU: published RAV/RAB. US: rate base per the last rate case. | DM regulated networks 5–8% p.a. real; Indian transmission/distribution 8–14% nominal gross-block growth *with capitalisation actually approved*. Below nominal GDP growth = a shrinking earnings engine. | Earnings = rate base × allowed return, so rate-base growth is essentially the entire organic EPS algorithm — and far more predictable than sales or margin. It also reveals whether reported capex is converting into earning assets or piling up in CWIP. |
| **Allowed ROE vs earned ROE (the under-earning gap)** | The return the regulator permits versus what is actually earned on regulated equity, plus the *reasons* for the gap: regulatory lag, disallowed capex, cost over-runs, incentive/penalty outcomes, under-recovery of fixed cost. | US allowed ROE typically 9.0–10.5%, earned within 50–100 bps of allowed is good, >150 bps under-earning is a flag. India: CERC norm 15.5% post-tax RoE on 30% notional equity for central generation/transmission; SERC distribution 14–16%. Chronic realisation below 80% of allowed = weak regulator or weak operator. | This single spread explains most of the valuation dispersion between regulated utilities. A reliable full-earner deserves ~1.2–1.6x book; a chronic under-earner deserves a discount to book regardless of what its P/E looks like. |
| **Availability (NAF) vs PLF vs CUF** | For regulated thermal: declared availability, which drives fixed-cost (capacity charge) recovery — distinct from PLF, which only drives variable-cost recovery. For renewables: Capacity Utilisation Factor / load factor. Compare to the regulator's normative level, not to zero. | Indian coal under CERC norms: 85% NAF for full fixed-cost recovery, incentive above; >88–90% is strong. All-India thermal PLF ~65–70%; a plant >80% is good. Solar CUF 19–25% (single-axis tracking to 26–28%); onshore wind 25–35%; offshore wind 45–55%; hydro 30–45%; nuclear >85%. | In a capacity-payment regime, availability — not generation — pays the fixed costs and the equity return. A plant at 45% PLF can be fully profitable at 90% availability. Sustained availability below the norm is direct, unrecoverable erosion of the allowed return. For renewables, CUF *is* the revenue line and is the number most often over-promised in project models. |
| **AT&C losses and collection efficiency** | Distribution: 1 − (billing efficiency × collection efficiency), capturing technical line losses plus theft, non-billing and non-collection. DM analogue: T&D loss plus bad-debt/uncollected revenue. | Best-in-class Indian private discoms and licensed franchises 6–9%; RDSS national target 12–15%; weak state discoms 20–35%+. DM T&D losses 4–7% with ~99% collection efficiency. | The cleanest single measure of distribution operating quality, and of whether privatisation/franchise value has actually been created. Every 100 bps of AT&C reduction drops almost entirely to EBITDA. It is also what regulators reward with incentive tariffs, and what determines whether a discom can pay its gencos. |
| **ACS–ARR gap** | Average Cost of Supply per unit less Average Revenue Realised per unit (₹/kWh) — the structural cash loss per unit sold. Largely an Indian construct; DM analogue is the revenue-requirement shortfall or deferred fuel balance. | Zero or negative. All-India has run roughly ₹0.30–0.50/kWh; weak states ₹1.0–1.5/kWh. Any positive gap not funded by an explicit, budgeted state subsidy is unsustainable. | Root cause of the entire Indian power receivable chain. A genco or transco selling into wide-gap states shows good P&L revenue and terrible cash conversion. It is also the best forward indicator of tariff shocks, PPA renegotiation attempts and payment default. |
| **Station heat rate, specific fuel consumption, auxiliary consumption** | Thermal efficiency (kcal/kWh, or Btu/kWh), specific coal/oil consumption (kg/kWh, ml/kWh), and in-house consumption as a share of gross generation — each measured **against the regulator's normative level**. | Indian supercritical coal normative SHR ~2,375–2,450 kcal/kWh (subcritical ~2,450–2,500); auxiliary consumption norm 5.25–6.5% coal, ~1–2% gas/CCGT. US/EU CCGT 6,400–7,000 Btu/kWh (50–60% net efficiency). Beating the norm is where real operating alpha appears. | Under cost-plus tariffs, performance versus norm is retained by (or charged to) the company — the only genuine efficiency lever a regulated genco has, and completely invisible in OPM because fuel is pass-through. For merchant plants, heat rate sets merit-order position and therefore dispatch hours. |
