Underwriting: Introduction

Sree Venkat / 2024-02-13


This post is intended to be an introduction to the term underwriting, its origin, significance, outcomes, components and the role of tech.


Underwriting is the process through which institutions evaluate the risk of an event to arrive at probability of failure and potential consequences. This helps better understand the cost of undertaking an event.

The term underwriting is believed to have been coined by the famed insurer Lloyd’s of London which, in its early days, would accept some of an event’s risk in exchange for a premium (for example, a sea voyage that features the possibility of a shipwreck and the subsequent loss of cargo and/or even the crewmembers). The individuals paying the premiums would literally write their names under the text describing the possession or event for which Lloyd’s was assuming some risk; hence, the term written under or underwriting. More on this here

They arrive at this by means of facts, policies and equations which are based on the characteristics of the source of the event(individual or institution).


Any institution which needs to assess and evaluate the risk of its business opportunities identify underwriting and underwriters as a foundational pillar. The longevity of the institution is directly proportional to the diversity of risk in its business portfolio.

Business opportunities in loans, securities and insurance are entirely driven by underwriting.


Given that underwriters evaluate and assess risk the outcome of their work ends up in what many industries refer to as risk profile in one or more ways. This profile helps them with arriving at:

  1. Eligibility of the applicant
  2. Controlled Exposure
  3. Pricing
  4. Terms
  5. Conditions



These are based on the characteristics of the applicant and they vary from credit history, medical history to company financials depending upon whether the sector is loans, insurance and securities respectively.

Some examples could be credit usage, repayment behaviour, age, gender, lifestyle, occupation, debt to income ratios, assets etc.


These are logical expressions built upon the facts that help identify eligibility and risk profiles.

Eligibility is primarily looked at from a perspective of identifying potential red flags or indicators that lead to denial.

Risk profiles are a representation that is used to classify segments of the business opportunities(applicants).


Based on sector there are varying calculations based on the risk profiles and facts derived from the characteristics of the business opportunities.

These data points are used to define pricing, terms and conditions based on which the institutions is willing to take up the risk of the success/ failure event presented by the business opportunity.

Role of tech

While underwriting has its pros with respect to risk assessment and mitigation, there are some cons from the sheer amount of time and resources necessary:

  • availability of underwriters with domain knowledge
  • human biases and error
  • repetitive nature of some aspects which can be attributed to toil
  • limited scope of experimentation

Here is how the industry today has been making use of automation:

  • Standardising fact generators (reduces error)
  • Streamlining policies to the respective sectors within a given institution (improves scope for experiments)
  • Use of ml models to predict probability of events based on a wide variety of past instances (reduces bias)
  • Rule engines which help simplify orchestration of policies(improves efficiency)
  • Audit logs which help maintain a trail of the outcomes at each and every step to review and analyse the system on a daily and long term basis (improves visibility and regulation)

Deep dives

Each sector of underwriting and the how different businesses have taken advantage of tech is a series of black holes worth diving deep into. Adding below a few sources which are good reads to start.

  1. Origin of the term underwriting
  2. Insurance: Underwriting Principles
  3. No Code Tech for Underwriting