The Umbrex Financial Services Industry Practice has prepared this guide to terminology, acronyms, shorthand, and insider language to help a newcomer to the fintech and insurtech sector get up to speed rapidly.
Card Payments
Four-Party Card Model
The four parties are the cardholder, issuer, merchant, and acquirer. The issuer provides the cardholder’s account, while the acquirer provides the merchant’s acceptance relationship. The card network routes messages and sets operating rules, but it usually neither lends the cardholder money nor sells the merchant’s goods.
This model explains why a single purchase can involve several independent decisions and fee layers. A decline usually comes from the issuer, merchant settlement comes through the acquirer, and network rules govern both. An on-us transaction occurs when one institution sits on both the issuing and acquiring sides, which can simplify routing and change the economics.
Gateway, Processor, and Acquirer
A payment gateway securely accepts transaction details from the merchant’s checkout. A processor formats, routes, and records payment messages. An acquirer holds the network relationship that allows the merchant to accept cards and receive settlement.
One company may perform all three functions, which is why sales presentations often blur them. In technical or commercial discussions, ask whether “processor” means an acquiring processor, an issuer processor, or merely the software layer connecting to another processor. Those are materially different positions in the payment chain.
Payment Facilitator (PayFac)
A payment facilitator, usually shortened to PayFac, boards submerchants under a master acquiring relationship rather than requiring every merchant to establish a conventional direct merchant account. The PayFac conducts merchant underwriting, monitors activity, manages funding, and accepts obligations imposed by its acquirer and the card networks.
A software platform that refers merchants to a processor is not automatically a PayFac. The distinction determines who controls onboarding, who appears in network records, who absorbs certain fraud or chargeback exposures, and how quickly submerchants can be activated.
Merchant of Record (MoR)
The merchant of record, or MoR, is the entity that legally sells to the end customer and typically appears on the customer’s statement. It ordinarily assumes responsibility for payment acceptance, refunds, chargebacks, indirect taxes, and customer-facing transaction terms.
A PayFac enables other merchants to accept payments; an MoR becomes the seller for payment and legal purposes. Platforms sometimes describe themselves loosely as the merchant of record when they merely process funds. That wording deserves scrutiny because the liability difference is substantial.
Authorization, Capture, Clearing, and Settlement
Authorization asks the issuer to approve and reserve funds. Capture confirms the amount the merchant wants to collect. Clearing exchanges final transaction records and calculates obligations. Settlement moves net funds among the relevant institutions.
An authorization is not payment. It can expire, be reversed, or differ from the captured amount. When practitioners say a transaction “went through,” the useful follow-up is whether it was merely authorized, captured successfully, or actually settled.
Interchange, Scheme Fees, and MDR
Interchange is generally paid through the acquiring side to the issuer. Scheme fees, also called network assessments, are charged by the card network. The merchant discount rate (MDR) is the merchant’s all-in acceptance price, incorporating those pass-through costs plus acquiring, processing, and other markups.
Interchange is not the processor’s entire revenue. It is usually a major cost component that varies by card type, transaction channel, merchant category, geography, and data quality. Confusing interchange with MDR can produce very imaginative payment margin estimates.
Interchange-Plus and IC++
Interchange-plus pricing passes through actual interchange and adds a stated provider markup. IC++ commonly separates interchange, card scheme fees, and the acquirer’s markup. The precise presentation varies by market and provider.
Blended pricing instead charges a simplified rate across transaction types. Blending is easier to understand but can conceal the underlying mix of card and network costs. IC++ offers more transparency, although it also produces invoices that look as if someone exported the card network’s filing cabinet.
Merchant Category Code (MCC)
A merchant category code, or MCC, is a four-digit card-network classification assigned according to the merchant’s primary business activity. It affects interchange qualification, cardholder rewards, network eligibility, fraud controls, regulatory treatment, and restrictions on certain transaction types.
An MCC is not simply a marketing label, and it is not equivalent to a general industrial classification such as NAICS. Misclassification can alter economics or trigger network scrutiny. In onboarding discussions, “What MCC are we using?” may be a question about both price and permissibility.
BIN, IIN, and BIN Sponsorship
A bank identification number (BIN), formally an issuer identification number (IIN), identifies the institution and card program associated with a payment credential. Practitioners still say BIN even as the standard expands from historical six-digit identifiers to eight-digit IINs.
BIN sponsorship allows a program to issue cards through a licensed network member, usually a bank. The sponsor owns the regulated and network-facing relationship; the program manager may control product design and customer experience. Access to a BIN does not make the fintech a bank or card-network member.
Card-Present, Card-Not-Present, and MOTO
Card-present (CP) transactions use an in-person acceptance method such as chip, contactless, or sometimes swipe. Card-not-present (CNP) transactions occur without the physical card at the acceptance device, most commonly online or in-app. MOTO means mail order or telephone order.
These classifications affect authentication, interchange, fraud exposure, and liability rules. Typing card details into a merchant terminal does not magically make an e-commerce purchase card-present. Networks classify the acceptance method, not the merchant’s optimism.
Network Token and Vault Token
A vault token is a provider-created surrogate that maps to a stored primary account number inside that provider’s secure environment. A network token is issued through card-network token services and may be restricted to a device, merchant, or transaction domain.
Network tokens support credential lifecycle updates when cards expire or are replaced, often improving authorization rates and reducing exposed card data. Vault tokens mainly reduce a merchant’s handling of raw credentials. Both are called tokenization, but they are not operationally interchangeable.
3DS2 and Strong Customer Authentication
EMV 3-D Secure 2 (3DS2) exchanges transaction and device information with the issuer to authenticate a cardholder. A low-risk transaction may follow a frictionless flow; a higher-risk transaction may trigger a challenge such as an app confirmation or one-time code.
In the European Economic Area and the United Kingdom, 3DS2 is a common mechanism for meeting Strong Customer Authentication (SCA) requirements under payment-services rules. Exemptions and liability shifts are conditional, not magical shields. A merchant can improve fraud outcomes while still damaging conversion if challenge strategy is poorly tuned.
Soft Decline, Hard Decline, and Retry Logic
A soft decline suggests the transaction may succeed after authentication, changed data, different timing, or another permitted attempt. A hard decline signals that retrying is unlikely to help, such as an invalid account or blocked credential. These labels are practitioner shorthand rather than perfectly standardized network categories.
Retry logic determines whether, when, and how a declined payment is attempted again. Intelligent retries can recover revenue; indiscriminate retries can generate fees, annoy issuers, violate network rules, and lower future approval rates.
The authorization rate is usually approved authorization attempts divided by submitted attempts. Always inspect the denominator. Some teams exclude technical failures, retries, or suspected fraud, producing several equally confident versions of “the” approval rate.
Chargeback, Representment, and Friendly Fraud
A chargeback reverses a card transaction through the network dispute process. The merchant may submit evidence through representment, after which the dispute can progress to pre-arbitration or arbitration depending on the network and case.
Friendly fraud describes a cardholder disputing a transaction that was actually authorized or received, sometimes through confusion and sometimes through opportunism. It remains a dispute operationally even when the merchant believes the customer is wrong.
Chargeback ratios may be calculated by count or value and can use different transaction periods. Network monitoring programs have precise definitions, so a reassuring internal ratio may not be the ratio that determines whether the program enters monitoring.