| **Variable cost per unit / merit-order position; spark and dark spreads** | Fuel plus variable O&M per kWh, benchmarked against the market clearing price (IEX DAM/RTM in India) or against the clean spark spread (gas margin after fuel and carbon) and dark spread (coal equivalent) in DM. | Indian pithead coal variable cost ₹1.7–2.6/kWh vs imported-coal plants ₹4–7/kWh; IEX DAM typically ₹3–6/kWh with scarcity spikes. Bottom-quartile variable cost in the national merit order = near-certain dispatch. | The single determinant of a merchant plant's earnings, and of a regulated plant's PLF (hence incentive income). Two plants with identical capacity and identical D/E can have opposite economics purely on fuel logistics — pithead vs imported, linkage vs e-auction, FSA coverage. |
| **Contracted vs merchant mix, weighted-average PPA tenor, hedge ratio** | Share of capacity (MW) *and* of EBITDA under long-term PPAs vs short-term/merchant; weighted-average remaining PPA life; counterparty credit quality; for merchant fleets the forward hedge ratio by year. | Investment-grade IPPs: >80–85% of capacity contracted with >10 years weighted-average remaining life. Merchant fleets: 85–100% hedged for the next 12 months, 50–70% year 2, 25–40% year 3. Under 5 years average PPA life with no re-contracting visibility is a cliff. | This sets the cost of capital and the multiple: contracted cash flows are bond-like (10–13x EV/EBITDA), merchant cash flows commodity-like (5–8x). It is also the most obscured number in consolidated reporting — rolling one-year "short-term PPAs" get described as contracted. |
| **FFO/Debt, FFO interest cover, Net Debt/EBITDA, Debt/RAB** | Rating-agency cash-flow leverage, adjusted for hybrids/perpetuals, leases, pension and decommissioning provisions, with debt split between recourse (holdco) and non-recourse (project SPV). | Regulated networks: FFO/Debt 13–20% for BBB (>18–20% for A), FFO interest cover 3.0–4.5x, Net Debt/EBITDA 4.5–6.0x, Debt/RAB 55–70%. Merchant generation: Net Debt/EBITDA below 3.5–4.0x, FFO/Debt above 25–30%. Indian IPPs: under 5x for contracted, under 3.5x for merchant-exposed. | Replaces D/E entirely — utilities are financed against cash-flow stability, not equity book. Downgrades are valuation events here, because the whole model rests on continuous access to cheap long-dated debt; a fall below investment grade can make the growth plan uneconomic overnight. |
| **Receivable days by counterparty, and the regulatory asset / deferral account balance** | DSO split by counterparty state/utility, overdues above 60 and 90 days, plus the balance-sheet stock of "regulatory assets" / Ind AS 114 regulatory deferral account balances — costs recognised in P&L but not yet admitted into tariff. | Indian gencos/transcos: 45–90 days healthy, above 150 days signals a weak counterparty book. Regulatory assets should be small and amortising on an approved, dated schedule; a balance exceeding 6–12 months of revenue is serious impairment risk. | This is where earnings quality lives. Profit is booked on accrual against a state counterparty that may not pay for years, and the regulatory asset is effectively an unsecured, unrecognised receivable on a *future regulator*. Rising regulatory assets alongside rising PAT is the classic sector accounting trap. |
| **Capex/depreciation, capitalisation rate, CWIP ageing** | Capex ÷ depreciation (growth intensity); the share of the year's capex actually capitalised and admitted into rate base; and the ageing of CWIP — how long each project has sat accruing IDC. | Capex/depreciation 1.3–2.0x indicates healthy rate-base growth; below 1.0x means shrinkage. CWIP under ~15–25% of gross block for a mature utility, turning over within the project cycle; individual line items older than 4–5 years are usually stranded. | Capex is the growth engine only if the regulator admits it. Aged CWIP is capital earning nothing while interest is capitalised into the asset — flattering the P&L — and is the most common precursor to a large impairment or a disallowance of cost over-runs. |
