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/insurance-brokers-services.md

View on GitHub
# Insurance brokers, TPAs and distribution — sector playbook

Use this when: the company **places, distributes or administers** insurance but does not underwrite it — retail commercial and personal-lines brokers, employee-benefit consultants, reinsurance brokers, wholesale brokers, MGAs/MGUs/coverholders, third-party administrators (TPA), claims and health-benefit administrators, insurance marketing firms, corporate agents, PoSP aggregator platforms and online insurance distributors.

**This is not an insurer. Do not route it to `insurance.md`.** A broker takes **no underwriting risk**, holds **no float**, and therefore has **no combined ratio, no solvency ratio, no reserve development, no persistency, no claims ratio, no embedded value and no VNB**. Those metrics do not exist for this company; if a data provider supplies one, it has mis-classified the entity. What exists instead is a commission-and-fee business with negligible fixed capital, high incremental margins and a working-capital profile dominated by money that belongs to other people. Economically this is a **high-margin professional-services roll-up** that happens to be paid as a percentage of somebody else's premium — closer to an asset manager or a staffing/consulting compounder than to a carrier.

Two facts govern everything below. First, revenue is a *percentage of premium*, so the broker's top line rides the underwriting cycle without taking any of its risk — a soft P&C market shrinks revenue with zero loss of clients or volume. Second, most listed players in this space are serial acquirers, so reported growth, reported margin, reported capital returns and reported book value are all artefacts of acquisition accounting until you separate the organic business from the bought one.

## 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

Unlike banks and insurers, most generic ratios are *computable* here — which is precisely the trap. They compute, they look plausible, and several of them are systematically wrong in a direction that flatters or punishes the wrong companies.

**Headline revenue growth — computable and near-worthless on its own.** A roll-up's reported growth is organic + acquired + FX + fiduciary investment income + disposals, four of which say nothing about the business. Never quote total revenue growth without the organic decomposition; if the company does not disclose organic growth on a consistent basis, say so and treat growth as unverified.

**"Revenue" itself is not one thing.** Some brokers report **net revenue** (commission and fees retained), others report gross billings or gross written premium handled, and wholesalers/MGAs vary in whether they gross up amounts paid away to sub-brokers and producers. Under Ind-AS 115 / ASC 606 the broker is an *agent*, not a principal, so the premium collected is not revenue — but the agent/principal call on sub-broker arrangements still differs across companies. Confirm the base before comparing any margin, because a gross-revenue reporter will show a structurally lower margin on identical economics.

**OPM / GAAP operating margin — mis-signed against acquirers.** Amortisation of acquired customer relationships and non-compete intangibles is a very large, entirely non-cash charge for a roll-up and zero for an organically grown peer. GAAP EBIT margin therefore ranks the acquirer below the organic operator on identical cash economics. The sector's own convention is **EBITDAC** (see below); use it, and also look at the GAAP number, but never compare the two forms across peers.

**EV/EBITDA — the right family of multiple, but both terms are usually built wrong.** Enterprise value must **add** deferred consideration and earnout/contingent-consideration liabilities (these are debt-like obligations to former owners, often several percent of market cap and frequently omitted by screeners) and must **not** net off fiduciary cash. Fiduciary cash is premium held in trust for insurers and clients; it is legally not the company's money and cannot repay debt. Netting it produces a fictitious low net-debt figure and a fictitious low EV. Screener EV for this sector is wrong far more often than it is right — rebuild it.

**D/E and equity-based leverage — broken at both ends.** After a decade of debt-funded acquisition, book equity is mostly goodwill, and many roll-ups run *negative tangible net worth*. D/E is then either meaningless or literally negative. The binding constraint is cash-flow leverage: **net debt / EBITDAC**, with earnouts included and covenant definitions read directly from the credit agreement (they usually permit generous "run-rate synergy" add-backs that make covenant leverage look 0.5–1.5x lower than the honest number).

**ROCE / ROE — distorted in opposite directions by the same fact.** Capital employed for an acquisitive broker is overwhelmingly goodwill from past deals, so ROCE measures the price paid for past books, not the return on the operating business. For an organically grown broker, capital employed is near zero or negative and ROCE explodes to a meaningless number. Neither reading is comparable. Use **return on tangible invested capital measured against cumulative acquisition spend** as the real test of whether the roll-up creates value.

**Current ratio, working capital, receivable and payable days — contaminated and undefined.** The balance sheet is grossed up by fiduciary assets and an almost exactly offsetting fiduciary liability, which pins the current ratio near 1.0x by construction regardless of financial health. Premiums receivable from clients and payable to insurers are pass-through items, not trade working capital; DSO computed on them measures the insurance settlement calendar, not the company. Compute working-capital metrics on own-account receivables (fee billings) only, or not at all.

**Free cash flow — genuinely meaningful here, unlike in banks and insurers, but with two traps.** Capex is typically 1–2% of revenue, so FCF should track EBITDAC closely; a persistent gap is the single most informative anomaly in this sector. Trap one: fiduciary balances swing operating cash flow with premium seasonality and rate moves, so use FCF before fiduciary movements. Trap two: earnout payments are split between operating and financing cash flow under both IFRS and US GAAP (the portion above acquisition-date fair value goes to operating), so acquisition cost leaks into "operating" cash flow in ways that vary by company.