Money Movement and Open Banking
ACH Credit, ACH Debit, and Return Codes
An ACH credit pushes funds from the originator’s account. An ACH debit pulls funds based on an authorization granted by the account holder. Both move through the US Automated Clearing House network rather than card rails.
ACH transactions can be returned using standardized NACHA return codes for reasons such as insufficient funds, invalid account information, or unauthorized debit. Return windows and consequences differ by reason and account type. Teams therefore track unauthorized returns separately from administrative or insufficient-funds returns.
RTP, FedNow, and Instant Payments
RTP is The Clearing House’s US real-time payment network, while FedNow is the Federal Reserve’s instant-payment service. Other markets have their own rails, such as SEPA Instant Credit Transfer in Europe and Faster Payments in the United Kingdom.
Instant payments generally provide rapid confirmation and final settlement. Same Day ACH is faster batch ACH, not an instant rail. The distinction matters for liquidity, fraud intervention, customer messaging, and whether a mistaken payment can realistically be recalled.
Push-Payment Finality and APP Fraud
Many instant rails use credit-push payments that become final quickly. This reduces settlement uncertainty but leaves little time to stop a payment after the sender approves it.
Authorized push payment (APP) fraud occurs when a victim is deceived into authorizing a transfer to a fraudster. The transaction is authorized in the authentication sense, but fraudulent in the broader sense. Reimbursement obligations vary by jurisdiction and are a major policy issue in markets with widespread instant payments.
Open Banking and Open Finance
Open banking gives authorized third parties access to bank-account information or payment initiation through customer-permissioned interfaces. Open finance extends the concept to products such as investments, pensions, mortgages, and insurance.
Practitioners may use “open banking” loosely for any bank-data aggregation. Technically, regulated API access, bilateral data-sharing arrangements, and credential-based scraping are different models with different consent, liability, and reliability characteristics.
AISP and PISP
Under European and UK payment-services frameworks, an Account Information Service Provider (AISP) accesses account data with the user’s consent. A Payment Initiation Service Provider (PISP) initiates account-to-account payments for the user.
A provider may hold one permission, both, or work through another regulated entity. AISP access supports aggregation and financial insights; PISP status concerns money movement. Saying a company “has open-banking access” does not establish which activity it is authorized to perform.
Screen Scraping and API Connectivity
Screen scraping uses customer credentials or delegated access to retrieve data from an institution’s online interface. API connectivity uses a structured interface designed for third-party data exchange, usually with tokenized consent.
APIs are generally more secure and predictable, but coverage and data completeness vary. Scraping may reach institutions or fields that APIs do not, yet it is more vulnerable to interface changes and credential restrictions. Aggregators often operate a mixed estate despite presenting a beautifully unified API to their customers.
Variable Recurring Payments (VRP)
A variable recurring payment allows a customer to authorize repeated account-to-account payments within agreed parameters, rather than approving every transaction individually. The concept is especially associated with UK open banking.
Sweeping VRP moves money between accounts belonging to the same customer, while broader commercial VRP can support recurring merchant payments. Practitioners watch VRP because it could compete with card-on-file and direct-debit models, but regulatory and bank coverage remain important constraints.
Confirmation of Payee (CoP)
Confirmation of Payee checks whether the recipient name entered by a payer matches the name associated with the destination account. Similar services may be called Verification of Payee in other markets.
CoP can reduce misdirected payments and some impersonation fraud, but it does not prove that the recipient is honest or that the underlying transaction is legitimate. A successful match means the name aligns with the account, not that the investment opportunity is suddenly real.
Banking Infrastructure and Embedded Finance
Embedded Finance and Banking-as-a-Service
Embedded finance places a financial product inside a nonfinancial customer journey, such as payments in a marketplace or a deposit account inside business software. Banking-as-a-Service (BaaS) supplies regulated banking capabilities and infrastructure to support such products.
Embedded finance describes the distribution experience; BaaS describes an enabling model. A fintech can embed a product without operating a broad BaaS platform, and a BaaS provider may serve multiple programs that customers never recognize as using common infrastructure.
Sponsor Bank
A sponsor bank provides the charter, regulatory permissions, network access, and account structure behind a fintech program. It may sponsor card issuance, deposit accounts, payments, or lending.
The fintech may own the user experience, but the sponsor bank remains accountable for activities conducted through the bank. When a sponsor requests transaction monitoring, complaint data, model documentation, or marketing approval, it is not behaving like an ordinary software supplier. It is managing regulated exposure attached to its charter.
Program Manager
A program manager coordinates the operating layers of a fintech program, potentially including the sponsor bank, processor, card network, compliance controls, ledger, and customer-facing product. The exact allocation varies widely.
The title does not by itself reveal who performs KYC, holds funds, approves cardholders, owns disputes, or bears fraud losses. Program diagrams should assign each regulated and operational responsibility explicitly rather than placing “program manager” in one reassuringly large box.
FBO Account, Omnibus Account, and Segregated Account
An FBO account is titled “for benefit of” underlying customers or users. An omnibus account pools funds for multiple beneficial owners, with ownership tracked in a subledger. A segregated account separates funds by customer, program, or legal purpose.
These labels overlap but are not synonyms. An FBO account can be omnibus, and segregation can exist at several levels. The practical questions are whose name appears on the bank’s records, whose money is legally held, and which system proves each customer’s balance.
Core Ledger and Subledger
The bank’s core ledger records the bank’s legal account balances. A fintech subledger allocates an omnibus balance across individual customers and transaction states. A sound subledger normally uses double-entry accounting so every movement has equal debits and credits.
The subledger may show pending, available, and posted balances that do not map one-for-one to the bank’s end-of-day balance. When practitioners say “the ledger is right,” the next question is which ledger and as of what cut-off time.
Virtual Account
A virtual account is a unique account identifier mapped to a physical settlement or concentration account. It helps attribute incoming and outgoing payments to a particular customer, invoice, wallet, or business unit without opening a separate bank account for each one.
A virtual account may look like an ordinary bank account to the payer, but it may not constitute a separately held deposit account. That distinction affects legal ownership, statements, deposit insurance analysis, and reconciliation design.
Pass-Through Deposit Insurance
Pass-through deposit insurance can allow each beneficial owner’s funds in a custodial or FBO arrangement to be insured up to applicable limits, provided the legal and recordkeeping requirements are satisfied. The insurance passes through the account holder of record to the underlying owners.
Marketing a product as “FDIC insured” does not create coverage. Account titling, ownership records, eligible deposit status, and aggregation with the customer’s other deposits at the same bank all matter. In program reviews, ledger accuracy becomes a deposit-insurance issue, not merely an accounting preference.
Deposit Sweep Network
A deposit sweep network allocates customer funds across multiple banks, often to increase aggregate deposit-insurance capacity or manage funding concentrations. The platform tracks where each customer’s money is placed and may rebalance allocations over time.
The headline insurance amount depends on customer eligibility, participating institutions, existing deposits, and accurate allocation records. A sweep program also introduces settlement, disclosure, liquidity, and reconciliation complexity behind an interface that may display one deceptively simple balance.