| **EBITDA per unit sold and cost per unit delivered** | EBITDA ÷ units generated/sold (₹/kWh, $/MWh); separately, O&M + employee + other opex per unit, per customer, per circuit-km or per MW. | Highly technology-specific: Indian regulated coal roughly ₹1.2–1.8/kWh EBITDA; solar ₹2.5–3.5/kWh; transmission better measured per ckt-km or per MVA. Judge trend and peer-relative, not level. Fixed O&M should track the regulator's escalation index (~4–6% p.a. in India), not exceed it. | Strips out fuel and purchased-power pass-through — exactly what OPM fails to do — and makes a discom, a genco and a transco comparable on unit economics. Opex per unit versus the normative allowance is where a regulated operator either keeps or loses the efficiency gain. |
| **Emissions intensity, compliance capex, remaining thermal asset life** | gCO₂/kWh; SOx/NOx/PM compliance status (FGD, SCR, ESP upgrades); carbon cost exposure (EU/UK ETS, US state programmes); remaining useful life of thermal assets versus the depreciation schedule *and* versus PPA tenor. | EU/UK integrated utilities largely below 200–300 gCO₂/kWh; Indian coal-heavy gencos near 900–1,000 gCO₂/kWh. Indian FGD has cost roughly ₹0.4–0.6 crore/MW and adds ~₹0.25–0.50/kWh to tariff. Flag thermal assets with 25+ years of book life left but a PPA ending within 10. | Drives mandatory capex (which may or may not be tariff-recoverable) and terminal-value risk. A coal plant depreciated over 40 years in a market decarbonising in 15 carries an unrecognised impairment. In DM, carbon cost is a direct hit to the dark spread and to dispatch order. |
| **Under-construction pipeline: MW by contracting stage** | Forward capacity split into (a) awarded, no PPA/PSA signed, (b) PPA signed, no financial closure, (c) under construction and financed — with the winning tariff and assumed equity IRR on each tranche. | For an Indian renewable IPP only the PPA+PSA-signed, financially-closed portion is meaningful. Target project equity IRR 13–16% India, 8–12% DM; a bid tariff implying sub-10% equity IRR at realistic CUF is value-destructive growth. Award-to-commissioning should be 18–24 months. | The gap between "MW awarded" and "MW earning" is the most systematically over-marketed number in the sector — multi-GW of Indian central and state auction awards have sat stranded for years without a signed power sale agreement. It is also the capital-discipline test: winning volume at uneconomic tariffs grows rate base while shrinking value. |
| **Grid and network operating KPIs** | Distribution/transmission reliability: SAIDI/SAIFI (or India's outage-hours and RDSS reliability metrics), transmission system availability against the CERC norm, transformer failure rate, circuit-km and MVA added, connections/customers added, and curtailment hours suffered by must-run renewables. | Indian transmission availability norm ~98–99% for full incentive; DM SAIDI targets are set explicitly by the regulator with penalty/reward attached. Rising curtailment above a few percent of generation is a direct revenue loss unless deemed-generation is contractually payable *and* actually paid. | These are the KPIs the regulator pays or penalises on, so they translate mechanically into allowed revenue. Reliability deterioration also precedes regulatory intervention, licence-condition disputes and consumer-facing political pressure — usually before it appears in the accounts. |
| **Distribution / dividend coverage from FFO less maintenance capex** | Dividend (or InvIT/YieldCo distribution) tested against FFO minus maintenance capex, not against EPS; plus the share of distribution that is return *of* capital. | Payout 50–80% of EPS is normal in DM; coverage on FFO-less-maintenance-capex above 1.0x on a multi-year basis. For InvITs/YieldCos, examine the NDCF bridge line by line. | Utilities are held as yield instruments, so a distribution cut is a de-rating event, not a rounding error. EPS-based coverage hides the fact that depreciation understates true asset replacement cost in an inflationary capex environment, and that some distributions are simply returning the unitholder's own capital. |