**P/E — usable but two-headed.** Reported GAAP EPS is depressed by intangible amortisation and distorted by fair-value remeasurement of earnout liabilities (a *gain* when acquired businesses underperform their targets — an earnings tailwind produced by bad news). The sector quotes "adjusted" or "cash" EPS adding back amortisation; that is directionally right but must be interrogated, because customer relationships genuinely do churn and retaining them costs real money.

**P/B and book value — meaningless, and this is the exact reverse of the underwriter case.** For a carrier, book value is economic capital and the primary valuation anchor. For a broker, there are no reserves, no investment portfolio and no regulatory capital worth speaking of; the assets are client relationships, producer teams, binding authorities and data. Book value is an accounting residue of deal prices. Do not compute P/B; do not treat negative tangible equity as distress by itself.

**Anything imported from `insurance.md` — undefined.** Combined ratio, loss ratio, solvency/RBC ratio, reserve triangles, persistency, VNB margin, embedded value, float, investment leverage. State explicitly in the output that these do not apply. The one exception: for an **MGA/MGU with delegated underwriting authority**, the *carrier's* loss ratio on the delegated book is not the MGA's P&L but is the key leading indicator of whether the binder gets renewed — track it as a business-continuity metric, not a profitability metric.

## The metrics that actually matter

Ranges are **indicative only**. They vary by market, sub-sector, point in the pricing cycle, interest-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 5–10 year record overrides every absolute band below.

