Two applicants with slightly different credit records are frequently offered exactly the same rate, while a small further difference produces a much larger jump. Banding explains both.
Pricing covers expected loss across a group
A lender cannot know which individual borrowers will default. It can estimate what share of a group with similar characteristics will, based on past behaviour of comparable applicants.
The rate charged must cover funding costs, operating costs and that expected loss, plus a return on the capital held against the possibility of worse-than-expected outcomes.
Since the loss estimate applies to the group rather than the person, everyone within the group is priced from the same starting point.
Bands make the process operable
A continuous pricing curve would require a distinct rate for every possible combination of characteristics, which complicates disclosure, comparison and regulatory oversight alike.
Grouping applicants into a manageable number of tiers allows rates to be published, marketed and compared, and allows decisions to be automated consistently.
The consequence is a step function. Moving within a band changes nothing, and crossing a boundary changes the rate noticeably.
Capital requirements reinforce the steps
Lenders must hold capital against their loan books, and the amount required rises with the assessed riskiness of the exposure.
Regulatory frameworks themselves define categories, so a loan that falls into a higher-risk bucket consumes more capital and must earn more to justify itself.
This transmits the banding from the regulatory framework into consumer pricing, independent of what any individual lender would otherwise choose.
Advertised rates apply to the best band
Headline rates are typically offered to a minority of applicants, with rules in many jurisdictions requiring that a stated proportion of successful applicants actually receive them.
Everyone else is offered a rate determined after assessment, which is why the figure quoted in advertising and the figure offered on application often differ.
Soft searches and eligibility checks exist to reveal the likely band without recording a hard enquiry that would itself affect the assessment.
Why small improvements sometimes pay off sharply
Because the pricing moves in steps, a modest improvement in an application can be worth nothing or can be worth a substantial rate reduction, depending on proximity to a boundary.
Reducing reported balances, correcting errors on a credit file or waiting for an adverse marker to age can each move an application across a threshold.
The bands themselves are not published, so the effect cannot be predicted precisely, but the step structure explains why outcomes appear disproportionate to the change made.