## How to value companies in this sector

Utilities are valued as **regulated capital stocks and contracted cash-flow streams**, not as earnings multiples. A diversified utility must be valued sum-of-the-parts; a single blended P/E on such an entity is close to meaningless.

**1. EV/RAB (or P/RAB) — primary for regulated networks.** The tool of choice for UK/EU transmission and distribution, and increasingly for Indian transmission and InvIT-held networks. You pay a premium or discount to the regulator-approved capital base depending on whether the company can out-earn its allowed return, beat its totex allowance and grow the RAB. Typical range 1.0–1.4x. Above ~1.4x embeds aggressive outperformance or aggressive growth; below 1.0x implies the market expects disallowance or chronic under-earning. It works because the regulator has effectively fixed both the numerator (allowed return) and the denominator (capital).

**2. P/B read against earned ROE — the standard cross-check** for US regulated utilities and Indian regulated PSUs. Justified P/B ≈ (earned ROE − g) / (Ke − g). Because allowed ROE is administratively set and stable, P/B is a far more stable and comparable anchor than P/E. A utility earning its allowed 15.5% RoE against a ~12% cost of equity supports a meaningful premium to book; one earning 9% against the same Ke should trade below book. US regulated utilities have historically traded 1.6–2.2x book; Indian regulated PSUs 1.0–2.5x depending on the growth and payout cycle.

**3. Long-horizon DCF, modelled control period by control period.** This is one of the few sectors where a 10–20 year explicit DCF is genuinely defensible, because the tariff formula, PPA tenor and asset life are contractually known. Model each regulatory period separately — allowed return, capex allowance, incentive/penalty, true-ups — and take terminal value off **closing RAB**, not a perpetuity growth rate. For contracted IPPs, model to PPA expiry with an explicit and conservative re-contracting or residual-value assumption; never a full-value perpetuity beyond the contract.

**4. DDM and dividend yield versus the sovereign bond.** Utilities are yield instruments and trade on the spread of dividend yield over the 10-year government bond; a widening spread is often the cleanest valuation signal in the sector. Test coverage against FFO less maintenance capex, not EPS.

**5. EV/EBITDA in preference to P/E**, because it neutralises capital structure, the enormous depreciation charge and differing tax positions (MAT, deferred tax, accelerated depreciation on renewables). Indicative bands: regulated networks 9–13x; contracted renewables 9–12x; contracted thermal 6–9x; merchant generation 4–7x on **mid-cycle** spreads. Always adjust EBITDA for capitalised interest and capitalised O&M, strip late-payment surcharge and other income, and neutralise regulatory deferral account movements.

**6. EV/MW and EV/kW as a replacement-cost sanity check.** India: greenfield coal roughly ₹8–9.5 crore/MW, solar ₹3.5–4.5 crore/MW, onshore wind ₹6–7.5 crore/MW, hydro ₹8–14 crore/MW; transmission per ckt-km or per MVA. US/EU: CCGT roughly $800–1,200/kW, utility-scale solar $900–1,300/kWac, onshore wind $1,300–1,700/kW, battery storage $250–400/kWh. Buying operating capacity materially below replacement cost is the classic entry signal in merchant power; paying well above it requires a demonstrable contract or grid-position advantage.

**7. Project-level equity IRR / NPV build-up for renewable and contracted IPPs, then aggregated.** Value each asset as discounted contracted cash flow to PPA end plus residual, net of project debt; add corporate overhead as a negative. This is how sponsors and lenders themselves underwrite, and it exposes auctions won at uneconomic tariffs that a consolidated EV/EBITDA hides completely.

**8. CAFD / NDCF yield for YieldCos and InvITs.** Indian power InvITs are valued on Net Distributable Cash Flow yield (typically 9–12%) and on NAV per unit versus market price, with careful attention to whether distributions are return *of* capital rather than return *on* capital, and to the residual asset life behind the yield. DM YieldCos are valued on CAFD yield plus DPS growth guidance.

**9. Merchant generation — value the fleet as a strip of spread options, not a stable EBITDA multiple.** Run mid-cycle and stressed spark/dark/clean-spark spreads through the plant's actual heat rate and dispatch profile; add explicit value for capacity-market payments and ancillary services; haircut for hedge roll-off. Merchant EBITDA must never be capitalised at a peak-of-cycle multiple — that is how power cycles are lost.

**Practical output for a diversified utility:** a sum-of-the-parts — regulated transmission at RAB or P/B, regulated distribution at P/B against earned RoE, contracted renewables at project DCF or EV/MW, merchant thermal at a low mid-cycle EV/EBITDA, and any manufacturing/EPC/mining arm on its own sector basis — less holdco net debt, less contingent liabilities, less a holdco discount (typically 15–25%).

**Do not use:** consolidated P/E on a mixed-asset utility; EV/Sales (revenue is a pass-through artefact); PEG; FCF yield on reported FCF; a perpetuity-growth terminal value on a contracted asset with a finite PPA; or peak-spread EV/EBITDA on merchant capacity.

## Peer set construction

A valid comparable shares the **regulatory regime, contracting structure and technology** — not merely the label "power". Mixing across these lines produces confidently wrong conclusions, because the multiple is a function of who bears price and volume risk.