Three-Way Reconciliation
In fintech programs, three-way reconciliation commonly compares the fintech subledger, the sponsor bank or settlement account, and a processor or payment-network record. The objective is to identify timing differences, missing transactions, duplicate postings, and actual breaks.
A balanced general ledger does not prove customer funds are correct. Each external cash position and customer liability must also reconcile. Persistent “timing items” have a habit of aging into things that are no longer timing items.
Digital Lending and Credit Risk
Loan Origination System and Loan Management System
A loan origination system (LOS) handles application intake, verification, underwriting, decisioning, documentation, and funding. A loan management system (LMS), often called a servicing platform, manages balances, payments, delinquency, statements, modifications, and payoff after origination.
Some platforms span both, but the workflows and control requirements differ. If a proposed change affects disclosures before funding, it is usually an origination issue. If it affects allocation of a late payment, it is a servicing issue.
Credit Box and Decisioning
The credit box defines which borrowers and exposures a lender is willing to approve, including score ranges, income tests, debt levels, loan sizes, geographies, and exclusions. Decisioning applies models and rules to determine approval, terms, referral, or decline.
A model estimates risk; the credit box translates risk appetite and policy into an approve-or-decline boundary. A lender can use the same model while changing its credit box, which is why approval rates can move even when “the model has not changed.”
Thin File, No File, and Cash-Flow Underwriting
A thin-file applicant has limited traditional credit history. A no-file applicant lacks enough bureau information to generate a conventional score. Cash-flow underwriting uses transaction-level income, balance, and expense patterns to assess ability and willingness to repay.
Cash-flow data can expand access and improve timeliness, but it does not remove the need for validation, fair-lending testing, explainability, and data-permission controls. “Alternative” data can still reproduce traditional disparities in less obvious forms.
Soft Pull, Hard Pull, Prequalification, and Preapproval
A soft pull generally does not affect the consumer’s credit score and is often used for eligibility screening. A hard pull is associated with an application for credit and may affect the score.
Prequalification is typically a conditional indication based on limited information. Preapproval often implies a stronger review, but the labels are not perfectly standardized. Marketing language must match the actual underwriting performed and the conditions still outstanding.
Adverse Action Notice and Reason Codes
An adverse action notice tells an applicant about a denial or other unfavorable credit decision and provides required reasons and disclosures. In the United States, the Equal Credit Opportunity Act and Regulation B are central, with additional Fair Credit Reporting Act requirements when consumer-report information is used.
Reason codes must reflect the principal factors that actually drove the decision. A generic “failed policy” explanation may be operationally convenient but legally inadequate. This is why model explainability becomes a customer-communication requirement rather than merely a data-science preference.
Bank Partnership and True Lender
In a bank-partnership model, a bank originates loans while a fintech provides technology, marketing, servicing, analytics, or funding support. The structure may use federal or state-bank lending authority and then sell receivables or participation interests.
True lender analysis asks whether the bank is substantively the lender or whether the nonbank partner should be treated as the lender. Courts and regulators may examine economics, control, risk retention, and program design. Merely placing the bank’s name on the agreement does not settle the issue.
BNPL and Pay-in-4
Buy now, pay later (BNPL) covers point-of-sale credit structures that defer or split payment. Pay-in-4 typically divides a purchase into four installments, often with the first paid at checkout and no stated consumer interest.
No-interest does not mean no economics. Providers may earn merchant fees, late fees, interchange, or cross-sell revenue. Regulatory treatment depends on product structure and jurisdiction, particularly around disclosures, disputes, credit reporting, and ability-to-repay expectations.
Warehouse Facility and Borrowing Base
A warehouse facility finances loans before they are sold, securitized, or held longer term. The lender can borrow against eligible receivables subject to an advance rate and a calculated borrowing base.
Eligibility rules, concentration limits, delinquency triggers, haircuts, and reserves determine how much funding is actually available. A portfolio may look large in gross terms while supporting much less borrowing once ineligible loans and required overcollateralization are removed.
Forward Flow Agreement
A forward flow agreement commits an investor to purchase qualifying loans originated over a future period. The agreement specifies eligibility criteria, pricing, volume, representations, warranties, and remedies for defective assets.
For a fintech lender, forward flow can provide repeatable funding without retaining every loan. The commercial tension lies in who controls the credit box and who bears deterioration when actual originations do not resemble the expected pool.
Vintage Analysis and Seasoning
A vintage groups loans by origination period, often month or quarter, and tracks their performance at comparable ages. Seasoning describes the accumulation of performance history as the loans mature.
Comparing total portfolio losses can hide deterioration because newer loans have had less time to default. Vintage curves answer whether recent cohorts are behaving worse at month six than older cohorts did at month six. That is usually more informative than admiring a blended average.
DPD, Roll Rate, Cure Rate, and First-Payment Default
Days past due (DPD) measures delinquency age. A roll rate measures movement from one delinquency bucket to a worse bucket, such as 30 DPD to 60 DPD. A cure rate measures return to current status.
First-payment default (FPD) occurs when the borrower misses the first scheduled payment under the lender’s defined threshold. Elevated FPD can indicate fraud, weak verification, poor acquisition quality, or a broken payment setup. Definitions vary, so always ask whether partial payments or grace periods count.
PD, LGD, EAD, and Charge-Off
Probability of default (PD) estimates the likelihood of default. Loss given default (LGD) estimates the percentage lost after recoveries. Exposure at default (EAD) estimates the balance exposed when default occurs. A simplified expected-loss expression is PD × LGD × EAD.
A charge-off is an accounting recognition that a receivable is unlikely to be collected under the applicable policy. It does not necessarily end collection activity. Net charge-offs equal gross charge-offs less recoveries, usually stated as a rate against average receivables.
APR and Finance Charge
The annual percentage rate (APR) expresses the annualized cost of credit using prescribed assumptions and included charges. The finance charge is the dollar amount of charges imposed as an incident to credit under applicable disclosure rules.
APR is not always the contractual interest rate, and fee inclusion can materially change it. Comparing APRs across products is useful only when the underlying timing and assumptions are understood, particularly for short-duration credit where annualization produces attention-grabbing numbers.
Insurance Distribution and Market Structure
Risk-Bearing Carrier, Paper, and Capacity
The risk-bearing carrier issues the insurance policy and assumes the contractual obligation to pay covered claims. Practitioners call the carrier’s licensed policy-issuing position its paper. Capacity is the amount and type of risk that the carrier or its reinsurers are willing to support.
An insurtech may control product design and distribution while another entity supplies paper and capacity. Hearing “we have paper” means a carrier relationship exists; it does not necessarily mean long-term capacity, approvals, reinsurance, or launch readiness are secured.
Managing General Agent and Managing General Underwriter
A managing general agent (MGA) is an insurance intermediary with delegated authority that may include underwriting, pricing, binding, and claims functions. A managing general underwriter (MGU) is a similar term used frequently in reinsurance or specialist underwriting contexts.
Usage varies by jurisdiction and market. What matters is the authority delegated, not the letters on the company website. Some MGAs merely distribute within tight rules; others effectively operate the product while the carrier retains formal risk-bearing status.
Fronting Carrier
A fronting carrier issues policies and provides licenses, regulatory filings, and claims-paying obligations while transferring much of the economic risk to reinsurers or other capital providers.
