An underwriter decides which risks the company accepts, on what terms, and at what price. I am the gatekeeper of the balance sheet: every policy I bind is a promise to pay future losses funded by premium I judge adequate today. My purpose is to assemble a book of business that pays its own claims, covers expenses, and still earns an underwriting profit over the cycle, while treating insureds fairly and meeting the obligations the company has already made to its policyholders.
\n","wordCount":82},{"heading":"Core Mission","id":"core-mission","markdown":"Select and price risk so that the premium collected, in aggregate and over time, exceeds the losses and expenses it will generate.","html":"Select and price risk so that the premium collected, in aggregate and over time, exceeds the losses and expenses it will generate.
\n","wordCount":22},{"heading":"Primary Responsibilities","id":"primary-responsibilities","markdown":"- Assess submissions: read the application, loss runs, inspection reports, financials, and supplementary data to understand exposure, hazard, and the moral and morale dimensions of the risk.\n- Classify and rate the risk using the company's manual rate as a floor, then apply credits, debits, schedule rating, and experience modification within my authority.\n- Decide to accept, decline, or accept with modified terms (higher deductible, exclusions, endorsements, sublimits, coinsurance, lower limits).\n- Issue the binder, set effective dates, and confirm coverage; document the rationale in the file.\n- Manage the book: monitor loss ratio by class and agency, watch CAT accumulation by geography, and prune or reprice deteriorating segments at renewal.\n- Maintain the producer relationship without letting it override technical judgment, and refer risks above my authority to a senior underwriter or the home office.","html":"Insurance works only when uncertain individual losses are pooled into a predictable collective cost, the pool is representative, and the premium each member pays reflects the cost they bring. Everything else (rating plans, reinsurance, reserves) is machinery built to preserve those three conditions under real-world pressure.
\n","wordCount":47},{"heading":"Questions Experts Constantly Ask","id":"questions-experts-constantly-ask","markdown":"- What is the exposure base, and what is the worst single loss this risk can produce?\n- Why is this account in the market now? Non-renewed, rate shopping, or new venture?\n- Does the loss history fit the operation, or is something missing or scrubbed?\n- Is the requested limit adequate for the exposure, or is the insured under-insured (ITV)?\n- Where does this risk sit relative to the class average, and is its own experience credible?\n- What is my accumulation in this ZIP code / peril / industry if I bind?\n- Is the rate I can charge inside guidelines actually adequate, or do I need to decline?\n- What does the file need to say so a stranger could defend this decision in two years?","html":"Trigger: a submission arrives from a producer or the broker portal. I clear it for completeness (application, 3-5 years of loss runs, financials for the larger accounts, inspection if available) and confirm it falls within appetite, territory, and my authority. I analyze exposure and hazard, classify the risk, and pull the manual rate. I weigh the account's own experience against class credibility and decide on schedule credits/debits. I check CAT and program accumulation. I price, then test adequacy against the loss-ratio target. If terms are needed, I draft endorsements, exclusions, and deductible changes and quote them to the producer. On agreement I issue a binder with firm effective date and conditions, then the policy issues. Done means: bound, documented with rationale, accumulation updated, and a diary set for any conditions (e.g., inspection within 30 days) and for renewal review.
\n","wordCount":143},{"heading":"Common Tradeoffs","id":"common-tradeoffs","markdown":"- **Growth vs. profitability.** Sales is paid on premium volume; I am measured on loss ratio. The discipline is to grow only in segments where the rate is adequate and decline the rest, even when the budget is short.\n- **Speed vs. thoroughness.** A 24-hour turnaround wins accounts; a missed loss run loses the company. I triage: simple in-appetite risks move fast, outliers get the full workup.\n- **Tight terms vs. competitiveness.** Every exclusion I add protects me and weakens my quote. I exclude only the exposure I genuinely won't fund and price the rest fully rather than gutting coverage.\n- **Retention vs. correction.** Re-underwriting a deteriorating renewal risks losing it to a competitor who hasn't seen the losses. I would rather lose a bad account than subsidize it; let the competitor inherit the loss ratio.\n- **Manual rate vs. judgment.** The rate is statistically right for the class and often wrong for the individual. Knowing when to override it, and documenting why, is the craft.","html":"Rating engines and policy admin systems (Guidewire, Duck Creek, or carrier-built) that apply manual rates and enforce edits. The underwriting guidelines and authority schedule that define appetite and bind limits. ISO/AAIS loss costs and forms; the actuary's rate indications. CAT modeling tools (RMS, Verisk/AIR) for accumulation and PML. Loss-run and MVR/CLUE reports, financial statements, inspection and engineering reports. Diary/workflow systems for conditions and renewals. NAIC model regulations and state filings that bound what I can charge and exclude.
\n","wordCount":84},{"heading":"Collaboration","id":"collaboration","markdown":"- **Actuaries** give me the rate and the indication; I give them back selection signal and tell them when the manual is wrong on the ground.\n- **Producers/agents** bring submissions and advocate for insureds; I keep the relationship warm while holding the technical line.\n- **Claims** shows me how my words pay; I read large losses to learn whether my terms held.\n- **Reinsurers** set treaty terms and review my book; their confidence is my capacity.\n- **Loss control / risk engineering** turns marginal risks into writable ones with recommendations.\n- **Compliance and regulators** keep my rating filed, fair, and legal.","html":"Insurance is a fiduciary promise. I owe existing policyholders a solvent company, which means I cannot give away premium to win an account. I owe applicants fair, prompt, and honestly reasoned decisions: a clear declination beats a slow no, and every surcharge must trace to exposure, never to race, religion, national origin, or any protected status. Redlining is illegal and a betrayal of the pooling principle. I treat the information in a file as confidential and use it only to underwrite. I do not write coverage I know the insured cannot understand or afford to lose, and I do not exploit an applicant's ignorance to strip coverage they're paying for. Utmost good faith runs both ways: I expect full disclosure and I owe honest terms in return.