**Splits that must not be mixed:**

- **Regulated network vs contracted generation vs merchant generation.** Different risk, different cost of capital, different multiple band. This is the primary split and it overrides every other similarity.
- **Generation vs transmission vs distribution.** Structurally different margin optics (85–92% vs 4–10% EBITDA margin), different KPI sets, different regulatory instruments. Never place a discom and a transco on the same margin chart.
- **Cost-plus regulated vs competitively-bid (TBCB/auction) assets, even inside the same company.** A cost-plus transmission line earns a formula return with true-ups; a TBCB line earns whatever the winning bid implied and carries the cost over-run risk itself.
- **Fuel type and fuel logistics.** Pithead coal, imported coal, domestic-linkage coal, gas/CCGT, hydro, nuclear, solar, wind, and storage have different variable costs, dispatch positions, capex per MW and terminal-value risk. A pithead and an imported-coal plant are not comparables.
- **Counterparty quality.** A genco selling to financially sound discoms or to central agencies is not comparable to one selling into states with a wide ACS–ARR gap and a renegotiation history, even at identical tariffs.
- **State-owned/PSU vs private.** Different cost of debt, different capital allocation incentives, government dividend and disinvestment overhangs, and different willingness to bid aggressively.
- **Corporate utility vs InvIT/YieldCo.** Yield vehicles have no retained growth, different tax treatment, an external sponsor/manager and a distribution-yield valuation basis. Do not put them in a P/E or P/B scatter with operating companies.
- **Consolidated vs standalone.** For groups with EPC, manufacturing, mining or trading arms, compare standalone regulated entities to each other and handle the rest in the SOTP.

**Also align:** regulatory jurisdiction and control-period stage (a company one year into a new control period faces a different lag from one at the end); accounting regime (Ind-AS 114 regulatory deferral accounts vs US GAAP ASC 980 regulatory assets vs IFRS without a regulatory-account standard — these are not comparable line items); fiscal year end (India April–March, most DM calendar); asset vintage (a fully-depreciated fleet flatters ROCE and depresses future rate base); and growth stage (a company mid-build will show weak ROCE and negative FCF for structural, not quality, reasons).

Aim for 5–8 peers. State the basis explicitly, and benchmark every operational metric **twice** — against peers, and against the regulator's normative level, which is the standard the company is actually paid against.