| Metric | Definition / how to compute | Indicative healthy range | Why it matters |
|---|---|---|---|
| **Organic revenue growth** | Total revenue growth excluding acquisitions and disposals (typically excluded for the first 12 months of ownership), excluding FX, and — for the honest version — excluding fiduciary investment income. Best practice is to reconcile: reported growth = organic + M&A + FX + fiduciary income + other. | Soft market 2–5%; hard market 7–12%; sustained 6–8%+ through a full cycle is best-in-class. Below system premium-rate growth is share loss. | **The single most watched number in the sector**, and the one management has most incentive to shape. It is also the only figure that separates a genuine franchise from a purchasing programme. Definitions are not standardised: check whether acquired revenue is folded into the organic base after 12 months (standard) or earlier; whether "same-store" excludes lost accounts; whether large one-off placements, book-of-business transfers between subsidiaries or the annualisation of a mid-year acquisition are counted; and whether the definition changed. A definition change in a weak quarter is a governance flag, not an accounting detail. |
| **Revenue mix: commission vs fee vs contingent/supplemental** | Split net revenue into base commission (% of premium, paid by insurer), client fees (fixed or hourly, paid by the client, common in large-account and benefits consulting), contingent/supplemental commissions, and fiduciary investment income. | Retail commercial: 60–80% commission. Large-account and benefits consulting can be 40–60% fee. Fee-heavy mix is more defensive; commission-heavy mix is more cycle-levered. | Determines cycle sensitivity, which is the whole game here. Commission scales with premium **rate**, so a fee-based consultant keeps revenue flat through a soft market while a commission-based broker loses 5–10% of revenue without losing a single client. Fee revenue is also higher quality (client-negotiated, disclosed, transparency-proof) but usually lower margin because it is people-delivered. A shift from commission to fee is often a mix effect of moving upmarket, not a deterioration — check which. |
| **Contingent and supplemental commissions** | Profit-sharing paid by insurers based on the volume, growth and **loss experience** of the book placed, plus fixed-rate "supplementals" agreed in advance. Express as % of total revenue and as % of EBITDAC. | 3–8% of revenue for a typical retail broker; higher for wholesale and MGA models. High conversion to profit (near-100% incremental margin). | Disproportionately important because it is almost pure margin — a company where contingents are 6% of revenue may be earning 15–20% of EBITDAC from them. But it is the lowest-quality revenue line: it depends on the *carrier's* loss experience, so it collapses after a catastrophe year or an adverse loss trend, arrives lumpily (often concentrated in one quarter), and carries a permanent conflict-of-interest and regulatory-disclosure risk. Strip it out and re-run organic growth and margin without it. Rising contingents masking flat base commission is a classic quality-of-earnings problem. |
| **EBITDAC margin** | Earnings Before Interest, Tax, Depreciation, Amortisation and **Change in estimated acquisition earn-out payables**. The "C" is the industry convention: it removes the non-cash fair-value remeasurement of contingent consideration, which otherwise makes operating profit move when acquired businesses beat or miss their earnout targets. Most companies also report "adjusted EBITDAC" after further add-backs. | Global large-cap retail brokers 25–33%; strong mid-cap retail 20–28%; wholesale/MGA 30–40% (fewer service obligations, higher revenue per head); reinsurance broking 25–35%; TPAs and benefit administrators 10–20%. | This is the sector's replacement for OPM and the denominator of nearly every valuation multiple used here, so getting the definition right decides the answer. Understand *why* the convention exists — without the "C", a company whose acquisitions are failing books an earnout write-back as operating income. Then police the gap between EBITDAC and **adjusted** EBITDAC: transaction costs, "integration costs" that recur every year, restructuring, and run-rate synergies not yet earned. Recurring add-backs are operating costs. Margin should expand slowly with scale; a sharp jump usually means mix (a wholesale or MGA acquisition) rather than efficiency. |
| **Compensation ratio** | Employee compensation and benefits (including producer commissions, share-based pay and incentive accruals) / net revenue. | Retail broking 50–60%; wholesale/MGA 40–50%; benefits consulting 55–65%. Stability matters more than level. | This is a people business; comp is 55–70% of the entire cost base, so the comp ratio *is* the margin story. It also reveals the bargaining balance between the firm and its producers: a falling comp ratio can mean scale and better back-office leverage, or it can mean the firm is underpaying producers who will leave with their books. Read against producer retention. Check share-based compensation separately — excluding it from "adjusted" margin while issuing it every year overstates profitability. |
| **Fiduciary investment income and fiduciary asset balance** | Interest earned on premiums held in trust between client collection and remittance to the insurer. Track the average fiduciary asset balance, the realised yield, and this income as % of revenue and % of EBITDAC. | Balance is a function of premium volume and settlement terms. Income can move from ~1% of revenue at zero rates to 5–8%+ at high short rates, and drops close to 100% to EBITDAC. | The purest rate-sensitivity in the business and a frequent source of analytical error. It is not float — the broker holds it briefly, cannot invest it in risk assets, and in most regimes must keep it segregated in trust. It carries almost no cost, so it converts to profit at nearly 100% and can single-handedly manufacture a margin expansion story in a hiking cycle and reverse it in a cutting cycle. Best-practice disclosure separates it from organic revenue; if the company does not, do it yourself and restate. Also check where the balances sit — bank counterparty concentration on fiduciary trust accounts is a real, under-discussed risk. |
| **Client retention rate** | Revenue retained from the prior-year client base, measured on revenue not client count. Distinguish gross retention (before new business, before rate) from net revenue retention (including rate, exposure and cross-sell on retained clients). | Commercial P&C 90–95%; large-account and benefits 92–96%; personal lines 85–92%; MGA program business more volatile. Below ~88% on commercial needs an explanation. | Retention is the compounding engine: at 95% retention a broker keeps the same book for two decades and every new placement is additive; at 85% it is on a treadmill. Because renewal commission requires almost no incremental work, retained revenue carries far higher incremental margin than new business. Retention is also the direct read on whether an acquisition worked — books bought from a departing founder frequently leak 10–20% over three years, which shows up nowhere in "organic growth" if the acquired unit is excluded from the base. |
| **Producer productivity, headcount and recruitment** | Revenue per producer; the vintage curve (producers typically take 3–5 years to become net contributors); validated vs unvalidated producer counts; annual producer hires, departures and lift-outs of teams from competitors; non-solicit/non-compete enforceability by jurisdiction. | Producer count growing with revenue; a rising share of validated producers; low involuntary and voluntary attrition among top-decile producers. | The book of business walks out of the door on two legs. Producer economics are the leading indicator of organic growth two to three years ahead: a firm that stops hiring saves money now and prints weak organic growth later, which is a favourite pre-sale cosmetic. Conversely, heavy hiring depresses near-term margin for the right reason. Team lift-outs (in either direction) and the litigation that accompanies them are the sector's most visible competitive event — track both sides. In the US, state-level non-compete enforceability materially changes book-portability risk. |
