The pen is portable

The skill is already entrepreneurial. Only the payslip isn’t. Here is the arithmetic of the leap.

Eleven papers in this series priced the running, growing and selling of a delegated-authority business. This one prices the decision that creates one. Inside every large insurer sit underwriters who already do the entrepreneurial part of the job: they hold an appetite in their heads, brokers ring them by name, and the portfolio they steer makes money for the balance sheet behind them. The market prices that skill in the millions. The payslip prices it at a salary (call it $250,000), and the difference compounds every year in someone else’s accounts. The question for that underwriter has never been ability. It has been arithmetic: how long until a firm of my own pays me what employment does, and what is the firm worth once it has?

For most of the modern market the honest answer was discouraging, because the toll was collected before the first policy was bound: a policy administration system built or bought over twelve to thirty months at $300,000 to $1 million and more, and an operations headcount hired ahead of revenue. The economics of starting were the economics of a mid-sized firm, imposed on a firm of one. That is the assumption that has expired. The machinery that made incumbents big — quote, bind, issue, bordereaux, reconciliation — is becoming something a new agency rents by the policy rather than builds in advance, which moves the cost of starting from capital expenditure to a line that scales with the book. When the toll falls, the arithmetic changes; the rest of this paper works it through.

The route itself is well trodden. Lloyd’s alone stands behind more than 4,000 coverholders across some eighty territories, with roughly two-fifths of its £57.9 billion of premium arriving through delegated arrangements. In Australia, underwriting agencies now write about $10.8 billion of premium — near enough one dollar in five of general insurance, up more than fourfold since 2014–15 — across roughly 300 agencies. Most of those agencies began the same way: an underwriter, an appetite, and a binder.

01 — The worked example

Take a concrete case, stated plainly so the assumptions can be argued with. An underwriter leaves a corporate seat on $250,000 and opens a specialty agency. The book writes $3 million of gross premium in year one, $6 million in year two, $10 million in year three, and grows to $20 million by year seven, held there. Gross coverholder commission is 27.5% of premium; 15% is paid away to producing brokers; the agency keeps 12.5% net, before any profit commission. The founder hires one underwriter-operator in year two and a second in year four — a firm of three at $20 million, which per-policy infrastructure makes an operating model rather than an aspiration. Every figure below is illustrative arithmetic on these stated assumptions: not advice, not a forecast, and deliberately conservative where it matters — profit commission, the usual sweetener of 15–25% of underwriting profit on a well-run binder, is excluded from the base case and shown only as upside.

Owner’s earnings = net commission revenue less all operating costs, before the founder pays themselves. Costs include staff (excl. founder), technology and data, compliance, PI, audit and office.

02 — The salary line

On these assumptions the crossing is early and specific. Year one is a pay cut and should be planned as one: roughly $120,000 of owner’s earnings against the $250,000 left behind, which is why the leap needs about a year of personal runway alongside the working capital. The line is crossed partway through year two, at roughly $5 million of annualised premium: the firm’s earnings pass the old salary, and every dollar of growth after that is upside employment never offered. By year three the founder earns roughly three times the corporate package; from year five onward, five times and more. The same ramp on the old cost base (the built system, the ops headcount hired with the book) crosses a year and more later and never earns within reach of the lean firm: at the $20 million plateau it delivers barely half the owner’s earnings from an identical book.

The cost of starting has always been the tax on underwriting entrepreneurship. What changed is that the tax is now largely optional.

03 — What the business is worth

Earnings are half the exit; the multiple is the other half, and — as the seventh paper in this series argued — the multiple is the buyer’s price for doubt. The market context is friendlier than most corporate underwriters suspect: 2025 transactions in delegated-authority businesses averaged 19.4× pro forma EBITDA, an all-time high, driven by platform-scale deals. A three-person agency will not command a platform multiple — small books carry key-person risk and capacity concentration, and buyers price both — so the honest planning band for a sub-$2 million-EBITDA firm is 6–12×, with the top of the band reserved for books whose records are self-evidently clean. On the worked example, after paying the founder a market salary:

Enterprise value, $m, on the stated assumptions. EBITDA is stated after a $250,000 market salary to the founder.

The last two rows matter most. The same $20 million book, run on a conventional cost base, produces $0.65 million of EBITDA and — carrying the diligence doubts of a manual firm — might transact at 8×: call it $5 million. Run lean on infrastructure that proves its own records, it produces $1.5 million and can argue for the top of the band: $15–18 million, and beyond it with profit commission. Infrastructure earns twice at exit — once in the earnings it doesn’t consume, and again in the doubt it removes. A five-year horizon offers an entirely respectable outcome; the ten-year figure is the one that reframes a career. Few salary negotiations close a gap measured in eight figures.

04 — What the arithmetic does not say

The numbers are necessary and not sufficient, and it would be a disservice to pretend otherwise. Capacity is the gate: a binder is granted on evidenced judgement, and a first-time founder is asking a market to price a track record that mostly lives in a former employer’s systems — the record you can lawfully prove is the record capacity will pay for, and assembling it takes longer than any software does. Employment contracts carry restraints that must be respected and planned around. Licensing — an AFSL or authorised-representative arrangement in Australia, coverholder approval at Lloyd’s — takes months and money, and both are in the worked example’s cost lines for a reason. One binder is also one counterparty: the concentration that makes the early firm simple makes it fragile, and diversification of capacity is as much a part of the ten-year plan as premium growth. None of this argues against the leap. It argues for making it the way underwriters make every other decision: eyes open, priced, and documented.

05 — The conclusion

This series has argued throughout that an insurance system is worth exactly what it does to four numbers: cost to start, cost to run, cost to connect, premium per person. This paper is those numbers read from the founder’s side of the table. Lower cost to start moves the leap from institutional to personal scale. Lower cost to run brings the salary line into year two instead of year four. Premium per person — three people, twenty million — is what turns a lifestyle practice into an asset. And infrastructure that proves its own records is what makes the asset payable, in cash, at the top of the band, on the day the founder decides the firm should outlive their ownership of it. The pen was always portable; now the machine it writes on is too.

Sources & basis. Lloyd’s market data (2025); APRA-published agency market statistics; MarshBerry 2025–2026 valuation update (May 2026) — delegated-authority transactions at 19.4× average pro forma EBITDA; MarshBerry 2024 — specialty M&A at 10–15×; MGA commission structures per published industry primers (base commission 10–25% of GWP plus profit commission). All firm-level figures are illustrative arithmetic on the stated assumptions and are not financial advice, a forecast, or an offer.

Cuttleflow Systems · Perspective 12 · 33°53′S · 151°16′E · Sydney