01 — OverviewExecutive Summary
Tanzania's external debt stock rose from roughly TZS 85.83 trillion to TZS 95.09 trillion between July 2025 and July 2026 — an increase of about TZS 9.26 trillion. The natural assumption is that this entire increase reflects new borrowing. It does not. Using the standard debt-dynamics decomposition that the IMF and World Bank apply in formal debt sustainability analysis, TICGL/TERI finds that only about six-tenths of that growth came from actual net new borrowing. The rest came from something no borrowing decision caused at all: the shilling losing value against the currencies Tanzania's debt is denominated in.
Five findings frame this analysis.
- Of the TZS 9.26 trillion increase in external debt over the past year, TZS 5.76 trillion (62.2%) came from net new borrowing, and TZS 3.50 trillion (37.8%) came purely from currency depreciation — a "valuation effect" that would have occurred even with zero new loans.
- Tanzania's currency exposure is concentrated, not diversified. 65.3% of external debt is denominated in a single currency, the US dollar, with a further 18.2% in Euros — leaving little room for one currency's movement to offset another's.
- A genuine zero-new-borrowing year would still not mean a flat debt stock. Holding the current USD-denominated debt stock exactly constant, a repeat of 2024's 9.04% shilling depreciation alone would add roughly TZS 8.6 trillion to the TZS-value of the debt — with no new loan signed.
- The shilling's movement has not been one-directional — Tanzania saw depreciation of under 3.5% in five of the last seven years, a sharp 9.04% depreciation in 2024, and a 2.30% appreciation in 2025. The risk is volatility and tail events, not guaranteed annual depreciation.
- This decomposition does not appear anywhere in Tanzania's official monthly debt reporting — a direct extension of a gap TICGL/TERI has flagged before, and one that is straightforward to close using data Tanzania's own institutions already publish.
- This year, the effect was mostly a "paper" one. Actual cash debt-service costs over the same twelve months were roughly unchanged by currency movement, since the shilling both strengthened and weakened at different points in the year — but a repeat of 2024's sharper single-year move would turn this into a real cash cost of a similar order of magnitude.
Read this alongside TICGL's flagship Dira 2050 policy-gaps analysis
A debt stock that can grow by trillions of shillings without a single new borrowing decision is a fiscal-space risk Dira 2050's financing plans need to price in explicitly, not discover after the fact. TICGL/TERI recommends reading this technical decomposition alongside the Dira 2050 policy-gaps piece.
Read: What's Next for Tanzania's Economy? The Policy Gaps Keeping $1 Trillion Out of Reach by 2050 →About TERI — TICGL's Research Institute
This decomposition was produced under the Tanzania Economic Research Institute (TERI), TICGL's affiliate research institute, applying the same public-debt-dynamics methodology used in IMF and World Bank Debt Sustainability Analyses to Tanzania's own published debt statistics — the kind of cross-referencing TERI's macroeconomics & public finance research pillar exists to do.
Visit TERI — teri.ticgl.com →02 — At a GlanceThe Decomposition in Numbers
Where the TZS 9.26 Trillion Increase Actually Came From
Source: TICGL/TERI decomposition of Bank of Tanzania, Monthly Economic Review, August 2026, Table A10 (external debt stock and end-of-period exchange rate, Jul-25 and Jul-26).
03 — MethodologyHow the Decomposition Was Built
Public debt dynamics analysis conventionally splits the change in a foreign-currency debt stock into two parts: a flow effect (the change in the underlying foreign-currency amount owed — new disbursements minus repayments, minus any debt relief) and a valuation effect (the change in the local-currency value of the debt that was already outstanding, caused purely by exchange-rate movement). This is the same logic the IMF and World Bank apply to the "residual" and exchange-rate terms in a formal Debt Sustainability Analysis.
ΔDebt (TZS) ≈ (ΔDebt in USD) × Exchange rate at end of period + (Debt in USD at start of period) × (Change in exchange rate)
Flow effect: (35,885.6 − 33,712.4) million USD × TZS 2,649.7/USD ≈ TZS 5.76 trillion.
Valuation effect: 33,712.4 million USD × (TZS 2,649.7 − TZS 2,545.8) ≈ TZS 3.50 trillion.
Sum: TZS 9.26 trillion — matching the Bank of Tanzania's own reported change in the TZS-converted external debt stock almost exactly, which is the internal-consistency check this kind of decomposition should pass.