| **Revenue per employee** | Net revenue / average total headcount. Compare only within the same market and sub-sector — the absolute figure is not comparable between a US broker and an Indian one. | Rises with mix toward wholesale/MGA and large accounts; falls with mix toward claims administration, TPA work and high-touch personal lines. Track the 5-year trend. | The cleanest single proxy for operating leverage in a people business, and the one that shows whether technology and offshoring spend is actually working. It is also the honest test of a "digital distribution" claim: a platform that genuinely automates placement should show revenue per employee climbing materially; one that has simply bought a call centre will not. Watch for offshore/GCC headcount shifting the ratio for cost reasons rather than productivity reasons. |
| **Free cash flow conversion** | (Operating cash flow before fiduciary movements − capex) / EBITDAC. Also compute cash EPS conversion and check the split of earnout payments between operating and financing cash flow. | 70–90% of EBITDAC for a healthy broker; capex 1–2% of revenue. Persistent conversion below ~60% needs a cause. | Because capex is trivial and there is no inventory or genuine trade working capital, an honest broker's EBITDAC should turn into cash almost immediately. A sustained gap is therefore diagnostic rather than incidental: the usual causes are capitalised commissions or capitalised software, earnout payments buried in operating cash flow, growing own-account receivables from disputed fees, cash interest on acquisition debt exceeding the picture EBITDAC paints, or add-backs that are actually cash costs. This is the metric that catches an over-adjusted EBITDAC. |
| **M&A programme economics** | Deals per year, annualised revenue acquired, total consideration, implied EV/EBITDA paid, share of consideration in earnouts and rollover equity, earnout liabilities on the balance sheet as % of market cap, and the retention of acquired books at 24 and 36 months. | Tuck-ins historically 6–12x EBITDA versus a listed acquirer at 12–20x; earnouts 15–35% of consideration. Both ends move with credit conditions and PE competition. | The entire equity story of most listed names in this sector is **multiple arbitrage**: buy at 9x, consolidate into a business rated 16x, and EPS rises by arithmetic alone with no operational improvement whatsoever. That is real value but it is finite, mechanical and reverses instantly if the buying multiple rises or the rating falls — so it must be *scored separately from operating quality*, never blended into "growth". The diagnostic question is whether acquired books retain and grow after year two. Also watch the direction of purchase multiples: paying up late in a cycle for books whose commissions are inflated by hard-market rates is buying peak earnings at a peak multiple. |
| **Leverage: net debt / EBITDAC** | Net debt **including** deferred and contingent consideration, finance leases and any seller notes, **excluding** fiduciary cash from the cash offset. Compare to the covenant definition in the credit agreement, which is usually more permissive. | Listed acquisitive brokers commonly 2.5–4.5x; PE-owned roll-ups 5.5–7.5x and higher. Above ~5x with organic growth below 4% is a fragile combination. | Serial acquisition funded with debt is the standard model here and moderate leverage is not distress — but leverage is what converts a soft-market revenue decline into an equity problem. Because revenue falls in a soft market with no volume loss and the cost base is largely fixed producer comp, EBITDAC is more operationally geared than the top line suggests. Model a 10% revenue decline through to covenant headroom. Also check the maturity wall, floating-rate share (many roll-ups are heavily floating and were rescued on the revenue line by the same rates that raised their interest cost), and whether earnouts fall due in the same window. |
| **Goodwill and intangibles vs tangible net worth; ROTIC** | Goodwill + acquired intangibles as % of total assets and of equity; tangible net worth; and return on tangible invested capital = post-tax EBITA / cumulative cash spent on acquisitions plus organic tangible capital. | Goodwill + intangibles of 60–80% of assets is normal, not alarming. Negative tangible equity is common and not by itself a warning. ROTIC comfortably above WACC and *rising* is the test. | Because P/B is meaningless here, the balance sheet's only real analytical use is to answer one question: did the money spent on acquisitions earn a return? ROTIC against cumulative deal spend is the discipline that a roll-up's own adjusted EPS bridge is designed to avoid. Also track goodwill by cash-generating unit and the headroom disclosed in the impairment note — impairment is the sector's delayed admission that a book was overpaid for or has run off. |
| **Premium rate exposure: organic growth decomposition** | Split organic growth into (a) premium **rate** change on renewed business, (b) **exposure units** — insured values, payroll, headcount, vehicle count, revenue of the client, (c) **net new business** won minus lost, and (d) commission-rate change. Benchmark (a) against a published commercial P&C rate index for the relevant market and lines. | In a hard market, rate can be most of reported organic growth; in a soft market it is negative and net new business must carry everything. | This is the decomposition that tells you whether you are looking at a good business or a good market, and it is the reason a broker must never be judged on organic growth in isolation. A broker printing 9% organic growth while market rates are up 10% is losing share; one printing 4% while rates are down 3% is winning. Exposure-unit growth also makes brokers a structural inflation beneficiary — nominal wage and property-value inflation raises insured values and therefore commission with no new clients. Commission-rate change is the silent line: carriers cut commission percentages when their own margins compress. |
| **Wholesale / MGA / MGU-specific: delegated authority and carrier concentration** | Share of revenue from binding authority and program business; number of capacity providers and top-3 carrier concentration; contract length and termination notice on each binder; the *carrier's* loss ratio on the delegated book; share of revenue from E&S (excess & surplus / non-admitted) lines. | Top-3 capacity concentration below ~50% of delegated revenue is comfortable; binder terms of 3+ years with limited termination rights are stronger than annual rolling ones. Carrier loss ratio on the program at or below the carrier's target. | A wholesale broker or MGA is a fundamentally different business from a retail broker and must never be blended into the same peer set. It earns higher margin and higher commission rates because it takes on underwriting selection, pricing and sometimes claims handling as a delegated service — but its revenue depends on a small number of carriers continuing to lend it their balance sheet. If the delegated book runs a poor loss ratio, capacity is withdrawn and the revenue line disappears in a single renewal cycle, regardless of client retention. E&S exposure also makes the model *more* cyclical, not less: E&S volume surges when standard markets contract and drains back when they reopen. |
| **TPA / benefits administration: lives, claims and contract concentration** | Lives or members administered; claims processed per period; fee basis (per-employee-per-month, per-claim, or % of premium); revenue per life; contract terms and renewal dates; top-5 client/carrier concentration; auto-adjudication rate and claims turnaround time; service-level penalties. | Client concentration above ~20% from one carrier or scheme is a material risk. Auto-adjudication rate rising; turnaround and grievance ratios improving. | TPAs and health-benefit administrators are a **contract services** business, not a broking business: revenue is per-life or per-claim, margins are structurally lower (10–20%), and the customer is often the insurer or a self-funded employer rather than a retail client. Their existential risk is not the pricing cycle but **in-sourcing** — insurers repeatedly decide to bring claims administration back in house, and a single contract loss can remove a double-digit share of revenue with 90 days' notice. Operational quality metrics (auto-adjudication, turnaround, grievance and repudiation-related complaints) are the leading indicator of contract renewal and, in India, of regulatory attention. |

