Purpose-Specific Financial Analytics for Entrepreneurial Credit

Market Screening

Samandarboy Sulaymanov1 Olimjonov Abbosjon Olimjonovich2

1 Tashkent State University of Economics, Tashkent, Uzbekistan

2 Tashkent State University of Economics, Uzbekistan

Received: July 09, 2026 Revised: September 01, 2026 Accepted: October 02, 2026 ⋆ Corresponding author

ABSTRACT

A lender choosing markets for small-enterprise finance may use account ownership as a shortcut for credit demand.

This article asks whether that shortcut identifies places where adults report borrowing to start or operate a business.

We analyze the 2024 country estimates in the World Bank’s Global Findex Database 2025: 98 economies with

complete business borrowing, account, formal-channel borrowing and formal saving measures. Account ownership

and business borrowing are negatively associated across those economies (Spearman ρ = −0.220). An account-only

model has a leave-one-economy-out mean absolute error of 5.529 percentage points, close to a constant forecast

(5.580); adding formal-channel borrowing barely changes it. Adding formal saving reduces error to 4.700 points,

although individual country errors remain substantial. Fifteen economies fall simultaneously above the sample’s

upper quartile of business borrowing and below its median account ownership. In 40 economies with paired gender

estimates, men have higher business borrowing in 19 and women in 21; account gender gaps do not reproduce a

uniform enterprise-credit gap. These are country-level diagnostics of reported borrowing, not loan applications,

unmet credit demand or evidence that account access causes enterprise borrowing. A staged market screen should

therefore use purpose-specific borrowing and validate borrower quality and local lending rules before capital

allocation.

Keywords: Business analytics Entrepreneurial finance Credit market screening Global Findex Out-of-sample

validation

1. INTRODUCTION

A bank or digital lender with a limited expansion budget must

choose where to investigate enterprise loans. Account penetration

is available for many markets and seems to offer a

ready-made ranking. Yet it measures the reach of a financial

relationship, while lending to start or operate a business concerns

a specific use of credit. An economy can have many

account holders and little reported business borrowing, or a

high business-borrowing rate with relatively few accounts.

Mistaking the first number for the second can redirect duediligence

spending.

We examine this failure of a proxy rather than estimating

a treatment effect. The decision is to identify economies

worth investigating, not to estimate the number of creditworthy

firms. The outcome is the share of adults, not registered

enterprises, who report business-purpose borrowing. This distinction

matters because one owner may have several firms,

and the survey does not link each borrowing purpose to a

lender, loan amount, application, rejection or business outcome.

The contribution is a transparent screen with held-out

forecast errors, contrasting evidence across lending and saving

channels, and a paired gender check. All source data and