00 — Overview & VerdictExecutive Summary: Visibility Has Doubled, but MSME Credit Value Has Fallen

Eight headline findings

• 1. Visibility has doubled while MSME credit value has fallen. Between 2023 and 2025 P2B value rose 106% and MSME loan accounts rose 105%, but the value of MSME bank loans fell 21% (index 100 to 79). Total bank loans rose 47%.
• 2. More accounts, much smaller loans. The average MSME bank loan fell from TZS 46.5 million (2023) to 34.1 million (2024) to 18.0 million (2025), a 61% fall. MSMEs' share of the bank loan book fell from 39.8% to 21.5%.
• 3. P2B now dwarfs the MSME loan stock. P2B value per shilling of MSME bank loans rose from 1.4 (2023) to 3.6 (2025).
• 4. There is a missing middle in ticket size. The average digital loan is about TZS 16,600; the average MSME bank loan is TZS 18 million; the BoT SME Credit Guarantee Scheme starts at TZS 50 million. The average enabled merchant receives about TZS 1.1 million a month through P2B, so a typical P2B-based working-capital advance would be TZS 0.5–2 million, below every one of these.
• 5. The data plumbing is the first bottleneck. In 2025 the operator–bureau collaboration registered more than 120,000 customers, 0.16% of 75.8 million active mobile-money accounts. The IFC–BoT agreement of October 2026 on alternative data is the right next step but is only starting.
• 6. Guarantees do not yet reach P2B-sized borrowers. The SME-CGS covers loans of TZS 50 million to 1 billion; the 2026 draft regulations for an independent Credit Guarantee Corporation (CGC) do not yet show a micro-ticket window. Cumulative SME-CGS approvals since resumption are only TZS 174 billion.
• 7. The opportunity is real but bounded. Central-case capacity is about TZS 1.6 trillion in 2025 (15% of the MSME bank loan stock); by 2030 pathway B implies capacity of TZS 5.6 trillion and a realised stock of about TZS 1.4 trillion.
• 8. The prize is value addition, not credit-to-GDP. Even in pathway C, P2B-based credit is about 0.5% of 2030 GDP. The reasons to pursue it are inclusion and productivity: cheaper working capital for micro-merchants, a bridge to formal registration, and demand for local tools.

Headline indicators, 2023–2025

Table 1. n/a = not reported on a comparable basis. MSME loan value is from BoT Table 2.3 and the text (Table 2.2 shows 10,498; see Sources & Data Notes).

01 — Question & FrameworkCan the Digital Trail That P2B Creates Become Working Capital and Investment?

The companion reports showed that Tanzania's payment rails are strong and its productive layers thin, and identified the weak data-to-credit link as the single highest-leverage gap. This report examines that link for the enterprises that matter most. MSMEs number five million or more, contribute roughly 35% of GDP and employ roughly half the workforce (TICGL), yet only about one in five has access to formal finance.

Why the question is urgent in 2026

Private-sector credit reached TZS 44.6 trillion in December 2025 (+23.5%) and grew 31.2% in the year to July 2026 (about TZS 53.8 trillion). The Central Bank Rate was raised in July 2026 to 6.25% and held there for the fourth quarter. A Bill tabled in Parliament aims to let banks accept movable assets as collateral and to raise private-sector credit from 22.5% of GDP (2024) to 50% by 2030. The BoT is turning its credit guarantee schemes into an independent corporation. The policy window for P2B-based finance is open now.

Six links from payment to capital

Working capital versus investment capital

Four diagnostics used in this report

Data and method

Method: index P2B and MSME credit series to 2023, compute average loan sizes and shares, build a ticket ladder, size P2B-based credit capacity from stated assumptions, test five hypotheses, and build 2030 pathways that reuse the P2B value paths of scenarios A–D in the P2B report. Two limits matter: MSME loan data cover accounts reported by banks, not unique borrowers; and no data link individual merchants' P2B receipts to their borrowing, so credit capacity is a model, not an observation.

02 — Look-back 2023–2025Visibility Doubled, MSME Credit Value Fell: The Scissors, the Thin Portfolio and the Cost of Credit

The scissors

Index the series to 2023. P2B value reached 206 by 2025. MSME bank loan accounts reached 205. Total bank loans reached 147. But the value of MSME bank loans fell to 79. The business sector's visible revenue doubled while the stock of bank credit to MSMEs shrank by about a fifth, in a year (2025) when overall private-sector credit grew 23.5%. The BoT attributes the lower MSME loan value mainly to repayments and notes that MSME non-performing loans fell to 2.8% from 3.4%. The accounts increase is linked to business formalisation and the expansion of microfinance. The result is a portfolio that has become broader and thinner: more borrowers, smaller loans, and a smaller share of the bank loan book.

Average loan size has collapsed

Dividing MSME loan value by accounts gives an average of TZS 46.5 million in 2023, 34.1 million in 2024 and 18.0 million in 2025 (−61%). MSMEs' share of the bank loan book fell from 39.8% to 35.9% to 21.5% (see the charts on the Overview tab). Part of this is a mix effect: new accounts are likely to be smaller microfinance-type facilities. But it also means that the "growth" in MSME accounts is not growth in the capital available to each enterprise.