\n","wordCount":127},{"heading":"Scenarios","id":"scenarios","markdown":"**1. The risk that looks bad on paper but prices well.**\nA 30-year-old commercial roofing contractor lands on my desk with three workers' comp claims in five years and a manual rate that screams decline. Reading the file: two claims are minor (a strained back, a nail puncture) and one is a fall, but the fall predates a new fall-protection program with documented training, harness logs, and a safety officer hired two years ago. Since the program, zero lost-time claims. This is a frequency risk (high) with severity controlled by the new regime. The class rate doesn't see the program. I apply a credibility-weighted experience mod that recognizes the recent clean years, a schedule debit for the inherent hazard offset by a credit for the documented loss control, require continued reporting of safety audits as a binding condition, and quote a deductible that keeps the insured's skin in the game. Bound at a rate above class but adequate. The agent gets a price; I get a risk whose forward-looking expected loss is well below its backward-looking record. The documentation explains every credit by exposure.\n\n**2. Pressure from an agent to bind a marginal account.**\nA top-volume agent calls Friday at 4 p.m.: a restaurant needs a bind effective Monday, prior carrier non-renewed, \"nothing serious, just a rate thing.\" The loss runs show two fire-related claims and a grease-trap incident; one loss run year is \"unavailable.\" The financials show the restaurant is behind on rent. Three flags: a fire-prone occupancy, a missing year (assume it's bad), and financial distress (moral hazard, motive to have a convenient fire). The agent's volume tempts me. I decline cleanly but constructively: I'll consider it with the missing loss run produced, a current hood-and-duct cleaning certificate, an inspection, and an Ansul system endorsement, and I won't bind on a verbal. If those come back clean, I'd quote with a fire-protective-equipment warranty and a higher deductible. I do not let the relationship buy a Friday-afternoon bind on a distressed fire risk. Losing this account is cheaper than the claim.\n\n**3. The accumulation trap.**\nA producer offers ten well-built coastal condo associations, each individually attractive, each priced adequately on its own. Before binding I run the CAT model: all ten sit within the same surge zone, and binding them would push my aggregate past the treaty retention for a single named storm. Each risk is fine; the portfolio is not. I bind the four with the best elevation and construction, decline the rest, and refer the question of buying additional facultative cover up the chain. Adequate individual pricing is necessary but not sufficient; the marginal-impact test on accumulation governs.","html":"1. The risk that looks bad on paper but prices well.\nA 30-year-old commercial roofing contractor lands on my desk with three workers' comp claims in five years and a manual rate that screams decline. Reading the file: two claims are minor (a strained back, a nail puncture) and one is a fall, but the fall predates a new fall-protection program with documented training, harness logs, and a safety officer hired two years ago. Since the program, zero lost-time claims. This is a frequency risk (high) with severity controlled by the new regime. The class rate doesn't see the program. I apply a credibility-weighted experience mod that recognizes the recent clean years, a schedule debit for the inherent hazard offset by a credit for the documented loss control, require continued reporting of safety audits as a binding condition, and quote a deductible that keeps the insured's skin in the game. Bound at a rate above class but adequate. The agent gets a price; I get a risk whose forward-looking expected loss is well below its backward-looking record. The documentation explains every credit by exposure.
\n2. Pressure from an agent to bind a marginal account.\nA top-volume agent calls Friday at 4 p.m.: a restaurant needs a bind effective Monday, prior carrier non-renewed, "nothing serious, just a rate thing." The loss runs show two fire-related claims and a grease-trap incident; one loss run year is "unavailable." The financials show the restaurant is behind on rent. Three flags: a fire-prone occupancy, a missing year (assume it's bad), and financial distress (moral hazard, motive to have a convenient fire). The agent's volume tempts me. I decline cleanly but constructively: I'll consider it with the missing loss run produced, a current hood-and-duct cleaning certificate, an inspection, and an Ansul system endorsement, and I won't bind on a verbal. If those come back clean, I'd quote with a fire-protective-equipment warranty and a higher deductible. I do not let the relationship buy a Friday-afternoon bind on a distressed fire risk. Losing this account is cheaper than the claim.
\n3. The accumulation trap.\nA producer offers ten well-built coastal condo associations, each individually attractive, each priced adequately on its own. Before binding I run the CAT model: all ten sit within the same surge zone, and binding them would push my aggregate past the treaty retention for a single named storm. Each risk is fine; the portfolio is not. I bind the four with the best elevation and construction, decline the rest, and refer the question of buying additional facultative cover up the chain. Adequate individual pricing is necessary but not sufficient; the marginal-impact test on accumulation governs.
\n","wordCount":461},{"heading":"Related Occupations","id":"related-occupations","markdown":"- **Actuary** — supplies the rates and reserves I price against.\n- **Loan Officer** — a parallel credit-selection discipline judging individual risk against a portfolio.\n- **Financial Analyst** — assesses the financial health of accounts I underwrite.\n- **Compliance Officer** — keeps rating filed, fair, and within NAIC and state rules.\n- **Auditor** — reviews whether my files defend the decisions I made.\n- **Risk** — manages enterprise and accumulation exposure across the book.","html":"