Trade Credit and Digital Records: Evidence from Côte d'Ivoire

How to cite

Wiles, Edward; Houeix, Deivy. 2026. "Trade Credit and Digital Records: Evidence from Côte d'Ivoire." AEA RCT Registry. AEARCTR-0018537. https://doi.org/10.1257/rct.18537-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Edward Wiles* , Deivy Houeix

* Primary investigator

Registry ID
AEARCTR-0018537
DOI
https://doi.org/10.1257/rct.18537-1.0
First registered
Last updated
Status
On going
Trial dates
2026-04-30 – 2027-07-31
Countries
Côte d'Ivoire ()
Registry record
http://www.socialscienceregistry.org/trials/18537

Abstract

Trade credit—suppliers allowing buyers to purchase inventory now and pay later—is central to firm-to-firm commerce but depends on trust. In low-enforcement environments, the same information frictions that ration trade credit also block access to formal lenders (MFIs, banks, fintechs). We ask: can firms use mobile money sales records to signal creditworthiness to suppliers and formal lenders — an open finance model — and how does this shape credit access and market outcomes? We run an RCT in Abidjan in partnership with a large mobile money operator and a fintech that specializes in trade credit. We randomly assign 3,000 retailers in the packaged drinks sector to one of three mobile money groups: control, access to a dashboard of their aggregated sales records, or dashboard access with an option to export a branded PDF containing a digital business card and these records. We cross-randomize these with an introduction to the fintech to test whether specialized lenders are better able to translate digital records into financing than typical suppliers. Outcomes include credit access, supplier relationships, mobile money usage, product variety, prices, and stockouts.

Experimental design

There are six total cells. There are three mobile money groups: control, access to a dashboard of their aggregated sales records, or dashboard access with an option to export a branded PDF containing a digital business card and these records. We cross-randomize these with an introduction to the fintech to test whether specialized lenders are better able to translate digital records into financing than typical suppliers.

Among firms in the two treatment groups with dashboard access, we also randomize (with equal probability) whether the dashboard displays the total mobile money fees paid (at the same horizon at the sales records).

Design details

Randomization method
Randomization done on a computer.
Randomization unit
Randomization is at the firm level. However, we also incorporate a saturation design, which is randomized at the neighborhood level.
Clusters
235
Observations
3000
Arms
500 each
Primary outcomes
Outcomes include credit access, supplier relationships, total input purchases, sales and profits, mobile money usage, product variety, prices, and stockouts.

Cited by

No citation data available for this registration yet.

Keywords