🧪📊💰

Nano Loan Pilot — Ghana

DR1 · DR10 · LN1 Graduation · Segment Comparison
Mar 2 – Apr 23, 2026 · All LN=0 First Loans · 8-Week View

📅 Mar 2 – Apr 23, 2026 🇬🇭 Ghana Production 📡 Snowflake Ghana Prod · ml.loan_info_tbl × ml.client_info 🔄 Live as of Apr 23, 2026
3,544
Nano Loans Disbursed
33.0%
Nano DR1 · On par with Counter Offer (32.4%)
29.8%
Nano DR10 · Now converged with CO/CG
37.3%
Nano Return Rate of Closed · vs 66% CG

📋 Pilot Context & Segment Definitions

Read First: Experiment Structure

  • Nano Loan launched Mar 2, 2026. This report covers 8 full weeks (Mar-02 → Apr-20) plus partial Apr-23 data. Now 7+ weeks old with maturing cohorts.
  • Nano targets first-time credit users who fail standard eligibility — lowest Fido score band (250–260), no prior repayment history.
  • Segment definitions: Nano = UCNLL_ product · Counter Offer = AC products · Personal = UCLL + GH_FIDO_MODEL% · Control Group = UCLL + GH_CONTROL_GROUP model.
  • Population sampling: ~5% of non-eligible (score 250–260) + 5% of CO-eligible (260–276) clients redirected to Nano. Controlled pilot, not full rollout.
  • DR methodology: IS_DUE_DTR1 / IS_DUE_DTR10 are CURRENT_DATE-based flags. Only mature cohorts (Mar-02, Mar-09 for DTR10; Mar-02 through Mar-16 for DTR1) give reliable rates. Later weeks have insufficient due-loan counts.
  • Maturity note: Nano has a slightly longer avg maturity (32.5 days vs ~29 days for others), which delays when DTR10 flags trigger relative to disbursement date.

Key Development Since Last Report (Apr 12 → Apr 23)

  • DR10 convergence resolves the main prior concern: Nano DR10 went from an apparent outlier (15.1% vs 1.1% CO on Apr 12) to fully converged (29.8% Nano, 28.7% CO, 30.5% CG). The earlier gap was a maturity timing artifact — CO and CG cohorts had not yet reached their DTR10 threshold. At the Mar-02 cohort level, all three segments showed identical DR10 (~30–32%) even at the time.
  • Nano volume grew +30%: 2,716 → 3,544 loans. Counter Offer +28% (2,922 → 3,755). Control Group +31% (1,722 → 2,248).
  • LN1 graduation gap has widened to 29pp: 37.3% of closed Nano clients returned vs 66.0% for Control Group. This is now the primary focus area.

⚡ Executive Summary — DR1 Across All Segments

Personal — DR1

22.6%
2,385 of 10,551 due loans. Score 276+, GH_FIDO_MODEL. Established baseline — lowest risk cohort by design.

Counter Offer — DR1

32.4%
463 of 1,429 due loans. Score 260–276. Primary benchmark for Nano Group A clients.

Nano Loan — DR1

33.0%
282 of 855 due loans. Score 250–260. Statistically indistinguishable from Counter Offer (32.4%) after 8 weeks.

Control Group — DR1

35.4%
326 of 920 due loans. Same UCLL product as Personal but lower scoring cohort (GH_CONTROL_GROUP model).

Core DR1 finding holds after 8 weeks: Nano DR1 (33.0%) remains statistically equivalent to Counter Offer (32.4%) — the established product serving the next score band up. With 855 due loans as the Nano denominator and 3,544 total disbursements, this is a robust signal.

📋 Segment Summary — Live Data (Mar 2 – Apr 23, 2026)

SegmentTotal LN0Avg Loan (GHS) Due DTR1Defaulted DTR1DR1 % Due DTR10Defaulted DTR10DR10 %
Personal 26,769GHS 281 10,5512,38522.6% 5,8471,14419.6%
Counter Offer 3,755GHS 148 1,42946332.4% 77122128.7%
Nano Loan 3,544GHS 60 85528233.0% 81524329.8%
Control Group 2,248GHS 279 92032635.4% 52115930.5%

DR10 interpretation: With all Mar-02/Mar-09 cohorts fully mature, these figures are stable and comparable across segments. Nano DR10 (29.8%) is in line with CO (28.7%) and CG (30.5%), confirming the previous report's elevated Nano DR10 (15.1%) was a timing artifact from immature cohort mix.

📉 Default Rate Analysis

DR1 = loan is >1 day past due and DTR > 1 (early missed payment). DR10 = loan is >10 days past due and DTR > 10 (persistent delinquency). Only loans where the due date has passed are included in each denominator.