This decomposition applies specifically to Tanzania's external (foreign-currency) debt, which made up 70.7% of the national debt stock at end July 2026. Domestic, TZS-denominated debt does not carry this valuation effect — its growth reflects actual new domestic borrowing and interest accrual, not currency movement. The analysis also uses end-of-period exchange rates as a simplifying convention, consistent with how Bank of Tanzania itself reports the converted debt stock each month.
1. Flow Effect vs. Valuation Effect: The Two Forces Behind Debt Growth
62.2% borrowing, 37.8% currencyOver the twelve months to July 2026, Tanzania's external debt grew by 6.4% when measured in the currency it was actually borrowed in — US dollars, Euros, and others. But once converted to shillings, the same debt grew by 10.8%. That 4.4-percentage-point gap is not a rounding artifact; it is the currency-valuation effect, and it is large enough to change how the headline growth number should be read.
| Component | TZS Trillions | Share of Total Increase |
|---|---|---|
| Net new borrowing (flow effect) | 5.76 | 62.2% |
| Currency depreciation (valuation effect) | 3.50 | 37.8% |
| Total increase | 9.26 | 100.0% |
Source: TICGL/TERI decomposition of Bank of Tanzania, Monthly Economic Review, August 2026, Table A10.
For every TZS 10,000 that Tanzania's external debt grew in shilling terms over the past year, roughly TZS 3,780 was not the result of any borrowing decision at all — the government, in effect, "owed more" purely because the shilling was worth less against the currencies that debt is denominated in.
2. The Zero-New-Borrowing Stress Test
Debt still rises under every depreciation scenarioTo make the point concrete, this section holds Tanzania's external debt stock exactly frozen at its July 2026 level in US dollars — USD 35,885.6 million — and asks a single question: with genuinely zero net new borrowing, what would the TZS-value of that same debt be under different currency scenarios?
| Shilling Depreciation | Implied TZS/USD Rate | TZS Value of Debt | Increase From Currency Alone |
|---|---|---|---|
| 0% (rate unchanged) | 2,649.7 | 95.09tn | — |
| 5% | 2,782.2 | 99.84tn | +4.75tn |
| 9.04% (2024's actual move, repeated) | 2,889.2 | 103.68tn | +8.60tn |
| 10% | 2,914.7 | 104.60tn | +9.51tn |
| 15% | 3,047.2 | 109.35tn | +14.26tn |
TICGL/TERI illustrative stress test, not an official forecast. The 9.04% scenario reuses Tanzania's actual 2024 annual-average currency move (see Section 4) as a realistic reference point, not a prediction.
External Debt Under Zero New Borrowing, by Depreciation Scenario
Source: TICGL/TERI calculation based on Bank of Tanzania external debt stock, Jul-26.
A debt ceiling, a borrowing plan, or a public commitment to "slow debt growth" that is defined only in terms of new loans signed can be technically honoured in full — zero new loans — and the debt-to-GDP ratio can still rise meaningfully. Any credible debt target needs to account for this scenario explicitly, not treat it as a residual surprise.
3. Why the Currency Concentration Makes This Worse
Two-thirds of external debt in one currencyThis valuation effect is not unique to Tanzania — every country with foreign-currency debt has some version of it. What amplifies it for Tanzania specifically is the lack of currency diversification within that external debt book: 65.3% is denominated in US dollars alone, with a further 18.2% in Euros. Only 16.5% is spread across the Chinese Yuan and all other currencies combined.
Currency Composition of Tanzania's External Debt
Source: Bank of Tanzania, Monthly Economic Review, August 2026, Table 2.6.4.
If Tanzania's external debt were spread evenly across a dozen currencies with imperfectly correlated movements against the shilling, currency strength in some would tend to offset weakness in others, dampening the overall valuation effect. With 83.5% of the book concentrated in just two currencies (USD and EUR), there is very little of that natural offsetting — when the dollar strengthens broadly against the shilling, the effect hits almost the entire foreign-currency book at once.
4. The Historical Volatility Record: Not Constant Depreciation, But Real Tail Risk
Mostly calm, with one sharp shock yearIt would be inaccurate to frame this as "the shilling always depreciates." It doesn't. Tanzania's annual-average exchange rate against the US dollar moved by less than 3.5% in five of the last seven years — genuinely modest by regional standards. But 2024 saw a sharp 9.04% depreciation in a single year, and 2025 partially reversed it with a 2.30% appreciation. The honest characterisation is volatility with tail risk, not a one-way trend.