## How to value companies in this sector

This is an asset-light, cash-generative, low-capex services business. Cash-flow-based and earnings-based methods work; balance-sheet methods do not.

**Primary — EV/EBITDAC (and EV/adjusted EBITDA).** The sector's own currency, and the multiple at which the businesses actually change hands in private M&A, which gives it an external validation the public multiple alone does not have. Build both sides properly: EV = market cap + gross debt + deferred and contingent consideration + finance leases − **own** cash, with fiduciary cash excluded from the cash offset. If the company reports both EBITDAC and adjusted EBITDAC, value on the less-adjusted figure and show the difference. Indicative ranges: global large-cap brokers have historically commanded high-teens to low-twenties EV/EBITDA in strong markets, mid-cap consolidators low-to-mid teens, TPAs and administration businesses high single digits to low teens — but these are cycle- and rate-dependent and must be re-anchored to the current peer set.

**Primary — P/E on a defined earnings figure, computed twice.** Compute GAAP P/E and adjusted/cash P/E (adding back acquisition-intangible amortisation and earnout remeasurement) and present both. Then take a view on the add-back: amortisation of customer relationships is non-cash, but if acquired books shrink and must be replaced with more acquisitions, the amortisation is proxying for a real, recurring cost of maintaining revenue. The test is empirical — check whether acquired books grow organically at 24–36 months. If they do, the add-back is fair; if they run off, cash EPS is overstating durable earnings.

**Primary — free cash flow yield and a DCF.** Because capex is trivial and working capital is largely pass-through, this is one of the few financial-sector-adjacent businesses where a straightforward FCFF or FCFE DCF is legitimate and informative. Model it explicitly with organic growth and acquisition spend separated: forecast organic revenue growth, EBITDAC margin, cash tax, cash interest and maintenance capex, then treat acquisition consideration (including earnout payments) as a **separate, explicit investing line** with its own return assumption. Never fold acquisitions into a single blended growth rate — that assumes free, infinite, constant-multiple deal supply, which is the sector's standard forecasting error.

**Sum-of-the-parts where the mix is genuinely mixed.** Retail broking, wholesale/MGA, reinsurance broking and TPA/administration deserve different multiples because they have different margins, different cycle sensitivity and different customer-concentration risk. A group that is 70% retail and 30% administration should not be valued on one blended multiple.

**Cross-checks.** EV/revenue is a useful sanity check *only within a tight sub-sector* given the high and fairly narrow margin band. Private-market transaction multiples for comparable books (widely reported in broker M&A commentary) anchor whether the public multiple is stretched. A reverse-valuation test is particularly powerful here: solve for the organic growth rate and the annual acquired-revenue run rate implied by today's price, then ask whether the deal pipeline and balance-sheet capacity can actually supply it.

**Do not use:**
- **Price-to-book and price-to-tangible-book.** No reserves, no investment portfolio, no regulatory capital base; book value is a residue of deal prices and is frequently negative on a tangible basis. This is the exact inverse of the underwriter case in `insurance.md`, where book value *is* the valuation anchor — and the single most common analytical error made with these companies.
- Any insurer metric: embedded value, P/EV, VNB multiple, price-to-GWP, combined ratio, solvency ratio.
- ROCE-based valuation screens, for the reasons in the section above.
- A blended growth-based PEG on total (organic + acquired) growth.

## Peer set construction

A valid comparable shares the **revenue model, cycle exposure and regulatory regime** — not merely the label "insurance". Mixing across these lines produces confidently wrong conclusions.

**Splits that must never be mixed:**

- **Brokers/intermediaries vs insurers.** Non-negotiable, and the reason this playbook exists. An intermediary and a carrier share a customer and nothing else — no risk, no reserves, no capital regime, opposite balance-sheet logic, opposite valuation anchor.
- **Retail broking vs wholesale / MGA / MGU.** Different margin structure (wholesale is materially higher), different customer (the retail broker is the wholesaler's client), different risk (carrier capacity withdrawal vs client attrition), different cyclicality (E&S flow is counter-cyclical to standard-market capacity).
- **Commercial P&C broking vs employee benefits vs personal lines vs reinsurance broking.** Fee/commission mix, retention, cycle sensitivity and regulatory exposure all differ. Reinsurance broking is a small-client-count, very high revenue-per-head, relationship-concentrated business that behaves nothing like retail.
- **Brokers vs TPAs and claims/benefit administrators.** Fee-per-life or fee-per-claim contract services with 10–20% margins, in-sourcing risk and client concentration — not a commission business. Never blend the two into a single margin comparison.
- **Serial acquirers vs organic-only operators.** Their reported growth, margin, EPS, capital returns and balance sheets are constructed differently. Compare them only on **organic growth, EBITDAC margin and FCF conversion**, never on GAAP EBIT margin, ROCE, P/B or reported EPS growth.
- **Listed intermediaries vs PE-owned consolidators.** PE-owned platforms run 5.5–7.5x leverage with aggressive covenant add-backs and often disclose only under bond documentation. Their EBITDA definitions are usually not comparable to a listed peer's.
- **India: IRDAI-licensed brokers vs corporate agents vs web aggregators vs insurance marketing firms vs PoSP-led platforms.** These are separate licence categories with different permitted activities, different tie-up limits and different commission economics (see below).
- **Digital-first distribution platforms vs traditional brokers.** A platform carrying heavy customer-acquisition marketing spend, negative EBITDA in growth phase and a take rate on low-ticket retail policies is closer to a consumer-internet business than to a commercial broker; consider whether `exchanges-payments.md` or `it-saas.md` unit-economics logic (CAC payback, cohort retention, contribution margin) should be run alongside this playbook.