Fronting is not risk-free for the carrier. It retains credit, operational, regulatory, and often some underwriting exposure. Fronting discussions therefore focus heavily on collateral, claims control, delegated authority, reinsurance security, and the quality of program data.
Third-Party Administrator (TPA)
A third-party administrator performs delegated administrative functions such as policy servicing, billing, claims intake, adjudication, or benefit administration. TPAs may require specific licenses depending on the product and jurisdiction.
A TPA does not ordinarily become the insurer simply by making claims decisions. The carrier remains responsible for policy obligations, although poorly controlled TPA activity can create regulatory, customer, and reserving consequences for the carrier.
Producer, Agent, and Broker
Producer is a licensing term covering persons or entities that sell, solicit, or negotiate insurance. An agent generally acts on behalf of an insurer, while a broker is commonly described as acting for the customer. Legal definitions and duties vary by jurisdiction.
Digital distribution does not eliminate producer licensing. If a platform recommends, solicits, or materially participates in selling coverage, the exact journey and compensation model may determine whether licensing is required.
Delegated Authority and Binding Authority
Delegated authority allows an intermediary to perform functions that would otherwise sit with the carrier, such as underwriting, binding, issuing documents, or handling claims. Binding authority specifically allows coverage to be committed within agreed parameters.
The authority is constrained by underwriting guidelines, limits, territories, referral rules, and reporting obligations. In reviews, an “out-of-authority bind” means the intermediary committed the carrier outside those boundaries, which is considerably more serious than a workflow exception.
Binder
An insurance binder is temporary evidence that coverage is in effect before the final policy documents are issued. It identifies essential terms such as the insured risk, effective date, coverage, limits, and insurer.
A binder is not the same as binding authority. The binder evidences coverage for a particular insured; binding authority is the contractual power allowing an intermediary to create that coverage.
Bordereau
A bordereau is a structured report sent by an MGA, coverholder, cedent, or administrator to a carrier or reinsurer. Premium bordereaux report policies and transactions; claims bordereaux report losses, reserves, and payments. The plural is bordereaux.
Timely, accurate bordereaux support accounting, reserving, exposure management, regulatory reporting, and reinsurance recovery. “The bordereau needs work” can mean the risk bearer lacks reliable visibility into the business it has legally agreed to support.
Admitted, Nonadmitted, and Excess and Surplus Lines
An admitted insurer is licensed in the relevant state and uses approved or permitted rates and forms under that state’s framework. A nonadmitted insurer is not licensed there but may write eligible business through the excess and surplus (E&S) lines market.
E&S supports unusual, high-risk, or rapidly evolving exposures that the admitted market may not accommodate. A surplus-lines broker typically handles required placement procedures, taxes, and filings. Nonadmitted does not mean unregulated; it means regulated through a different mechanism.
Embedded Insurance and Affinity Distribution
Embedded insurance places coverage within the purchase or use of another product, such as travel protection during booking. Affinity distribution offers insurance to members or customers of a group with a shared relationship, such as an association or platform.
The crucial questions are whether coverage is optional, who is licensed, how consent is obtained, and whether the product is appropriate for the underlying purchase. Teams often track attach rate, the percentage of eligible transactions or users that purchase coverage.
Insurance Underwriting and Product
Appetite, Eligibility, and Rating
Underwriting appetite describes the types and concentrations of risk an insurer wants. Eligibility rules determine whether a submission can be considered. Rating determines the premium and terms for an eligible risk.
These stages answer different questions: Do we want this class of business, can this specific risk enter the product, and what should we charge? A pricing model can produce a rate for a risk that the carrier does not want to write at any price.
Exposure Unit and Class Code
An exposure unit is the basis against which risk and premium are measured, such as payroll, sales, vehicle-years, insured value, or member-months. A class code groups risks with similar expected loss characteristics, particularly in commercial and workers’ compensation insurance.
Incorrect exposure bases or class codes distort both pricing and regulatory reporting. A premium that looks reasonable at quotation can become inadequate after audit if the actual payroll, revenue, or other exposure is higher than declared.
Rate Filing and Form Filing
A rate filing submits pricing rules, factors, and actuarial support to an insurance regulator. A form filing submits policy wording, endorsements, notices, and related documents. Depending on jurisdiction and product, filings may require prior approval, use-and-file treatment, or another review standard.
Product configuration cannot safely outrun approved filings. A user-interface change that alters eligibility, coverage wording, or displayed price may have regulatory consequences even if the technology team views it as a minor release.
Quote-Bind-Issue and Straight-Through Underwriting
Quote-bind-issue (QBI) describes the path from generating a premium indication, to committing coverage, to issuing the policy documents. Straight-through underwriting completes that path without manual underwriter intervention.
A referral sends the application to a human because it falls outside automated authority or requires judgment. A high straight-through rate is not automatically good if automation is approving business that should have been referred.
Telematics and Usage-Based Insurance
Telematics captures behavioral and contextual data such as mileage, speed, braking, time of day, or phone handling. Usage-based insurance (UBI) uses such data in pricing, discounts, underwriting, or customer feedback.
Models must distinguish exposure, such as miles driven, from behavior, such as harsh braking. Consent, sensor quality, proxy discrimination, scoring stability, and customer understanding all affect whether a technically predictive model becomes a viable insurance product.
Parametric Insurance
Parametric insurance pays when a predefined, objectively measured trigger occurs, such as wind speed, rainfall, earthquake intensity, or flight delay. Payment depends on the trigger, not on a conventional adjustment of the policyholder’s actual loss.
The central issue is basis risk: the trigger may occur without matching the insured’s loss, or the insured may suffer a loss without the trigger being met. Parametric products can pay quickly, but only if the trigger is clear, reliable, and legally structured as insurance where required.
Accelerated Underwriting and Evidence of Insurability
In life insurance, accelerated underwriting uses digital applications, third-party data, and predictive models to issue some policies without traditional medical exams. Evidence of insurability (EOI) is the health and risk information required before certain coverage can be approved.
Accelerated underwriting does not mean no underwriting. It means replacing or selectively waiving parts of the conventional evidence process. Cases outside model limits may still require medical records, laboratory tests, or full underwriting.
Anti-Selection
Anti-selection, also called adverse selection, occurs when people with greater expected risk are more likely to buy, retain, or increase coverage than the insurer’s pricing assumptions anticipate.
Digital convenience can intensify anti-selection if customers can recognize favorable mispricing faster than the insurer can. Waiting periods, eligibility rules, benefit limits, and data verification often exist partly to control this effect, not merely to make the application longer.
Actuarial Indication and Credibility
An actuarial indication estimates the rate change needed for expected premiums to cover projected losses and expenses while meeting the target return or contingency provision. It is an analytical result, not automatically the rate that will be filed or charged.
Credibility describes how much statistical weight can reasonably be placed on a body of experience. Sparse insurtech data may be blended with industry benchmarks or broader classes. A highly precise model output can still rest on a very small and uncooperative sample.
Catastrophe Model, AAL, and PML
A catastrophe model combines hazard, exposure, vulnerability, and financial terms to simulate losses from events such as hurricanes, earthquakes, floods, or wildfires. Average annual loss (AAL) is the long-run modeled average loss per year.