Tanzania's Annual TZS/USD Movement, 2019-2025
Source: Bank of Tanzania, Monthly Economic Review, August 2026, Table A1 (annual-average exchange rate, 2018-2025).
Cumulatively, the shilling depreciated by about 12.1% against the US dollar between 2018 and 2025 — an average of roughly 1.6% a year compounded. But that average masks the reality that almost all of the cumulative move happened in a single year, 2024. Debt planning built around a smooth "average" depreciation assumption would have been badly wrong-footed in 2024 specifically, and could be again.
5. Closing the Reporting Gap: Why This Belongs in Official Publications
Currently absent from monthly and annual debt tablesThis decomposition is not exotic. It is a standard, well-established component of the IMF-World Bank debt sustainability framework, and Tanzania's own institutions already publish every input it requires — the external debt stock in USD, and the exchange rate, both monthly. What's missing is the calculation itself, published as a standard line item rather than left for outside analysts to reconstruct.
This is a direct extension of a gap TICGL/TERI has identified before: that Tanzania's monthly debt reporting shows the debt stock as an unscaled number, with no debt-sustainability framing attached (see What Doesn't Tanzania's Monthly Economic Review Show?). The flow-vs-valuation split described here is a concrete, low-cost addition that would close part of that same gap.
- Separate "new borrowing" accountability from "currency risk" accountability — a finance ministry team responsible for new loan approvals should not be judged by a debt-growth number that is partly outside their control.
- Set more credible medium-term debt targets — a target expressed only as "gross new borrowing per year" leaves a currency-sized blind spot; a target that accounts for expected valuation effects is more likely to hold up.
6. The Scale of the Current Impact: What TZS 3.5 Trillion Actually Means Today
A real "paper" effect — this year's cash cost was a near-washIt would be easy to overstate this analysis by implying Tanzania physically paid TZS 3.50 trillion in extra cash this year because of the shilling. It didn't — and being precise about why matters for credibility. External debt-service obligations are fixed in the currency they were borrowed in and scheduled years in advance; the valuation effect changes the reported TZS-equivalent value of the stock, not the USD amount actually owed on any given payment date.
TICGL/TERI totalled Tanzania's actual external debt-service payments over the trailing twelve months (August 2025 - July 2026): USD 2,654.1 million, costing TZS 6.679 trillion at the exchange rates prevailing each month. Had the exchange rate simply stayed flat at July 2025's level the entire year, that same USD 2,654.1 million would have cost TZS 6.757 trillion — actually slightly more. The reason: the shilling partially strengthened between August and December 2025 before weakening later in the year, so the path was not one-directional, and the net cash effect of currency movement on this particular year's debt-service bill was close to a wash.
So where did the TZS 3.50 trillion go, if not into this year's cash debt-service bill? It sits in the reported stock — the number used for debt-to-GDP ratios, rating-agency and IMF debt sustainability assessments, and the base against which any nominal debt ceiling is measured. That matters for three concrete reasons, even without an immediate cash cost:
TICGL/TERI calculation comparing the valuation effect (Section 1) to Bank of Tanzania, Monthly Economic Review, August 2026, Chart 2.5.2 (Central Government Expenditure, June 2026).
This year's near-neutral cash outcome was a function of timing, not a reason to discount the risk. Applying 2024's actual 9.04% shock uniformly to this same twelve months of debt-service payments would have cost approximately TZS 7.28 trillion instead of TZS 6.68 trillion — about TZS 0.60 trillion more, in cash, in a single year. The stock effect is the leading indicator; a repeat of a 2024-style shock is what would turn it into a genuine cash-flow problem.
- This year: the currency effect was a real TZS 3.50 trillion addition to the reported debt stock, but roughly neutral for actual cash debt-service spending — a distinction worth stating plainly rather than glossing over.
- Next time the shilling moves sharply in one direction, as it did in 2024: the same mechanics that produced a near-wash this year would instead produce a real, additional cash cost of a similar order of magnitude to the TZS 0.60 trillion single-year estimate above — competing directly with development and wage spending for the same government revenue.
06 — SynthesisCross-Cutting Synthesis
62.2% of the past year's external debt growth reflects genuine new net borrowing — the majority, but well short of the full picture.
37.8% of the increase — TZS 3.50 trillion — came from the shilling alone, entirely independent of any borrowing decision.
83.5% of external debt sits in just two currencies (USD and EUR), leaving little natural offsetting when either moves.
Every input this decomposition needs is already published monthly — publishing the calculated split itself is the missing step.