Personal loans are still standing in for business loans

The BoT reports that personal loans of TZS 1,165 billion were used to finance small businesses in 2025, more than double the TZS 533 billion of 2024, even though the number of such accounts fell from 57,054 to 24,391. The Bank explains this as MSME operators using personal facilities because of collateral challenges. Personal loans remain 35–36% of private-sector credit overall and are described by the BoT as largely extended to small and medium enterprises. The pattern is that entrepreneurs without collateral borrow as individuals at personal-loan terms; P2B records could let them borrow as businesses.

The chain narrows sharply after payments

There are 75.8 million active mobile-money accounts and 2.79 million enabled merchants, but 578,692 MSME bank loan accounts, and only 120,000 customers registered through operator–bureau collaboration in 2025. Definitions differ (accounts, merchants and customers are not the same units) so the comparison is indicative, but the drop-off of one to two orders of magnitude after the payment layer is clear.

Cost of credit

Formal credit is not cheap, and it is much cheaper than the digital alternative. The BoT reports an overall lending rate of about 15.4% in 2025 (negotiated rate 12.9%). Microfinance Tier 2 institutions and digital lenders price much higher; the BoT has noted high interest rates as a key challenge for Tier 2 borrowers, and TanzaniaInvest reports a BoT-acceptable ceiling of about 4% a month for digital loans (date not stated in the source). The gap of up to three times between bank and digital pricing for the same three-month loan is the opportunity: a P2B-based, automatically repaid facility that lenders can price closer to bank rates because collection risk is low.

Appendix data: BoT MSME and bank credit

Table A1. Source: BoT Annual Financial Inclusion Report 2025, Tables 2.2 and 2.3, as published.

Table A2. Source: TICGL calculation.

03 — Barriers 1–3Where Conversion Breaks: The Ticket Gap, Data Access and Guarantee Fit

Barrier 1: the ticket gap

Credit products in Tanzania cluster at two ends. At the bottom are digital loans averaging about TZS 16,600 and small state programmes (the NEEC funds averaged about TZS 280,000 per beneficiary in 2025, reaching 8.7 million MSMEs with TZS 2.44 trillion). At the top are bank MSME loans averaging TZS 18 million, local-government 4-4-2 loans averaging about TZS 9.7 million, and the SME Credit Guarantee Scheme, which covers loans of TZS 50 million to TZS 1 billion. A P2B-typical merchant, receiving about TZS 1.1 million a month on average (or about TZS 3 million a month for an active Vodacom merchant, indicative), sits in between.

Barrier 2: data access and credit infrastructure

Merchants' P2B records sit with e-money issuers. The BoT has made progress on infrastructure: it reviewed the Credit Reference System Regulations (2019) in 2025, deployed a Credit Reference Data system with bank onboarding under way, and cooperation between mobile network operators and credit bureaus registered 120,000 customers. National ID coverage reached about 81% of eligible adults, with 138 institutions integrated for e-KYC. The IFC agreement of October 2026 will explore alternative data in credit assessments. But a consented, portable merchant record that any licensed lender can use does not yet exist at scale.

The legal base exists (Personal Data Protection Act Cap. 44 and its 2023 regulations; Financial Consumer Protection Regulations amended in 2025; the 2024 Guidance Note on Digital Lenders). Recent legal analysis identifies two gaps relevant here: no requirement for transparency in automated credit scoring, and coordination between the BoT, the Personal Data Protection Commission and self-regulatory associations.

Barrier 3: collateral and guarantee fit

Collateral is the stated reason MSME operators use personal loans. The Movable Property Security Rights Bill, if enacted, will let banks take movable assets (stock, equipment, receivables) as security. The guarantee architecture is being rebuilt: seven schemes operate, the BoT SME-CGS has approved a cumulative TZS 173.98 billion since resumption (TZS 42.89 billion in 2025), and the 2026 draft regulations would create an independent CGC jointly owned by Government and the private sector. The problem is fit. The SME-CGS targets loans from TZS 50 million to 1 billion for firms with turnover up to TZS 10 billion; the average enabled merchant's annual P2B receipts are about TZS 13 million.

Table A3. Credit support instruments.

04 — Barriers 4–6Product Design and Consumer Protection, Informality, and Tools and Capability

Barrier 4: product design, pricing and consumer protection

Digital lending has expanded fast, with women holding 46.3% of digital loan accounts, but conduct problems have accompanied it. The BoT closed 69 unlicensed digital lenders in 2024 and legal analysis documents unclear pricing, short repayment periods and automatic wallet deductions that can empty a borrower's wallet. A P2B-based advance repaid by deduction from incoming payments is efficient, but the deduction rate must be visible and adjustable. The Afya Mkopo product allows borrowers to choose a repayment share of 33% to 100% of each incoming payment, with 33% as default, which is a useful design benchmark.

Barrier 5: informality

Only about 12.5% of MSMEs are formally registered and about 72% of SMEs are informal (TICGL). Lenders need a legal entity and a licence, and the mandatory electronic payments regime (GN 158C, in force 1 July 2026) now requires proof of a digital payment facility for business licensing. The BoT notes that formalisation drove the doubling of MSME loan accounts, which suggests that formalisation and credit are linked in practice. For P2B-based lending, the practical design question is whether a merchant with a payment history but no licence can borrow a small amount, or whether the credit file requires formal registration first.