**Also align:** market and pricing cycle (a US broker and an Indian broker are at different points of a different P&C cycle); accounting regime (Ind-AS vs IFRS vs US GAAP, and whether revenue is reported gross or net); fiscal year (India April–March); interest-rate environment, because fiduciary income is a large and non-comparable margin component across rate regimes; and size band, since scale drives comp ratio, carrier leverage and access to national accounts.

Aim for 5–8 peers. State the basis explicitly, and benchmark every metric twice — against peers *and* against the company's own multi-year record.

## Sector-specific red flags

- **The organic growth definition changes, or stops being disclosed.** The most important number in the sector is also the least standardised. Watch for acquired revenue entering the organic base early, "same-store" definitions that quietly exclude lost accounts, mid-year acquisitions annualised into the base, book transfers between subsidiaries counted as new business, and fiduciary investment income left inside organic revenue during a hiking cycle. A definition change coinciding with a slowdown is a governance flag.
- **Growth entirely from M&A while organic decelerates.** The classic roll-up decay pattern: as the base grows, each tuck-in moves the needle less, so deal size and purchase multiples rise, leverage rises with them, and the arbitrage narrows. Chart organic growth and acquired revenue separately over five years; if organic is trending toward zero, the equity story is a financing structure, not a business.
- **Acquired book run-off hidden by the exclusion window.** Books bought from a retiring founder commonly leak clients once the earnout period ends and the seller's non-compete expires. Because acquired revenue sits outside the organic base for 12 months and inside it thereafter, the leakage shows up as mysteriously weak organic growth two to three years after a deal-heavy period. Ask for retention of acquired books at 24 and 36 months.
- **Earnout accounting used as an earnings lever.** A downward remeasurement of contingent consideration is a P&L *gain* generated by an acquisition **failing**. The EBITDAC convention exists to strip it out — verify it actually has been, both in the operating line and in adjusted EPS.
- **Adjusted EBITDAC add-backs that recur.** "Integration costs", "transaction costs", restructuring and "one-off" legal charges appearing every year for five years are operating costs. Track the cumulative gap between GAAP operating profit and adjusted EBITDAC; the FCF conversion metric is the referee.
- **Margin expansion sourced from fiduciary investment income or contingents rather than operations.** Both drop nearly 100% to profit and both are outside management's control. Restate margin excluding them before concluding anything about efficiency. The corollary risk is symmetric and under-appreciated: a rate-cutting cycle removes this margin with no operational cause.
- **Contingent commissions rising as a share of revenue.** Higher dependence on carrier profit-sharing means earnings quality is deteriorating and a single bad catastrophe or loss-trend year will take it away. It also raises conflict-of-interest and disclosure risk — this is the exact practice that produced the major US broker contingent-commission enforcement actions of the mid-2000s and led to years of disclosure reform.
- **Producer attrition, team lift-outs and litigation.** Departure of top producers, a wave of poaching suits, or a sharp fall in unvalidated-producer hiring all precede organic weakness by two to three years. So does an abnormally falling comp ratio.
- **Carrier concentration and binder loss (wholesale/MGA).** Loss of a capacity provider, non-renewal of a binding authority, or deterioration in the loss ratio on a delegated program can remove a revenue block instantly. Read the capacity-provider disclosure and the program renewal calendar.
- **Client or carrier concentration and in-sourcing risk (TPA).** A single insurer or large self-funded scheme worth a double-digit share of revenue, on a short-notice contract, is an existential dependency. Insurers periodically bring claims administration in house.
- **Fiduciary funds handling.** Any indication of commingling client/insurer money with own funds, delayed remittance to insurers, use of fiduciary balances to fund operations, or a regulatory finding on trust-account segregation is a first-order integrity flag — it is the intermediary equivalent of a bank misusing depositor money, and it is how brokers actually fail.
- **Leverage plus soft market.** Net debt/EBITDAC above ~5x (including earnouts) with organic growth below 4% and a floating-rate debt stack is the configuration that turns a normal cyclical revenue decline into a restructuring. Test covenant headroom against a 10% revenue fall.
- **Goodwill impairment, or conspicuously thin headroom in the impairment note.** The delayed acknowledgement that books were overpaid for. Read the CGU-level disclosure, not the group total.
- **Revenue recognition timing.** Under IFRS 15 / ASC 606 placement commission is generally recognised when the policy is bound rather than spread over the policy period, which front-loads revenue and amplifies seasonality around major renewal dates. Watch for policy changes, for estimates of variable consideration (contingents accrued in advance), and for aggressive accrual of profit-sharing not yet confirmed by the carrier.
- **Capitalised commissions and capitalised software** growing faster than revenue — the most common cause of an EBITDAC-to-cash gap.
- **Regulatory and conduct exposure.** Commission-disclosure regimes, remuneration transparency rules, mis-selling and claims-handling penalties, and — in India — IRDAI inspection findings, licence-renewal conditions, and the periodic scrutiny of payouts routed through non-broking entities or group companies to sidestep commission and expense limits.