TZS 3.50 trillion moved the reported debt stock; actual cash debt-service cost was roughly neutral this year, since the shilling's path wasn't one-directional.
A repeat of 2024's 9.04% single-year shock would have added an estimated TZS 0.60 trillion in real cash debt-service cost — the tail risk this year's timing happened to avoid.
07 — RecommendationsPolicy Recommendations
Priority 1 — Publish the Flow-vs-Valuation Split Monthly
- Add a standard line to Tanzania's monthly and annual debt reporting decomposing the change in external debt into net-new-borrowing and currency-valuation components, using the methodology set out here.
Priority 2 — Diversify New Borrowing's Currency Mix Where Feasible
- Weigh currency diversification explicitly when structuring new external financing, to reduce the concentration in USD and EUR that currently amplifies a single exchange-rate move across most of the debt book.
Priority 3 — Build Reserve and Hedging Buffers Against Tail-Risk Years
- Given that most of the cumulative 2018-2025 depreciation happened in a single year (2024), size foreign-exchange reserve buffers and consider hedging instruments against the possibility of another single-year shock, not just a smooth average trend.
Priority 4 — Define Debt Targets to Explicitly Account for Valuation Effects
- Any medium-term debt ceiling or target should state how it treats currency-driven changes in the TZS-value of external debt, so that a zero-new-borrowing year is not mistaken for a zero-debt-growth year.
"Tanzania could freeze every new external loan today and its debt, measured in the currency the budget is actually planned in, would still grow the next time the shilling weakens. That is not a criticism of borrowing policy — it is a reminder that debt policy and currency policy are the same policy for two-thirds of what Tanzania owes abroad."
— TICGL / Tanzania Economic Research Institute (TERI)
08 — Sources & Data NotesReferences, Data Sources and Limitations
Bank of Tanzania, Monthly Economic Review, August 2026 — Table A10 (National Debt Developments, including monthly external debt stock and end-of-period exchange rate), Table 2.6.4 (currency composition of disbursed outstanding external debt), and Table A1 (annual-average exchange rate, 2018-2025). The flow-vs-valuation decomposition, zero-borrowing stress test, and all percentage splits are TICGL/TERI's own calculations applying standard public-debt-dynamics methodology to these published figures.
- Primary data: Bank of Tanzania — Monthly Economic Review, August 2026.
- Methodology reference: The flow-versus-valuation decomposition follows the standard approach used in IMF and World Bank Debt Sustainability Analyses (DSA) for separating exchange-rate effects from net borrowing in public debt dynamics.
- Known limitations: The decomposition uses end-of-period exchange rates as a simplifying convention and does not separately isolate debt forgiveness, interest capitalisation, or valuation changes on non-USD currencies within the "flow effect" term — a fully currency-by-currency decomposition would refine, but is unlikely to materially change, the approximate 60/40 split found here. The zero-new-borrowing scenarios in Section 2 are illustrative stress tests, not forecasts.
09 — Quick AnswersFrequently Asked Questions
Can a country's debt grow even if it doesn't borrow any new money?
Yes, if a meaningful share of that debt is denominated in foreign currency. When the local currency depreciates, the local-currency value of foreign-currency debt rises automatically, even with zero new borrowing. This is known as a valuation or exchange-rate effect, and it is a standard component of the IMF-World Bank public debt dynamics framework.
How much of Tanzania's recent debt growth came from currency depreciation rather than new borrowing?
Of the TZS 9.26 trillion increase in Tanzania's external debt stock (in shilling terms) between July 2025 and July 2026, TICGL/TERI's decomposition finds that approximately TZS 5.76 trillion (62.2%) came from net new borrowing, while approximately TZS 3.50 trillion (37.8%) came purely from the shilling's depreciation against major currencies during that period.
Why is Tanzania particularly exposed to this currency effect?
Because 65.3% of Tanzania's external debt is denominated in a single currency, the US dollar, with a further 18.2% in Euros. This concentration means a large share of the debt stock moves in local-currency terms with a small number of exchange rates, rather than being spread across a diversified currency basket that might partially offset each other.
Has the shilling always depreciated against the dollar?
No. Tanzania's exchange rate history from 2018 to 2025 shows mostly small annual moves (under 3.5% in five of seven years), one sharp depreciation of about 9.0% in 2024, and an appreciation of about 2.3% in 2025. The risk is not constant depreciation but volatility, including the possibility of another sharp single-year move.