Probable maximum loss (PML) usually refers to loss at a selected return period or probability level, although conventions vary. A “one-in-100” result is not the worst conceivable loss and does not mean the event occurs neatly once every century.
Policy Administration and Claims
Policy Administration System (PAS)
A policy administration system manages policy records, quotations, issuance, endorsements, billing instructions, renewals, cancellations, and document generation. It is the operational system of record for policy terms and status.
Insurtech interfaces often sit above a carrier or vendor PAS. If the front end and PAS disagree, claims and regulatory teams generally care about the policy record that legally controls, not the screen that looked nicer at checkout.
Endorsement, Rider, and Declarations Page
An endorsement modifies policy terms, coverage, or limits. A rider performs a similar function and is especially common in life and health products. The declarations page summarizes policy-specific details such as insureds, limits, deductibles, locations, and effective dates.
The declarations page is not the entire contract. Coverage depends on the forms, exclusions, definitions, and endorsements attached to it. Newcomers often treat the summary as if it contains all operative wording.
First Notice of Loss (FNOL)
First notice of loss (FNOL) is the initial report that an insured event or claim has occurred. It captures information needed to identify the policy, establish the claim, assess urgency, and begin triage.
A digital FNOL can improve speed and data quality, but it is only the intake stage. It does not confirm coverage, liability, or claim value. “We automated FNOL” may still leave nearly every consequential decision with an adjuster.
Coverage Determination and Reservation of Rights
A coverage determination assesses whether the policy responds to the reported loss. A reservation of rights notifies the insured that the carrier is investigating or defending while preserving its ability to deny some or all coverage.
This is distinct from deciding who caused the loss or how much it is worth. Claims automation that conflates coverage, liability, and valuation can produce fast answers to the wrong question.
Case Reserve and IBNR
A case reserve is the estimated unpaid amount assigned to a known claim. Incurred but not reported (IBNR) reserves cover claims that have occurred but are not yet reported, plus development not fully reflected in case reserves under some reserving conventions.
Case reserves are claim-level estimates; IBNR is generally established at an aggregate level using actuarial methods. Low reported claims can look favorable while IBNR rises because actuaries expect more losses to emerge later.
Indemnity, ALAE, and ULAE
Indemnity is the amount paid for the insured loss itself. Allocated loss adjustment expense (ALAE) can be assigned to specific claims, such as defense counsel or an independent adjuster. Unallocated loss adjustment expense (ULAE) supports claims handling more broadly.
Loss-ratio definitions differ on whether and how adjustment expenses are included. Two teams can report different loss ratios from the same claims portfolio and both be internally consistent, which is why the numerator definition matters.
Frequency and Severity
Claim frequency measures how often claims occur relative to an exposure base. Severity measures the average cost per claim. Expected loss is broadly driven by both.
A portfolio can deteriorate because more claims occur, because each claim costs more, or because the business mix changes. Claims reviews that say only “losses are up” have not yet completed the diagnostic portion of the meeting.
Subrogation and Salvage
Subrogation allows an insurer that paid a claim to pursue recovery from a responsible third party. Salvage is value recovered from damaged property after a claim, such as the sale of a totaled vehicle.
Both reduce net claim cost, but through different mechanisms. Recovery potential should be identified early because evidence, notice, and asset value can disappear while the claim remains in an ordinary workflow queue.
Special Investigation Unit (SIU)
A Special Investigation Unit investigates claims or applications with indicators of fraud, organized abuse, material misrepresentation, or other suspicious behavior. Referral criteria may combine rules, models, network analysis, and adjuster judgment.
A referral is not a fraud finding. Excessive referrals create delay and expense; too few can signal weak detection or pressure to preserve straight-through processing. The useful measure is not simply SIU volume but the quality and disposition of referrals.
Claims Leakage and Straight-Through Claims
Claims leakage is the avoidable difference between what a claim actually costs and what it should have cost under proper coverage, investigation, negotiation, recovery, and handling. It may arise from overpayment, missed subrogation, poor vendor control, or process error.
Straight-through claims processing resolves eligible claims without manual handling. Automation can reduce handling expense and cycle time, but weak controls can convert many small errors into very consistent leakage.
Insurance Economics and Reinsurance
Written, Earned, and Unearned Premium
Written premium is recorded when coverage is written, subject to the applicable accounting basis. Earned premium is the portion recognized as the insurer provides coverage over time. Unearned premium represents the remaining coverage obligation.
A rapidly growing program can report strong written premium while much of it remains unearned. Claims ratios are generally compared with earned premium because both should relate to the same exposure period.
GWP, GEP, NWP, and NEP
Gross written premium (GWP) is premium written before reinsurance. Gross earned premium (GEP) is the earned portion before reinsurance. Net written premium (NWP) and net earned premium (NEP) are after the effect of ceded reinsurance.
Growth claims based on GWP may say little about retained economics. A program can generate substantial gross volume while ceding most premium and risk to reinsurers.
Loss Ratio
The loss ratio compares losses with earned premium. A common form is incurred losses ÷ earned premium, but definitions differ on whether loss adjustment expenses are included.
Low loss ratios can reflect sound selection, conservative pricing, immature claims, favorable reserve assumptions, or simply a benign period. In young portfolios, development patterns matter more than a single early ratio.
Expense Ratio
The expense ratio compares underwriting expenses with premium. Depending on the reporting framework, the denominator may be written premium, earned premium, or another basis.
Distribution commissions, acquisition expenses, carrier fees, administration, and operating costs may be treated differently across organizations. Never combine expense ratios from different sources without checking the accounting basis.
Combined Ratio
The combined ratio is broadly the loss ratio plus the expense ratio. A result below 100 percent indicates an underwriting profit under the stated basis; above 100 percent indicates an underwriting loss.
Combined ratio = loss ratio + expense ratio
It does not include all investment income, financing costs, taxes, or capital effects. A program can have a sub-100 combined ratio and still disappoint economically, or exceed 100 while remaining attractive for other reasons.
Ceded and Retained Risk
Ceded risk is transferred to a reinsurer. Retained risk remains with the insurer after reinsurance. The same distinction applies to premium, losses, and limits.
“We cede 80 percent” is incomplete without identifying which layer, expenses, commissions, exclusions, and event definitions apply. Reinsurance often reallocates risk in a more structured way than a simple percentage suggests.
Quota Share
A quota-share treaty transfers a fixed percentage of defined premiums and losses to a reinsurer. If 60 percent is ceded, the reinsurer generally receives 60 percent of covered premium and pays 60 percent of covered losses, subject to treaty terms.
Quota share supports capacity and capital management while aligning the parties across the portfolio. Economics are heavily influenced by ceding commissions, profit commissions, exclusions, and whether the underlying business performs as expected.
Excess of Loss and Attachment Point
Excess-of-loss reinsurance pays losses above an attachment point up to a stated limit. It may protect against individual large claims or the accumulation of losses from one event.
A layer described as 5 million excess of 5 million generally covers the next 5 million after the cedent absorbs the first 5 million, subject to the contract. Reinstatements, aggregates, exclusions, and event definitions determine how much protection is actually available.
Treaty and Facultative Reinsurance
Treaty reinsurance automatically covers a defined portfolio of risks meeting agreed terms. Facultative reinsurance is placed for an individual risk or specific exposure.
