Charta is on Google Play: practise on the go. Get the app
Charta

Lay out the stages of the customer relationship cycle and the trade-off that arises between how creditworthy a client is and how profitable.

FRM® Part II · Credit Risk Measurement and Management · Credit Scoring and Retail Credit Risk Management · FRM2.CR.7.g

The idea

The customer relationship cycle runs from acquisition (marketing, application scoring and pricing), through account management (behavioural scoring, limit changes, cross-selling) and retention (attrition models and offers) to collections and recovery. Each stage uses different data and decisions. The trade-off is that creditworthiness and profitability are not the same thing. The safest customers often pay in full each month and generate mainly interchange income. Riskier revolvers carry balances and pay interest and fees, so some risky customers are worth more than their expected loss. In expected loss, PD is the probability of default, LGD the loss given default and EAD the exposure at default, as the next line shows. EL = PD × LGD × EAD. Account profit is what remains of income after costs and that expected loss, as follows. Profit = interest margin + fees − operating cost − EL. Customer lifetime value adds retention and discounting to this profit. Exam questions test that a risk score alone is not a profit score and that cutoffs can be set on profit.

Formulas

Annual profit per card account

profit = interest margin + fees − operating cost − PD × LGD × EAD

interest margin
the balance revolved × (card rate − funding rate)
fees
interchange and fee income
operating cost
the cost of running the account
PD
the probability of default
LGD
loss given default
EAD
the exposure at default

Expected loss is charged on EAD, not on the average balance. Safe customers who pay in full may earn little margin.

Customer lifetime value with level profit and constant retention

CLV = m ÷ (1 + d − retention)

m
the profit received at the end of each year while the customer stays
d
the discount rate
retention
the probability of staying for each further year

Assumes the first year's profit is certain and later ones are weighted by retention. Ignoring attrition turns it into a perpetuity, m ÷ d, which overstates value.

A practice question

From the question bank, with the reason each wrong option is wrong.

Which sequence lists the stages of the customer relationship cycle in a retail lender?

  1. Credit scoring, hedging, securitisation and financial reporting.Why not: Lists internal functions, not the stages of the customer relationship.
  2. Underwriting, pricing, syndication and settlement.Why not: Describes a wholesale deal process, not the customer cycle.
  3. Acquisition, account management, retention and collections. - correct
  4. Regulation, funding, trading and write-off.Why not: Mixes regulatory, funding and trading activities that are not customer relationship stages.

The customer relationship runs from winning the customer (acquisition) to managing the account, retaining the customer, and finally collecting on any delinquent accounts. Each stage has its own scores and decisions.

A harder one

A card issuer plans to send pre-approved limit increases to existing customers with strong payment records. Which stage of the customer relationship cycle is this, and which score mainly supports it?

  1. Acquisition, supported by an application score.Why not: Acquisition covers winning new customers; these customers are already on the books.
  2. Account management, using a behavioural score. - correct
  3. Retention, supported by a collection score.Why not: A collection score deals with delinquency, not with strong performers.
  4. Collections, supported by a fraud score.Why not: Collections deals with delinquent accounts, and a fraud score is for illicit use.

Limit management and cross-selling to existing customers are account management, and the behavioural score, based on account history, is the relevant tool.

The hardest tier

A card customer is expected to generate USD 300 of profit at the end of each year while they remain a customer. The probability of staying for each further year is 80%, and the first year's profit is certain. The discount rate is 10%. The customer's lifetime value is closest to:

  1. USD 3,000Why not: Treats the profit as a perpetuity discounted at 10% and ignores customer attrition.
  2. USD 800Why not: Applies the retention rate to the first year's profit, which was stated to be certain.
  3. USD 1,000 - correct
  4. USD 1,500Why not: Uses the expected lifetime of five years without discounting.

Value = sum over t of 300 x 0.8^(t-1) / 1.1^t = 300 / (1 + 0.10 - 0.80) = USD 1,000. Retention and discounting both matter for customer value.

In the app: 6 practice questions on this outcome, spaced review that brings it back before you forget it.

Start free →

More in Credit Scoring and Retail Credit Risk Management