DR1 — Overall by Segment

Control Group
35.4%
Nano Loan
33.0%
Counter Offer
32.4%
Personal
22.6%

Nano, CO and CG are clustered within 3pp of each other. Personal is structurally lower due to higher Fido score eligibility (276+).

DR10 — Overall (Mature Cohorts, Converged)

Control Group
30.5%
Nano Loan
29.8%
Counter Offer
28.7%
Personal
19.6%

Nano DR10 is now converged with CO and CG. The 15% vs 1% gap in the Apr 12 report was entirely a cohort maturity timing artifact, now fully resolved.

Weekly DR1 by Disbursement Cohort (Mar-02 → Apr-20)

▼

DR1 is reliable only for cohorts whose loans have had their due dates pass. Mar-02 through Mar-16 are largely mature. Mar-23 onwards have few due loans — treat as directional only.

WeekNano DisbNano DR1CO DisbCO DR1CG DisbCG DR1Personal DisbPersonal DR1Maturity
Mar-0239333.3%41034.9%27935.8%3,01524.7%Mature
Mar-0947534.9%47931.9%31935.1%3,57021.9%Mature
Mar-1648430.8%52832.0%31736.3%3,85622.3%Mature
Mar-2351016.7%*58410.5%*2950.0%*3,8418.3%*Partial
Mar-3047621.4%*49912.5%*28525.0%*3,5018.8%*Partial
Apr-064600.0%*50525.0%*2690.0%*3,2962.7%*Immature
Apr-13501—494—342—3,930—Too early
Apr-20245—256—142—1,760—Too early
TOTAL3,5443,7552,24826,769

* Based on small due-loan counts — directional only. Apr-20 is a partial week through Apr 23.

Weekly DR10 — Mature Cohorts (Mar-02 & Mar-09)

▼

Only Mar-02 and Mar-09 have sufficient DTR10-due loans. The convergence across Nano, CO and CG is visible at the individual cohort level.

WeekNano Due DTR10Nano DR10CO Due DTR10CO DR10CG Due DTR10CG DR10Personal Due DTR10Personal DR10
Mar-0239330.5%41031.7%27930.5%3,01520.9%
Mar-0937330.6%33926.8%23231.5%2,64019.2%
Mar-163023.3%*120.0%*616.7%*1102.7%*

* Mar-16 based on very few DTR10-due loans — not statistically reliable.

Mar-02 cohort (most mature): Nano DR10 = 30.5%, CO DR10 = 31.7%, CG DR10 = 30.5%. All three non-Personal segments perform identically on persistent delinquency at the cohort level.

📈 Disbursement Volumes — 8 Weeks

Nano is disbursing at ~450–510 loans/week on full weeks, comparable to Counter Offer. Average loan size remains GHS 60 — ¼ of a CO loan (GHS 148) and ⅕ of Personal/CG (~GHS 280). Total Nano portfolio since Mar 2: approximately GHS 212,640 across 3,544 loans.

WeekNano LoanCounter OfferControl GroupPersonal
Mar-023934102793,015
Mar-094754793193,570
Mar-164845283173,856
Mar-235105842953,841
Mar-304764992853,501
Apr-064605052693,296
Apr-135014943423,930
Apr-20 (partial)2452561421,760
TOTAL3,5443,7552,24826,769

🔄 Return Disbursement Rate — LN0 → LN1 Graduation

Overall rate = % of all LN0 clients who received a subsequent disbursement (LN≥1). Rate of closed = same, restricted to clients who fully repaid (closeddate IS NOT NULL) — the cleaner post-repayment engagement measure.

Return Rate — Overall (All LN0 Clients)

Personal
31.8%
Counter Offer
29.0%
Control Group
28.2%
Nano Loan
16.9%

Overall rate is lower for Nano because many recent loans haven't matured yet. The "of closed" metric controls for this.

Return Rate — Of Closed LN0 Clients Only

Control Group
66.0%
Personal
63.4%
Counter Offer
60.3%
Nano Loan
37.3%

Nano's 37.3% vs Control Group's 66.0% = −28.7pp gap. This is the primary operational concern for this cycle.