Treaty coverage offers repeatable capacity; facultative placement supports risks that are unusually large or outside normal treaty terms. If a submission “needs fac,” it generally cannot be bound under ordinary portfolio authority without separate reinsurance approval.
Ceding Commission and Profit Commission
A ceding commission is paid by the reinsurer to the cedent to compensate for acquisition and administration expenses, and sometimes to reflect expected profitability. A profit commission provides additional compensation when results meet defined performance thresholds.
These mechanisms can materially change MGA, carrier, and reinsurer economics even when the headline quota-share percentage stays constant. Definitions of profit, loss development, carryforward, and calculation periods deserve close attention.
Fronting Fee and Reinsurance Collateral
A fronting fee compensates the issuing carrier for licenses, regulatory obligations, administration, capital usage, and retained exposure. Reinsurance collateral protects the carrier against the reinsurer’s failure to pay and may take the form of trusts, letters of credit, or withheld funds.
The fee is not pure margin if the front has meaningful operating or credit obligations. Collateral requirements can become a decisive constraint for an insurtech or capital provider that planned around premium volume rather than trapped capital.
Identity, Financial Crime, and Fraud
CIP, KYC, CDD, and EDD
A Customer Identification Program (CIP) defines how a regulated institution obtains and verifies identifying information. Know Your Customer (KYC) is the broader practitioner term for identity and customer-risk controls. Customer Due Diligence (CDD) assesses the customer’s identity, activity, purpose, and risk.
Enhanced Due Diligence (EDD) applies deeper review to higher-risk customers or situations. Passing identity verification does not mean CDD is complete; a real person can still present unacceptable money-laundering, sanctions, or fraud risk.
KYB and Ultimate Beneficial Owner
Know Your Business (KYB) verifies a legal entity, its registration, status, activities, ownership, and authorized representatives. An ultimate beneficial owner (UBO) is a natural person who ultimately owns or controls the entity under the relevant threshold and rule.
KYB is not merely KYC with a company name. Corporate structures, nominees, subsidiaries, and control rights can make ownership difficult to establish. Fintech onboarding often stalls after the business itself is verified but before its beneficial owners are satisfactorily resolved.
AML Transaction Monitoring
Anti-money laundering transaction monitoring uses rules, scenarios, models, and investigative workflows to identify activity that may indicate laundering or related financial crime. Examples include structuring, rapid movement of funds, unusual counterparties, or activity inconsistent with the customer’s profile.
A monitoring alert is not a finding of suspicious activity. It is an invitation to investigate, often one of several thousand invitations. Effectiveness depends on data completeness, calibration, case quality, and whether investigators can reconstruct the actual flow of funds.
Sanctions Screening and PEP Screening
Sanctions screening compares customers, counterparties, and transactions with restrictions issued by authorities such as OFAC, the United Nations, the European Union, or the United Kingdom. Politically exposed person (PEP) screening identifies people whose public position may create elevated corruption risk.
A PEP is not automatically prohibited, while a sanctions match may legally block or reject activity depending on the regime. Name similarity creates false positives, so screening requires matching logic, disposition procedures, and escalation for genuine matches.
SAR, STR, and CTR
A Suspicious Activity Report (SAR), or Suspicious Transaction Report (STR) in many jurisdictions, reports activity meeting the applicable suspicion standard. SAR filing and confidentiality rules are tightly controlled.
In the United States, a Currency Transaction Report (CTR) reports qualifying cash transactions above the regulatory threshold. A CTR is not itself an allegation of wrongdoing. Attempts to evade CTR reporting may create separate suspicion, commonly called structuring.
Identity Proofing and Authentication
Identity proofing establishes that a claimed identity corresponds to a real person and that the applicant is that person. Authentication verifies that a returning user is entitled to access an existing account or credential.
A customer can be correctly proofed at onboarding and later suffer account takeover. Conversely, a fraudster can authenticate perfectly to an account created under a synthetic identity. The controls address different stages of the identity lifecycle.
Documentary and Non-Documentary Verification
Documentary verification uses evidence such as passports, driver’s licenses, or corporate records. Non-documentary verification uses databases, credit files, phone records, account ownership, device signals, or other independent sources.
Many programs combine both. Document checks can detect an invented identity but may fail against sophisticated stolen documents; database checks can confirm that a person exists without proving the applicant is that person.
Liveness and Presentation Attack Detection
Liveness detection assesses whether biometric input comes from a live person rather than a photograph, replay, mask, or generated image. Presentation attack detection (PAD) is the broader discipline of detecting attempts to fool a biometric capture system.
Passive liveness analyzes the capture without requiring explicit gestures; active liveness may ask the user to move or respond. Deepfakes and injection attacks mean teams must also consider whether the camera feed itself has been manipulated.
Synthetic Identity Fraud
Synthetic identity fraud combines real and fabricated identity elements to create a persona that appears legitimate. The fraudster may cultivate credit history over time before drawing down funds or exploiting accounts.
This differs from conventional identity theft, where the fraudster impersonates an existing person. Synthetic identities can pass isolated verification checks because some constituent data is valid, making cross-source consistency and network analysis particularly important.
Account Takeover (ATO)
Account takeover, or ATO, occurs when an unauthorized actor gains control of an existing customer’s account. Methods include credential stuffing, phishing, SIM swapping, malware, session theft, and social engineering of support staff.
ATO detection focuses on changes from established behavior, device, location, payee, or session patterns. Strong onboarding controls do little to stop an attacker who compromises the customer six months later.
Mule Account
A mule account receives, moves, or converts proceeds for criminals. The account holder may be knowingly involved, recruited under false pretenses, or manipulated into facilitating transfers.
Mule detection looks beyond individual transactions to networks of counterparties, rapid pass-through behavior, shared devices, and unusual account-opening clusters. Closing one receiving account without examining connected accounts usually treats the leaf rather than the tree.
Velocity Rule and Device Fingerprint
A velocity rule flags excessive activity within a defined period, such as many cards used on one account or many accounts opened from one device. A device fingerprint combines technical attributes and behavioral signals to recognize or link devices.
Both are probabilistic. Shared households and corporate networks can resemble fraud rings, while sophisticated fraudsters can rotate devices and attributes. The value comes from combining signals rather than treating one fingerprint as a physical fingerprint.
Alert, Case, Disposition, and False Positive
An alert is generated by a monitoring rule or model. Related alerts may be grouped into a case for investigation. The investigator records a disposition, such as clearing the activity, escalating it, restricting the account, or considering a regulatory report.
A false positive is activity flagged as suspicious or fraudulent that is ultimately legitimate. High false-positive rates consume investigators and inconvenience customers, but an extremely low rate may indicate that controls are missing difficult cases rather than achieving enlightenment.
Regulatory Perimeter
Money Transmission, MSB, and MTL
Money transmission generally involves receiving money or monetary value from one party for transmission to another. In the United States, a qualifying provider may be a federal money services business (MSB) and may also need state money transmitter licenses (MTLs).
Federal MSB registration does not replace state licensing. Whether an activity constitutes money transmission depends on the exact flow of funds, contractual role, custody, exemptions, and jurisdiction. Product diagrams often become legal analysis once arrows begin carrying money.