## Sector-specific red flags

- **Rising regulatory asset / Ind AS 114 deferral balance alongside rising reported PAT.** Costs are being recognised in the P&L against a tariff recovery the regulator has not yet approved. The sector's single most common earnings-quality trap; multiple Indian discoms have carried regulatory assets exceeding a full year of revenue. Ask: is there a dated, approved amortisation schedule, and has the regulator ever disallowed part of it?
- **Aggressive interest and overhead capitalisation.** A jump in the capitalisation rate, capitalised IDC growing faster than CWIP, employee cost and O&M capitalised into projects, or (India-specific) use of the Ind AS 21 D13AA exemption to capitalise forex losses on long-term borrowings into fixed assets instead of the P&L. Each flatters EPS and inflates the asset base simultaneously.
- **Aged, static CWIP.** Projects sitting 4+ years while IDC accrues, with no commissioning date. Stranded capital dressed up as growth; it ends in impairment or in a regulator refusing to admit the cost over-run into rate base.
- **Extension of asset useful life or a change in depreciation policy** — particularly stretching coal plant life to 30–40 years in a decarbonising market, or departing from the CERC depreciation schedule. Boosts EPS immediately and disguises stranded-asset risk.
- **Discom receivables managed by financing rather than collection.** Heavy use of late-payment-surcharge instalment schemes, receivable factoring or securitisation, bill discounting, or trade-receivable sales that move overdues off balance sheet and into "improved" operating cash flow. Check gross receivables and the >90-day ageing bucket, not just DSO.
- **Disputed or provisional tariff income booked as revenue.** Compensatory-tariff claims, change-in-law claims, carrying-cost claims and arbitration awards recognised while under appeal. Indian appellate history (APTEL and the Supreme Court) is full of reversals of exactly these.
- **PPA cliff and renegotiation risk.** A large share of capacity with PPAs expiring within 5 years and no re-contracting visibility; or exposure to states with a record of attempting retrospective tariff renegotiation. Also watch capacity awarded by central or state agencies with no downstream power sale agreement signed — it may never generate revenue.
- **Merchant exposure disguised as contracted.** Rolling one-year "short-term PPAs" presented as long-term contracting, a hedge book with a near-term maturity wall, or PPA counterparties that are trading intermediaries rather than end discoms. Read the hedge ratio *by year*, not the headline contracted percentage.
- **Related-party fuel, EPC or O&M contracts.** Fuel imported through a group trading arm, EPC awarded to a promoter entity, O&M priced above the regulator's norm. Benchmark fuel cost per unit and capex per MW against listed peers — a persistent unexplained premium is the tell, and over-invoicing allegations against Indian coal importers are the cautionary case.
- **Persistent under-recovery of fixed cost.** Availability below the normative level (below 85% NAF in India), frequent forced outages, or heavy partial-load operation. This permanently erodes the allowed return and is often buried in "other expenses" or a prior-period line.
- **Asset drop-downs into a sponsor-controlled InvIT or YieldCo at valuations generating large book gains.** The gain is recycled related-party value, not operating performance, and it can leave the listed sponsor holding the residual, harder-to-monetise assets.
- **Leverage that is not what it appears.** Hybrid or perpetual instruments counted as equity; non-recourse SPV debt carrying an undisclosed parent guarantee or sponsor support undertaking; large contingent liabilities relative to net worth. Reconstruct leverage on a fully-recourse basis before applying any coverage test.
- **Growth funded by repeated equity dilution, or by winning auctions at uneconomic tariffs.** Aggressive low bids in solar/wind or transmission auctions grow reported MW and rate base while destroying equity value. Test each win against realistic CUF, cost of debt and equipment cost.
- **Under-provisioned long-tail obligations.** Ash pond and mine reclamation, nuclear decommissioning funds, pension deficits — and the discount rate used on each. A 50 bps discount-rate change can move the reported liability materially and is a common quiet lever.
- **EBITDA quality contamination.** Late-payment surcharge income, treasury income on surplus cash, one-off insurance recoveries, government grant amortisation and consumer-contribution amortisation sitting inside "operating" EBITDA. Rebuild EBITDA from units sold × unit economics and compare.
- **Promoter share pledging and holdco leverage (India), plus a widening gap between consolidated and standalone cash flow.** Dividends may not be reaching the listed entity from the SPVs where the cash actually sits — project-finance covenants create dividend traps, and consolidated EBITDA can be structurally unavailable to the parent.
- **Regulatory relationship deterioration.** Delayed or unfiled ARR/true-up petitions, repeated disallowances, tariff orders issued years late, a state commission operating without full quorum, or a state failing to release budgeted subsidy on time. The regulator is the counterparty; its behaviour is a fundamental, not a footnote.
- **Resource shortfall hidden behind "deemed generation" claims (hydro and renewables).** Poor hydrology, low wind years, soiling and degradation above the modelled 0.5–0.7% p.a. for solar, or grid curtailment despite must-run status. Compare actual generation against P50/P90 estimates over 3–5 years, never one good year.

## Cycle and structural context

**There are two different cycles, and which one applies depends on the asset.** Regulated networks run on the **regulatory cycle** — the control period. Returns are reset every 3–5 years (CERC control periods in India, RIIO price controls in the UK, rate cases in the US), and the reset is the single largest scheduled valuation event a network faces. Early in a period, outperformance against the totex/opex allowance is retained; late in a period, the regulator has seen the outperformance and claws it into the next allowance. Always know where in the control period a company sits, and what the draft determination looks like.

**Merchant generation runs on the classic capacity cycle**, and it is brutal in both directions: high spreads attract capacity additions, additions crush spreads for years, low returns stop investment, reserve margins tighten, spreads spike again. The reliable rule is that merchant EBITDA at the top of the cycle should never be capitalised at a normal multiple, and merchant assets bought below replacement cost at the bottom are where the sector's largest returns have historically been made. Reserve margin and the forward capacity-addition pipeline are the leading indicators.

**Rate cycle matters twice over.** Utilities are long-duration bond proxies, so the equity de-rates when the sovereign yield rises — the dividend-yield-to-bond spread is the practical gauge. Separately, rising rates raise the cost of the debt that funds rate-base growth, and there is a lag before the regulator resets the allowed return to compensate; that lag is a real earnings hit for a company mid-build. Check the fixed/floating debt mix and the maturity wall.