## Cycle and structural context

**The P&C pricing cycle is the sector's exogenous driver, and it moves revenue without moving volume.** Commission is a percentage of premium, so a hard market — rising rates after catastrophe losses, adverse reserve development or capital withdrawal — lifts broker revenue with essentially zero incremental cost, and a soft market compresses it with zero client loss. This is the most important sentence in this playbook after the "not an insurer" warning: **you cannot read a broker's revenue decline as commercial failure, or a revenue surge as commercial success, without the rate index.** Always decompose organic growth into rate, exposure and net new business, and always state where the local P&C market is in its cycle.

**Exposure units make brokers a structural inflation beneficiary.** Premiums scale with insured values, payrolls, headcount, fleet size and client revenue. Nominal inflation raises all of them, so commission rises with no new business at all. This is why brokers have historically compounded through inflationary periods that damaged most services businesses — and why real, volume-driven growth must be isolated before crediting management.

**Interest rates enter through fiduciary income, not through underwriting.** A rate cycle changes broker margin directly and quickly, with essentially no lag and no operational cause. A cutting cycle removes a margin tailwind that a hiking cycle handed over.

**Capital-cycle interaction with wholesale/MGA.** When carriers retrench, business flows to E&S and delegated-authority markets, benefiting wholesalers and MGAs; when capacity returns, it flows back to standard markets. So the wholesale sub-sector is counter-cyclical to standard-market capacity while still being levered to overall rate levels — a two-factor exposure that retail brokers do not have.

**Consolidation is the defining structural feature.** Insurance distribution is one of the most fragmented professional-services markets in the world, and private-equity capital has been an aggressive competing bidder for the same tuck-ins that listed consolidators buy. That competition sets purchase multiples, which sets the arbitrage, which sets the acquirers' incremental returns. Track the direction of private-market multiples as a sector-level input.

**Structural threats to score explicitly.** Carrier direct-to-consumer distribution in personal lines and small commercial, where the broker's advisory value-add is weakest. Commission-rate compression as carriers manage their own expense ratios. Regulatory transparency initiatives that expose remuneration to clients. Automation of low-touch placement and quote comparison. And in the other direction, structural supports: rising insurance penetration in emerging markets, growing complexity in cyber, climate, D&O and specialty lines where advice is genuinely valued, and the shift of employers to self-funded benefit structures that need administrators.

**For TPAs specifically, the cycle is contractual, not pricing-driven.** Their risk is the periodic decision by insurers to in-source claims, and their opportunity is regulatory or technological complexity that makes outsourcing cheaper. Model them on contract renewal dates and lives administered, not on premium rates.

## India vs global notes

| Dimension | India | US / global |
|---|---|---|
| Regulator and licence categories | IRDAI. Distinct registrations with distinct rules: **direct broker** (life / general / composite), **reinsurance broker**, **composite broker** — under the IRDAI (Insurance Brokers) Regulations, 2018, with minimum capital requirements scaled by category (broadly ₹75 lakh direct, ₹4 crore reinsurance, ₹5 crore composite) and net-worth maintenance obligations. Separately: **corporate agents**, **web aggregators**, **insurance marketing firms**, **motor insurance service providers (MISP)** and **point-of-sales persons (PoSP)**. Licences are periodically renewed and can be conditioned or suspended | US: state-by-state producer and surplus-lines licensing, no single federal regulator; NAIC model rules. UK: FCA authorisation, with client-money (CASS 5) rules governing premium trust accounts. EU: IDD (Insurance Distribution Directive) with demands-and-needs and remuneration-disclosure requirements |
| Broker vs agent distinction | Legally central. A **broker represents the client** and may place with any insurer; a **corporate agent represents the insurer** and may tie up with a limited number of insurers per line (broadly up to three each in life, general and health under the open-architecture rules). This single distinction determines the revenue model, the conflict-of-interest profile and the competitive set | The retail/wholesale/MGA distinction matters more than broker/agent; independent agents and captive agents coexist, and MGAs hold delegated underwriting authority under binding agreements |
| Commission regulation | Historically product-level commission caps. The **Expenses of Management (EOM) regulations effective FY2024** replaced product-wise commission caps with an overall board-approved EOM ceiling at the insurer level, giving insurers flexibility on what they pay intermediaries within that limit — a material change to the sector's revenue mechanics and a live source of channel-by-channel commission repricing | Commission rates are market-determined; disclosure regimes (US state rules post the mid-2000s contingent-commission settlements, UK/EU under IDD) govern transparency rather than level |
| Bancassurance | A dominant channel, particularly for life. Banks act as **corporate agents**, distributing for a small number of tied insurers — frequently a group insurance company — which raises persistent mis-selling and captive-distribution concerns and has drawn both IRDAI and RBI attention. Bancassurance economics compete directly with independent broking for the same premium pool, so treat it as part of the competitive analysis for any Indian distributor | Bank distribution exists but is a smaller share of the market in the US; more significant in parts of Europe and Asia |
| Digital distribution | Web aggregators are a licensed category; several online platforms have converted to full **direct broker** licences to widen permitted activity. **Bima Sugam**, the IRDAI-sponsored electronic insurance marketplace, is a structural disintermediation and margin question for every online distributor — model it as a scenario, not a certainty. PoSP networks are the standard way digital brokers scale low-ticket distribution | Online comparison and digital MGA models are unlicensed-category-free but subject to state producer licensing; no central state-sponsored marketplace equivalent |
| TPAs | Licensed under the IRDAI (Third Party Administrators – Health Services) Regulations, 2016; paid a percentage of premium or a per-claim/per-life fee. Structural pressure from insurers building in-house health claims capability, and from regulator focus on claims turnaround, repudiation and grievance metrics | Health-benefit administration is dominated by ASO (administrative services only) arrangements for self-funded employers, plus pharmacy benefit and specialty administrators; PEPM fee structures are the norm |
| Listed universe and disclosure | Thin and skewed toward online distribution; most traditional brokers and TPAs are unlisted, promoter-owned or PE-owned, so peer sets are hard to build and often require unlisted or cross-border comparables — say so explicitly. Ind-AS 115 governs agent-vs-principal revenue presentation. Reporting in ₹ crore/lakh, April–March fiscal year, quarterly results plus concall. Check promoter holding and pledge, related-party flows to group insurers or group distribution entities, and CARO qualifications | Deep listed universe across global, national and specialist brokers. 10-K/10-Q on EDGAR; organic growth, EBITDAC and adjusted EPS are standard non-GAAP disclosures with mandated reconciliations — read the reconciliation tables, not the headline |
| Taxes and other | GST at 18% on broking commission and services; TDS obligations on commission payouts; sub-broker/PoSP payouts are a large pass-through cost line to check for gross-vs-net presentation | Sales tax generally not applicable to commissions; surplus-lines taxes and stamping fees apply in the US E&S market and are collected by the wholesaler |