Barrier 6: tools and capability

A payment history is hard to underwrite when the merchant has no bookkeeping, no separation between business and personal wallets, and no inventory record. Smartphone penetration was 41.8% in 2025 (feature phones 87.1%), so USSD-first tools matter. MSME e-commerce use was 15% in 2023 against a FYDP IV target of 60% by 2031. The BoT is training Certified Financial Educators (1,167 in 2025) but delivery was concentrated in Dar es Salaam (about half of outreach). Simple bookkeeping embedded in merchant wallets is the missing complement to P2B data.

05 — Five Pathways & CapacityFive Conversion Pathways and What They Could Be Worth

The research notes provided for this study set out three pathways: transaction-based lending, cash-flow underwriting and embedded finance. The evidence supports those three and suggests two more: portfolio guarantees and asset finance. Each is assessed below.

Table 3. Five pathways.

Sizing P2B-based credit capacity

s is the share of P2B value coming from merchants whose records are credit-visible (active for six months or more, regular inflows, identifiable) and a is the advance as a multiple of monthly receipts. The realised stock is capacity multiplied by take-up. The shares and multiples are assumptions, not observations.

Table 4. P2B-based credit capacity in 2025, TZS trillion outstanding, from P2B value of TZS 37.52 trillion. Compare MSME bank loan stock of TZS 10.41 trillion. The 5%/10% illustration in the TICGL research note (TZS 190 billion) falls within this range.

Two tiers of merchant, two kinds of product

The BoT does not publish the distribution of merchant receipts. As an illustrative test, suppose the top 5% of merchants (about 139,000) hold half of P2B value. Their average receipts would be about TZS 11 million a month and a one-month advance would be about TZS 11 million, in the range of the average MSME bank loan. The other 95% would average about TZS 0.6 million a month, and an advance of 0.5–2 times receipts would be TZS 0.3–1.2 million. The research note's example merchant (TZS 8–15 million a month) is therefore a top-tail merchant, not the typical one. Vodacom's data point the same way: its active merchants receive about TZS 3 million a month on average (2026, indicative), nearly three times the all-merchant average of TZS 1.1 million (2025).

06 — Capital Ladder & HypothesesFrom Working Capital to Investment, and Five Hypotheses Tested

Working capital is what most P2B-based products can deliver. Investment capital needs more. The table sets out the evidence each rung requires.

Table 5. The capital ladder.

The investment rungs depend on formalisation and collateral reform, which are the Movable Property Bill and the BRELA online registration system, more than on payment data. The role of P2B is to make the transition from micro-advance to line-of-credit faster by building a record, and to make the larger loans safer by supplying evidence of cash-flow.

Other sources of capital also matter

Testing five hypotheses

Four hypotheses are supported; H5 awaits pilot evidence.

07 — Outlook 2026–2030 & ImpactFour Pathways to 2030: P2B-Based Credit Is Real but Bounded

The pathways reuse the P2B value paths of scenarios A–D in the companion P2B report (TZS 99.4, 121.7, 147.1 and 70.9 trillion in 2030). They add three assumptions for each: the credit-visible share s, the advance multiple a, and take-up u. All are assumptions.

Table 6. Source: TICGL scenario model; P2B paths from the companion P2B report; GDP from the companion digital economy report. Capacity = P2B ÷ 12 × s × a. Disbursement assumes an average three-month tenor (four turns a year). Illustrative.

Economic impact and limits

What P2B-based credit can and cannot do

The scale is modest by national standards. Pathway B's realised stock is about 0.4% of 2030 GDP, and the Movable Property Bill's target would require private-sector credit to rise by about 28 percentage points of GDP. P2B-based finance will not deliver that target. Its purpose is different: it widens the base of borrowers, lowers the cost of working capital for micro-merchants, builds credit files and brings enterprises into the formal system.

08 — Policy & KPIsEight Actions and a KPI Dashboard for P2B-Based MSME Finance

Eight actions follow from the evidence. They are aligned with NFIF3 (2023–2028), the 2026/27 Budget measures on credit guarantees, GN 158C and the IFC–BoT agreement.

Table 7. Policy and market actions.

Proposed KPI dashboard

Table 8. Targets are TICGL proposals, not official government commitments.

09 — Risks & Data GapsRisks, Data Gaps and the TICGL/TERI Research Agenda

Risks

Data gaps

Research agenda for TICGL/TERI

1. Merchant panel survey. About 1,000 merchants in Dar es Salaam, regional towns and rural markets, on receipts, borrowing, tool use and registration.
2. Pilot of Pathways 1 and 2. Design with one operator and one bank or microfinance lender, on a consented sample, to measure take-up, cost and repayment.
3. Data annex. Update the P2B report's MSME credit-account figure to the BoT 2025 data and publish a joint three-report data annex.
4. KPI tracking. Track the KPI dashboard annually and review the credit-capacity assumptions as data become available.