**The secular structure is changing faster than the accounting.** Score these explicitly:

- **Decarbonisation and stranded assets.** Coal fleets carrying decades of remaining book life in markets targeting steep emissions cuts hold unrecognised impairment. Terminal value, not near-term EPS, is where this shows up.
- **Renewable cost deflation and auction competition** have compressed IPP returns; a portfolio built at old tariffs is a different asset from a pipeline being bid today.
- **Intermittency, storage and firming.** As renewable penetration rises, the value shifts from energy to **firm, dispatchable and flexible** capacity — storage, hydro, gas peakers, and ancillary services. Capacity markets, firm-and-dispatchable (FDRE) tenders and round-the-clock contracts are the commercial response. A pure-solar IPP faces curtailment and price-cannibalisation risk that its historical CUF does not capture.
- **Demand growth returning after two decades of stagnation in DM**, driven by data centres, electrification of transport and heat, and industrial electrification. This is a genuine change to the rate-base growth algorithm and the first real volume story utilities have had in a generation — but check whether the load is *contracted* and who pays for the network reinforcement.
- **Grid constraint as the binding scarcity.** Transmission and interconnection queues, not generation capex, are increasingly the bottleneck — which is structurally favourable for transmission RAB growth.
- **Distributed generation, rooftop solar and open access** erode the discom's most profitable industrial and commercial customers, worsening cross-subsidy economics and the ACS–ARR gap. In India, open access and captive generation removing high-tariff C&I load is a direct threat to distribution licensee earnings.

**Regulation is the business model, not a constraint on it.** In India: the Electricity Act 2003 framework, CERC and state SERC tariff regulations, the current control period's norms, RDSS conditionality and distribution privatisation/franchising, the late-payment-surcharge rules that forced discom payment discipline, RPO and carbon-credit obligations, and coal linkage and pricing policy. In the US: state PUC rate cases, FERC for transmission and wholesale markets, allowed-ROE trends, and formula-rate mechanisms. In the UK/EU: RIIO price controls, RAB indexation to inflation, and the ETS. A change in any of these repriced the sector before it appeared in any financial statement.

## India vs global notes

| Dimension | India | US / UK / EU |
|---|---|---|
| Regulator | CERC (central generation, interstate transmission, interstate trading); state SERCs (distribution, intra-state); APTEL as appellate body, then Supreme Court. Ministry of Power and state governments set policy and subsidy | US: state PUCs for retail rates, FERC for transmission and wholesale markets. UK: Ofgem (RIIO). EU: national regulators plus ACER. Independence and predictability generally higher |
| Return mechanism | Cost-plus with normative parameters: 15.5% post-tax RoE on a 30:70 notional equity:debt structure, normative O&M, availability-linked fixed-cost recovery, incentives for beating heat rate/availability norms. Increasingly displaced by competitive bidding (TBCB for transmission, reverse auctions for renewables) | US: rate cases setting allowed ROE (~9–10.5%) and equity ratio, with formula rates and trackers. UK/EU: RAB × allowed WACC with totex incentive sharing and RAB inflation indexation |
| Reporting | Ind-AS; Ind AS 114 permits regulatory deferral account balances presented as separate line items; ₹ crore/lakh; FY April–March; quarterly results, investor presentation and **analyst concall** — treat concall Q&A as a primary source on PPA status, receivables and capitalisation | 10-K/10-Q on EDGAR under US GAAP with ASC 980 regulatory assets/liabilities; UK/EU annual report under IFRS, which has no equivalent regulatory-asset standard, so RAB reconciliations sit in the regulatory accounts, not the IFRS statements |
| Counterparty risk | The dominant risk. State discoms are the buyers; ACS–ARR gaps, delayed subsidy release and long receivables are structural. Watch state-level payment data and any securitisation/bailout scheme | Largely absent for regulated utilities; DM merchant generators face market rather than counterparty risk. Credit risk sits in offtaker rating for contracted PPAs |
| Ownership and governance | Central PSUs, state utilities and private groups coexist; promoter holding, promoter share pledging, and government disinvestment overhangs matter. CARO reporting, related-party disclosures and auditor observations are the standard India-specific governance reads | Widely held; regulated utilities are typically single-jurisdiction operating companies under a holdco. Governance risk is more about capital allocation and M&A than promoter conduct |
| Yield vehicles | InvITs (power transmission and renewables) with SEBI-mandated NDCF distribution, external manager/sponsor structure, and unit-level yield valuation | YieldCos and listed infrastructure funds; MLP-like structures in some markets; distribution based on CAFD |
| Fuel and inputs | Coal linkage/FSA vs e-auction vs imported coal drives the entire variable-cost dispersion; gas availability limited; hydro subject to monsoon and state royalty/free-power obligations | Gas-dominated marginal pricing in the US, with carbon cost embedded in EU/UK dispatch; coal largely retiring in DM |
| Carbon | RPO obligations, carbon credit trading scheme in development, no broad economy-wide carbon price yet | EU ETS and UK ETS are a direct, quantified cost in the dark/clean-spark spread; US state and regional programmes plus tax-credit regimes (ITC/PTC and transferability) that materially change renewable project economics |
| Typical leverage convention | Project SPVs at 70:30 or 75:25 debt:equity by regulatory design; holdco recourse debt often separate and less visible | Networks geared 55–65% of RAB; rating agencies drive the target, and the rating is managed as a policy variable |