## Checklist

- [ ] Confirm the company takes **no underwriting risk** — if it does on any material book, route that part to `insurance.md` and analyse separately.
- [ ] State explicitly in the report that combined ratio, solvency ratio, reserve development, persistency, claims ratio, float, embedded value and VNB do not exist for this company.
- [ ] Delete P/B, P/TBV, ROCE and receivable/payable-day metrics; explain why book value is not an anchor for an asset-light intermediary.
- [ ] Establish whether revenue is reported gross or net, and whether sub-broker/PoSP payouts are grossed up, before comparing any margin.
- [ ] Get organic revenue growth, read the company's **definition** of it, check whether the definition has changed, and reconcile reported growth = organic + M&A + FX + fiduciary income.
- [ ] Decompose organic growth into premium **rate**, **exposure units** and **net new business**; benchmark rate against a market P&C rate index and state where the cycle is.
- [ ] Split revenue into base commission, client fees, contingent/supplemental commissions and fiduciary investment income; restate growth and margin excluding the last two.
- [ ] Compute EBITDAC margin; then compute the gap to **adjusted** EBITDAC and list every add-back, flagging anything that recurs annually.
- [ ] Track fiduciary asset balance, realised yield and fiduciary income as % of EBITDAC; model the effect of a 200 bps rate move in each direction.
- [ ] Check client retention (revenue-weighted), and separately the retention of acquired books at 24 and 36 months.
- [ ] Review producer economics: revenue per producer, validated vs unvalidated mix, hiring, attrition, lift-outs and litigation; check the comp ratio trend.
- [ ] Compute revenue per employee over five years as the operating-leverage and automation test.
- [ ] Rebuild EV from scratch: add deferred and contingent consideration, **exclude fiduciary cash** from the cash offset. Never use a screener's EV here.
- [ ] Compute net debt / EBITDAC including earnouts, compare to the covenant definition, and stress a 10% revenue decline against covenant headroom.
- [ ] Score the M&A programme separately from operating quality: deals, revenue acquired, multiples paid, earnout share, and ROTIC on cumulative acquisition spend.
- [ ] Reconcile EBITDAC to free cash flow; investigate any conversion below ~60–70% (capitalised commissions/software, earnouts in operating cash flow, cash add-backs).
- [ ] Present GAAP and adjusted EPS side by side and take an explicit view on whether the intangible-amortisation add-back is economically justified.
- [ ] Value on EV/EBITDAC, P/E on defined earnings, and an explicit DCF with acquisition spend modelled as a separate investing line — never a single blended growth rate.
- [ ] Wholesale/MGA: check capacity-provider concentration, binder terms and renewal calendar, and the carrier's loss ratio on delegated programs.
- [ ] TPA/administration: check lives and claims administered, fee basis, top-5 contract concentration, renewal dates, and in-sourcing risk.
- [ ] Test fiduciary-funds integrity: segregation, remittance timing, any regulatory finding on trust accounts, bank counterparty concentration.
- [ ] India: identify the exact IRDAI licence category (direct/reinsurance/composite broker, corporate agent, web aggregator, IMF, PoSP-led) and its permitted activities and tie-up limits.
- [ ] India: assess EOM-regulation-driven commission repricing, bancassurance competition, Bima Sugam disintermediation risk, and related-party flows to group insurers or group distribution entities.
- [ ] Peer set: same sub-sector (retail / wholesale-MGA / reinsurance / TPA), same market and rate environment, same acquisitive-vs-organic profile, same revenue presentation — stated explicitly.
- [ ] Benchmark every metric twice: against the peer set and against the company's own 5–10 year history. No conclusion from a single metric.
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