SegmentTotal LN0Closed LN0Returned LN1+Return Rate (Overall)Return Rate (of Closed)Gap vs CG
Personal 26,76913,4228,509 31.8%63.4%−2.6pp
Control Group 2,248960634 28.2%66.0%Baseline
Counter Offer 3,7551,8051,089 29.0%60.3%−5.7pp
Nano Loan 3,5441,602598 16.9%37.3%−28.7pp

LN1 graduation gap is the primary concern. Of 1,602 Nano clients who fully repaid, only 598 (37.3%) received a second loan. Control Group — same UCLL product, similar risk cohort — achieves 66.0%. The 28.7pp gap suggests a routing or product-matching failure in the LN0→LN1 graduation pathway, not a demand problem. Repaid Nano clients are likely not being presented with a clear and accessible LN1 offer at the right moment.

📅 Loan Maturity Terms by Segment

SegmentTotal LoansAvg Maturity (Days)Median (Days)Min (Days)Max (Days)
Personal26,76928.8301033
Counter Offer3,75528.9301033
Control Group2,24828.930933
Nano Loan3,54432.5301052

Maturity structure note: Nano's longer avg maturity (32.5 vs ~29 days) and wider range (max 52 days) means Nano loans take slightly longer to cross the DTR10 threshold. This had inflated Nano's apparent DTR10 advantage in earlier snapshots — not faster repayment, just later due dates. With 8 weeks of history, this effect is fully resolved in the overall figures.

🔍 Key Findings

DR1 Parity with Counter Offer — Holds After 8 Weeks

Nano DR1 (33.0%) has been statistically indistinguishable from Counter Offer (32.4%) across all mature cohorts. With 3,544 total disbursements and 855 due loans, this is a robust and stable finding. First-time borrowers in the 250–260 Fido score band miss payments at the same early rate as the established 260–276 product.

Positive signal — validated

DR10 Convergence — Previous Risk Concern Resolved

The Apr 12 report flagged Nano DR10 (15.1%) as 14× higher than CO (1.1%). With cohorts now fully mature, DR10 has converged: Nano 29.8%, CO 28.7%, CG 30.5%. At the Mar-02 cohort level all three segments showed identical DR10 (~30–32%) even in April. The earlier gap was entirely driven by immature cohort mix in the overall figure — CO and CG cohorts had not yet reached their DTR10 threshold at the time of that snapshot.

Risk concern resolved

LN1 Graduation Gap — 28.7pp Below Control Group

37.3% of repaid Nano clients received a second loan, vs 66.0% for Control Group. The gap was 16pp in the Apr 12 report (38.8% vs 54.5%) — it has widened as the Control Group's re-engagement rate matured. This is the highest-priority issue: repaid Nano graduates are either not being identified for LN1 offers, not being routed to the right product, or not converting when offered. This is not a credit quality issue — it is a product pathway and routing issue.

Primary concern — action required

~30% DR10 Across Riskier Segments — Structural Risk Baseline

Nano, CO and CG all show ~29–31% DR10 on mature cohorts. This establishes the structural persistent-delinquency rate for the 250–276 Fido score population. About 1 in 3 borrowers in this band will be 10+ days late at some point after their due date. For Nano at GHS 60 average, collections economics at DTR 2–10 require careful calibration — the low loan amount reduces the revenue available to fund outreach relative to CO (GHS 148) or Personal (GHS 281) loans.

Monitor & calibrate

Disbursement Pace Steady — Pilot Scale Growing

Nano is disbursing at ~475–510 loans per full week, comparable to Counter Offer (~480–585/week). Apr-13 showed the highest Nano single-week volume (501 loans), suggesting continued organic growth. Total 3,544 Nano loans represents a statistically meaningful pilot scale for policy decisions on Tier B/C activation.

Healthy pace

📊 Delta vs Apr 12 Report

MetricApr 12, 2026Apr 23, 2026ChangeInterpretation
Nano Loans Disbursed2,7163,544+828 (+30%)Steady disbursement pace
Counter Offer Disbursed2,9223,755+833 (+28%)Steady
Control Group Disbursed1,7222,248+526 (+31%)Steady
Nano DR133.6%33.0%−0.6ppStable — no deterioration
Counter Offer DR133.2%32.4%−0.8ppStable
Nano DR1015.1%29.8%+14.7ppMaturity convergence — all segments rose comparably
Counter Offer DR101.1%28.7%+27.6ppConfirms Apr 12 CO figure was a timing artifact
Control Group DR106.5%30.5%+24.1ppConfirms Apr 12 CG figure was a timing artifact
Nano Return Rate (of closed)38.8%37.3%−1.5ppGraduation gap persists — slight decline
CG Return Rate (of closed)54.5%66.0%+11.5ppCG maturing strongly — gap vs Nano widened to 29pp
Personal Return Rate (of closed)49.4%63.4%+14.0ppPersonal also maturing — cohort re-engagement rising

🎯 Recommended Actions