## Checklist

- [ ] Classify every material business into regulated network / contracted generation / merchant generation / trading / EPC-manufacturing before computing anything.
- [ ] Refuse to report consolidated OPM, ROCE, D/E and FCF for a mixed-asset utility; state why in the report.
- [ ] Recompute margin on **net revenue** (ex fuel and purchased power) and as **EBITDA per unit sold**.
- [ ] Identify the regulatory regime, the current control period, its start and end, and what the next reset is expected to do.
- [ ] Establish rate base / RAB and its multi-year growth; check how much capex was actually **admitted** into it.
- [ ] Compute allowed ROE vs earned ROE on regulated equity, and explain every basis point of the gap.
- [ ] Pull the operational KPIs against the **norm**: availability/NAF, PLF, CUF, heat rate, auxiliary consumption, AT&C losses, transmission availability, SAIDI/SAIFI.
- [ ] For distribution: AT&C losses, collection efficiency, ACS–ARR gap, and whether subsidy is budgeted and released.
- [ ] Map contracted vs merchant mix by MW and by EBITDA, weighted-average PPA life, counterparty by name, and hedge ratio year by year.
- [ ] Replace D/E with FFO/Debt, FFO interest cover, Net Debt/EBITDA and Debt/RAB; split recourse from non-recourse and re-add hybrids treated as equity.
- [ ] Check receivable days by counterparty and the >90-day ageing bucket; test whether any DSO improvement came from factoring, securitisation or a bailout scheme.
- [ ] Read the regulatory deferral account / regulatory asset note in full: balance, trend vs PAT, approved amortisation schedule, disallowance history.
- [ ] Age the CWIP, check the capitalisation rate and capitalised IDC, and flag anything sitting 4+ years.
- [ ] Compute capex/depreciation; negative FCF with capex/depreciation of 1.3–2.0x is a growth signal, not distress — say so explicitly.
- [ ] Rebuild EBITDA excluding late-payment surcharge, other income, grant and consumer-contribution amortisation, and regulatory-account movements.
- [ ] Check depreciation policy and asset-life assumptions against PPA tenor and decarbonisation timelines; quantify stranded-asset exposure.
- [ ] Verify the pipeline stage by stage — awarded / PPA+PSA signed / financially closed / under construction — and back out the implied equity IRR on recent wins.
- [ ] For merchant capacity, run mid-cycle and stressed spreads through the actual heat rate; never capitalise peak-cycle EBITDA.
- [ ] Value sum-of-the-parts: RAB or P/B for regulated, project DCF or EV/MW for contracted, low mid-cycle EV/EBITDA for merchant; deduct holdco debt, contingent liabilities and a holdco discount.
- [ ] Cross-check with EV/MW against replacement cost, and dividend yield against the 10-year sovereign.
- [ ] Test distribution coverage on FFO less maintenance capex; for InvITs/YieldCos, decompose NDCF and identify return of capital.
- [ ] India: read the CARO report, related-party note (fuel, EPC, O&M), promoter pledge disclosure, contingent liabilities, and the latest tariff order and true-up petition status.
- [ ] Check long-tail provisions — ash pond, mine reclamation, decommissioning, pension — and the discount rates applied.
- [ ] Compare actual renewable/hydro generation to P50/P90 over 3–5 years; quantify curtailment and whether deemed generation was actually paid.
- [ ] Peer set: same regulatory regime, sub-sector, contracting structure, technology and consolidation basis — stated explicitly, 5–8 names.
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: 115kUpdated Claude CodeCodexVerifiedInstall
add-runner-evalpaperclipai/paperclipAdd or extend a Paperclip Runner protocol evaluation definition, roster, assertion, or report fixture with provenance and narrow validation.Stars: 115kUpdated Claude CodeCodexVerifiedInstall
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: 115kUpdated Claude CodeCodexVerifiedInstall
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: 115kUpdated Claude CodeCodexVerifiedInstall