Payment Institution and Electronic Money Institution
In European frameworks, a Payment Institution (PI) is authorized to provide specified payment services. An Electronic Money Institution (EMI) may issue electronic money in addition to providing permitted payment services.
Electronic money represents stored monetary value issued on receipt of funds and accepted beyond the issuer under the applicable definition. An EMI is not a deposit-taking bank, and customer funds are generally protected through safeguarding rather than deposit insurance.
Safeguarding
Safeguarding requires certain payment and electronic-money firms to protect customer funds, commonly through segregation in designated accounts or permitted insurance arrangements. It is intended to keep customer money separate from the firm’s own assets and available if the firm fails.
Safeguarding is not the same as deposit insurance. Reconciliation frequency, account designation, insolvency treatment, and the timing of fund receipt all matter. A bank balance that equals total customer liabilities by coincidence is not a safeguarding framework.
Agent-of-Payee Exemption
An agent-of-payee exemption may allow a platform receiving payment on behalf of a seller to avoid money-transmitter treatment when payment to the platform legally satisfies the buyer’s obligation to the seller. Availability and conditions vary by jurisdiction.
The exemption depends on genuine agency and contractual structure, not merely inserting the word “agent” into terms and conditions. Marketplaces use it carefully because a failed exemption analysis can expose the platform to licensing obligations across multiple states.
PCI DSS
The Payment Card Industry Data Security Standard (PCI DSS) sets security requirements for entities that store, process, or transmit cardholder data, and for systems that affect that environment. Compliance obligations depend on role, architecture, and transaction volume.
Tokenization and hosted payment fields can reduce scope, but outsourcing card handling does not eliminate all responsibility. PCI compliance is a card-industry requirement rather than a government license, although failures can produce contractual penalties, forensic reviews, and loss of card acceptance.
Regulation E Error Resolution
US Regulation E implements the Electronic Fund Transfer Act and establishes consumer protections for qualifying electronic fund transfers, including error notices, investigation deadlines, provisional credit, and liability rules.
Operational teams must distinguish an unauthorized transfer from a merchant dispute, service complaint, or authorized scam. The customer’s description may be imprecise, but classification determines deadlines and required treatment.
UDAAP
Unfair, Deceptive, or Abusive Acts or Practices (UDAAP) is a US consumer-financial standard applied to product design, disclosures, servicing, collections, complaints, and actual customer outcomes. The related term UDAP omits “abusive” and appears under other authorities.
Technically accurate disclosure may not cure a product whose overall operation misleads or harms consumers. In fintech reviews, UDAAP often appears when the legal text says one thing but defaults, interface design, or servicing behavior produce another practical result.
Model Risk Management and SR 11-7
Model risk management governs model inventory, development, validation, change control, monitoring, documentation, and use. In US banking, Federal Reserve and OCC guidance commonly associated with SR 11-7 strongly influences sponsor-bank expectations.
A vendor score or machine-learning service can still be a model for governance purposes. Calling it a “rules engine” does not eliminate model risk if it transforms inputs into consequential estimates or decisions.
Consumer Duty
The UK Consumer Duty requires firms to act to deliver good outcomes for retail customers across products and services, price and value, consumer understanding, and support. It emphasizes foreseeable harm and evidence of outcomes rather than disclosure alone.
For fintech and insurtech firms, this reaches interface design, friction, vulnerable customers, product testing, support access, and distribution chains. A smooth digital journey is not necessarily a good outcome if it smoothly leads customers into poor-value products.
Wealthtech and Brokerage
Registered Investment Adviser and Broker-Dealer
A US Registered Investment Adviser (RIA) provides investment advice under a fiduciary framework. A broker-dealer effects securities transactions and operates under broker-dealer conduct, supervision, capital, and market rules.
A platform may involve both entities, with advice supplied by an RIA and execution or custody provided by a broker-dealer. The distinction affects recommendations, disclosures, compensation, account agreements, and which entity the customer is actually dealing with at each step.
Robo-Adviser and Model Portfolio
A robo-adviser delivers automated or digitally assisted investment advice, often using questionnaires and algorithms to assign a customer to a model portfolio. The model specifies target allocations, instruments, and rebalancing rules.
Automation does not reduce the activity to software distribution. Suitability or fiduciary analysis, disclosure, model governance, trading, and exception handling remain regulated functions. A five-question risk survey may be elegant, but elegance is not the same as sufficient investor profiling.
Direct Indexing and Tax-Loss Harvesting
Direct indexing holds individual securities intended to approximate an index rather than holding a pooled index fund. This permits customization, exclusions, factor tilts, and account-level tax management.
Tax-loss harvesting realizes losses to offset gains while maintaining desired market exposure, subject to tax rules such as wash-sale restrictions. The benefit depends on tax circumstances, tracking error, transaction costs, and future gains, not merely the number of losses harvested.
Fractional Shares
Fractional shares allow customers to hold less than one whole share. The broker typically maintains whole shares at an omnibus level and allocates fractional economic interests through internal books and records.
Fractional interests may have different transferability, voting, execution, and corporate-action treatment from whole shares. A customer may own the economic exposure without holding a separately transferable fraction registered directly in the market infrastructure.
Payment for Order Flow, NBBO, and Best Execution
Payment for order flow (PFOF) is compensation paid to a broker for routing customer orders to a particular execution venue. The National Best Bid and Offer (NBBO) displays the best quoted US prices across protected venues.
Best execution requires reasonable diligence to obtain favorable execution under the circumstances and is broader than simply matching the NBBO. Price improvement, speed, likelihood of execution, order size, and conflicts all matter.
ACATS
The Automated Customer Account Transfer Service (ACATS) supports transfers of eligible brokerage assets between participating firms in the United States. It coordinates account validation, asset transfer, and residual settlement.
Not every asset transfers cleanly. Fractional shares, proprietary funds, alternatives, and unsupported positions may be liquidated, rejected, or handled separately. “We support ACATS” does not mean every customer account will arrive unchanged.
Omnibus and Fully Disclosed Clearing
Under omnibus clearing, an introducing firm holds customer-level records while the clearing firm sees a pooled account. Under a fully disclosed arrangement, customer accounts are carried individually on the clearing firm’s books.
The structure affects books and records, customer statements, margin, asset transfers, regulatory reporting, and who can see end-customer activity. It also determines where a fintech must build or obtain a reliable customer subledger.
T+1 Settlement
T+1 means a securities trade settles one business day after the trade date. Settlement transfers cash and securities after execution and clearing obligations are calculated.
A shorter cycle reduces counterparty exposure but compresses allocation, affirmation, funding, foreign-exchange, and exception-management timelines. The trade can execute instantly while the operational work still continues into the next day.
Digital Assets
Custodial and Noncustodial Wallet
A custodial wallet relies on a provider to control the cryptographic keys and execute transactions for the user. A noncustodial wallet gives the user direct control of the keys.
Custody determines who can authorize transfers, recover access, freeze assets, and respond to legal process. A wallet interface may look similar in both cases, but the control and insolvency implications are very different.
Hot Wallet and Cold Storage
A hot wallet has keys accessible to online systems for routine transfers. Cold storage keeps keys offline or otherwise isolated to reduce exposure to remote compromise.