What can policymakers do about currency-driven debt growth?
Common approaches include diversifying the currency composition of new borrowing, using hedging instruments for large foreign-currency debt-service payments, building foreign-exchange reserve buffers ahead of major repayments, and publishing the borrowing-versus-valuation decomposition of debt growth as a standard part of official debt reporting so the distinction is visible to policymakers and markets.
Did Tanzania actually pay more cash for debt service this year because of the shilling?
Not materially. TICGL/TERI found that actual external debt-service payments over the trailing twelve months (August 2025-July 2026) cost about TZS 6.68 trillion, close to the roughly TZS 6.76 trillion it would have cost had the exchange rate stayed flat, because the shilling's path included both strengthening and weakening over the period. The TZS 3.50 trillion valuation effect shows up in the reported debt stock, not primarily in this particular year's cash debt-service bill — though a sharper single-year move, like 2024's, would change that.
Muhtasari kwa Kiswahili
Je, Nini Kingetokea Tanzania Ikisimamisha Mikopo Mipya Kesho? Deni Bado Lingezidi Kuongezeka — Hivi Ndivyo Kiasi Chake. TICGL/TERI wamefanya uchambuzi wa kitaalamu, ukitumia mbinu ile ile inayotumiwa na IMF na Benki ya Dunia kwenye tathmini rasmi za uwezo wa kulipa deni, kubainisha ni kiasi gani cha ongezeko la deni la nje la Tanzania kilitokana na mkopo mpya halisi, na kiasi gani kilitokana tu na shilingi kupoteza thamani.
Matokeo makuu: Kati ya Trilioni 9.26 za ongezeko la deni la nje la Tanzania (Julai 2025 hadi Julai 2026), Trilioni 5.76 (asilimia 62.2) zilitokana na mkopo mpya halisi, wakati Trilioni 3.50 (asilimia 37.8) zilitokana tu na shilingi kushuka thamani dhidi ya Dola na sarafu nyingine — bila mkopo mpya wowote. Kwa sababu asilimia 65.3 ya deni la nje liko kwa Dola pekee (na asilimia 18.2 zaidi kwa Euro), Tanzania haina "kinga ya asili" inayotokana na kuwa na sarafu mbalimbali. Hata kama Tanzania ingesimamisha mikopo mipya yote leo, marudio ya mshtuko wa mwaka 2024 (shilingi kushuka thamani kwa asilimia 9.04) pekee yangeongeza takribani Trilioni 8.6 kwenye thamani ya deni hilo hilo kwa shilingi — bila mkopo mmoja mpya. Trilioni 3.50 hizi ni sawa na miezi 1.8 ya bajeti nzima ya maendeleo ya Serikali, au asilimia 74 ya matumizi yote ya Serikali ya mwezi mmoja. Kwa uaminifu zaidi: malipo halisi ya deni mwaka huu (fedha taslimu zilizotoka) hayakuongezeka sana kwa sababu ya sarafu, kwa kuwa shilingi ilikuwa na nguvu kwa baadhi ya miezi — lakini marudio ya mshtuko wa 2024 yangeweza kuongeza takribani Trilioni 0.6 za malipo halisi ya ziada ndani ya mwaka mmoja tu. Historia inaonesha shilingi haishuki thamani kila mwaka kwa kiwango kikubwa (miaka mitano kati ya saba iliona mabadiliko chini ya asilimia 3.5), lakini mwaka 2024 ulionesha hatari halisi ya "tail risk" ya mshtuko wa ghafla.
- Ongezeko la Deni la Nje (Julai 2025-2026): Trilioni 9.26 za TZS
- Kutokana na Mkopo Mpya: Trilioni 5.76 (asilimia 62.2)
- Kutokana na Sarafu Pekee: Trilioni 3.50 (asilimia 37.8) — sawa na miezi 1.8 ya bajeti ya maendeleo
- Deni la Nje kwa Dola: asilimia 65.3 — utegemezi wa sarafu moja
- Mfano wa "Hatari Kubwa": Marudio ya mshtuko wa 2024 (asilimia 9.04) = Trilioni 8.6 za ziada kwenye deni, na Trilioni 0.6 za ziada kwenye malipo halisi ya mwaka mmoja
Chanzo: Tanzania Economic Research Institute (TERI), kwa ajili ya TICGL, uchambuzi wa athari ya sarafu kwenye deni la Tanzania, ukitegemea takwimu za Benki Kuu ya Tanzania (BOT), Agosti 2026.