Hot wallets support liquidity and speed; cold storage supports security. Institutional custody usually combines both with approval policies, withdrawal limits, key ceremonies, and controlled transfer processes.
Private Key and Seed Phrase
A private key authorizes transactions from a blockchain address. A seed phrase can deterministically generate a collection of private keys and therefore restore access to associated wallets.
Possession of the key generally means practical control of the asset. Losing it can make assets inaccessible, while exposing it can allow irreversible theft. A seed phrase should not be treated like an ordinary password that a support desk can simply reset.
On-Ramp and Off-Ramp
An on-ramp converts fiat currency into digital assets. An off-ramp converts digital assets back into fiat or enables withdrawal into conventional financial accounts.
These points connect blockchain activity to banks, cards, and payment systems, making them central to KYC, sanctions, fraud, liquidity, and licensing analysis. The blockchain transfer may be decentralized while the customer journey remains heavily dependent on regulated intermediaries.
Stablecoin and Reserve Assets
A stablecoin seeks to maintain a stable value relative to a reference asset, commonly a fiat currency. Fiat-backed stablecoins rely on reserve assets and redemption mechanisms; other designs may use crypto collateral or algorithms.
Reserve composition, custody, liquidity, redemption rights, legal claims, and disclosure quality determine how credible the peg is. “Backed one-to-one” is incomplete without asking backed by what, held where, for whose benefit, and redeemable under which conditions.
On-Chain, Off-Chain, Smart Contract, and Gas
On-chain activity is recorded on the blockchain. Off-chain activity occurs in internal ledgers or other systems without an immediate blockchain transaction. Centralized exchanges commonly execute many customer trades off-chain.
A smart contract is code deployed to a blockchain that executes defined logic. Gas is the computational fee required to process transactions on certain networks. Smart contracts automate execution, not legal interpretation, data truth, or good judgment.
Proof of Reserves and Proof of Liabilities
Proof of reserves seeks to demonstrate that a custodian or platform controls specified assets, often through cryptographic attestations and wallet verification. Proof of liabilities addresses what the platform owes customers.
Assets without complete liabilities do not establish solvency, and a point-in-time proof may not reveal encumbrances, borrowed assets, or subsequent transfers. The meaningful analysis reconciles controlled assets with customer obligations and legal ownership.
Real-World Asset Tokenization
Real-world asset (RWA) tokenization represents rights relating to conventional assets, such as funds, bonds, invoices, or property, using blockchain-based tokens. The token may represent ownership, a contractual claim, or access to an intermediary vehicle.
The legal wrapper matters more than the visual fact that a token exists. Transfer restrictions, securities laws, custody, servicing, redemption, and enforceability determine whether the holder owns the underlying asset, a claim against an issuer, or merely a technically impressive symbol.
The Phrase Translator
“We can lift auth with network tokens and smarter retries.”
It may mean: Card approvals could improve by replacing stale credentials and retrying recoverable declines more intelligently. Someone should still check network retry rules and whether the reported authorization-rate denominator is flattering.
“We’re a PayFac, not the merchant of record.”
It may mean: We facilitate payments for submerchants, but we do not want responsibility for being the legal seller, handling every tax obligation, or owning the underlying customer transaction.
“The sponsor wants daily FBO reconciliation before opening the BIN.”
It may mean: The bank will not permit card issuance until the fintech can prove that customer-level ledger balances reconcile to the pooled bank account and processor records every day.
“Same Day ACH will not solve the instant-payment use case.”
It may mean: Faster batch processing still does not provide the immediate confirmation, continuous availability, or settlement finality expected from RTP or FedNow.
“CoP helps, but it does not remove APP exposure.”
It may mean: Matching the recipient’s name can prevent some mistakes and impersonation scams, but customers can still be persuaded to send money to a correctly named fraudster.
“The flow is soft-pull prequal, then hard pull at acceptance.”
It may mean: Initial offers are screened without a score-impacting inquiry, but the lender performs a formal credit inquiry when the customer proceeds. Marketing and consent language need to reflect that handoff.
“The latest vintage is seasoning above the loss curve.”
It may mean: Recent loans are producing more delinquency or loss at the same age than the forecast or prior cohorts. The blended portfolio average may not have admitted this yet.
“The model decline needs adverse-action reasons, not just a score.”
It may mean: The lender must identify the principal factors actually driving the negative credit decision. A sophisticated probability estimate with no defensible explanation is operationally incomplete.
“The MGA has paper, but next year’s capacity is not locked.”
It may mean: A carrier currently issues the policies, but the carrier or reinsurers have not firmly committed enough future risk capacity. Growth forecasts should therefore remain emotionally modest.
“Quote-to-bind is fine; bind-to-issue is falling out in the PAS.”
It may mean: Customers accept quotes successfully, but bound transactions fail or require manual repair before policy issuance because the policy administration workflow or data mapping is breaking.
“The combined ratio works only after the ceding commission.”
It may mean: Underwriting economics depend materially on compensation from the reinsurer. Without that commission, acquisition and operating expenses would make the program unattractive.
“FNOL is digital, but coverage still goes to manual review.”
It may mean: Claim intake is automated, but the consequential question of whether the policy responds still requires a human. The front door is digital; the house behind it remains traditional.
“Case reserves are stable, but IBNR is developing adversely.”
It may mean: Known claims do not yet look worse individually, but actuarial analysis expects additional unreported claims or later development. The apparent calm in the claims system may be misleading.
“KYC passed; KYB and the UBO chain are the blockers.”
It may mean: The individual applicant was verified, but the business’s legal existence, ownership, or controlling persons remain unresolved. No one is opening the account merely because the founder’s selfie looked convincing.
“Proof of reserves is not proof of solvency.”
It may mean: The platform may demonstrate control of assets without proving complete customer liabilities, absence of encumbrances, or legal ownership. One side of the balance sheet is not the whole balance sheet.
Net Net
Fintech and insurtech language is difficult because regulated entities, software platforms, payment rails, credit models, insurance contracts, customer ledgers, and risk-transfer structures often occupy the same product journey. A familiar word such as “reserve,” “bind,” “settle,” or “authorize” can refer to a very specific legal, accounting, or operational state.
- Which rail, account structure, policy system, or transaction stage does this term refer to?
- Who is the regulated entity, risk-bearing carrier, lender, custodian, issuer, or merchant of record in this flow?
- Is the item authorized, captured, settled, bound, issued, earned, delinquent, charged off, or merely pending?
- Which classification controls here, such as MCC, class code, delinquency bucket, admitted status, or customer-risk tier?
- What is the exact numerator, denominator, exposure basis, and observation period for the metric being quoted?
- Does the conclusion depend on a model estimate, regulatory definition, policy wording, network rule, or contractual allocation?
- Which jurisdiction and license determine whether this payment, lending, insurance, or digital-asset activity is permitted?
- What evidence supports the position: a settlement file, subledger, bordereau, filing, model output, claims record, or customer authorization?
- Is the authority retained by the bank or carrier, or delegated to a program manager, MGA, TPA, processor, or platform?
- What event would shift liability, require referral, trigger a return, create a reserve, or change the economics?
Real fluency does not require memorizing every acronym. It requires recognizing which specialist system of meaning is in play, understanding what decision the language is carrying, and asking the question that makes the hidden assumption visible.