Global Maize Shocks, Exchange Rates, and Domestic Consumer Prices in Ghana, Nigeria, Kenya, and South Africa

How a dollar maize shock does—or does not—become a local food-price shock once currency movements and domestic price structures intervene.

Published: 2026-06-28 | KANA AI Research

Executive Summary

Global maize prices matter for African inflation, but they do not pass cleanly or uniformly into domestic consumer prices. The literature is clear on the mechanism: the world-price shock arrives in US dollars, the exchange rate either amplifies or cushions it, and domestic market structure, trade policy, inventories, and food-basket weights determine how much of that shock reaches households.1234 In African settings, pass-through from international food prices to headline inflation is typically partial—often in the 5–15% range for aggregate inflation—while transmission into food-specific indices is usually materially larger.13

Drawing on monthly series for the global maize benchmark, national exchange rates, and domestic consumer prices, the clearest empirical result in this four-country set is not a broad, uniform maize-to-CPI corridor. It is a narrower pattern. Ghana shows the strongest evidence that exchange-rate movements help predict food-price movements: the cedi/USD rate Granger-causes Ghana’s food CPI at the 5% level over 1998–2026, while the direct global maize-to-food-CPI link is weak.1213 Kenya and Nigeria show no robust short-run predictive transmission from the global maize benchmark or the exchange rate into headline CPI over the available monthly samples. South Africa’s available exchange-rate overlap is too short for a reliable dynamic system estimate, so the South African result is best read from the literature and descriptive evidence rather than from a full multivariate monthly model.459

The practical conclusion is straightforward. The exchange rate is the more credible transmission corridor than the world maize price alone, but even that corridor is country-specific and depends heavily on the domestic price endpoint being measured. Ghana’s food CPI is the only food-specific monthly endpoint in this set, and it is precisely there that the exchange-rate link is clearest.13 Nigeria, Kenya, and South Africa are assessed mainly through headline CPI, which is a broader and more diluted basket than food CPI and therefore a weaker detector of maize-price transmission.17

Key findings:

• African evidence points to only partial pass-through into headline inflation—typically about 5–15%—with stronger transmission into food-specific price indices than into aggregate CPI.13
• In Ghana, the monthly sample spans 337 observations from January 1998 to January 2026 for food CPI, and the exchange rate helps predict food CPI changes at the 5% level ($p=0.048$), while global maize prices do not.1213
• In Ghana, the strongest simple association is between the exchange rate and food CPI, with peak cross-correlation around 0.98, versus only about 0.33 between global maize prices and food CPI.
• In Nigeria, the aligned monthly multivariate sample is 96 observations from 2016 to 2024, and monthly changes show near-zero correlations: 0.015 for maize vs NGN/USD, -0.095 for NGN/USD vs headline CPI, and -0.009 for maize vs headline CPI.111415
• In Kenya, no national maize-price or food-specific CPI series was confirmed, so the analysis uses headline CPI only; Granger-causality p-values are 0.42 for maize → CPI and 0.10 for exchange rate → CPI, which is not strong enough to support a short-run transmission claim.11
• In South Africa, the overlapping monthly exchange-rate sample is only 26 observations, below the depth needed for a reliable VAR or impulse-response estimate, so no decision-grade dynamic pass-through estimate is reported.45
• The clearest transmission corridor in this set is Ghana’s exchange-rate-to-food-price channel, with the relevant horizon in the short run—within roughly 12 months—while the direct world-maize-to-domestic-price corridor is weak across the four-country evidence assembled here.1213

1. Theoretical Framework

The literature gives a clear transmission map. A global maize shock starts as a US-dollar price movement. It then passes through the local currency via the exchange rate, and only then reaches domestic wholesale, retail, and consumer prices. In import-dependent African economies, that pass-through is significant but incomplete because governments intervene, traders hold inventories, transport and processing costs matter, and not every food item in the consumer basket is directly tied to imported grain.123

Two distinctions matter for interpretation.

First, world-price shocks and exchange-rate shocks are not the same shock. A 10% increase in the dollar maize price and a 10% depreciation of the local currency both raise the local-currency landed cost of maize, but they operate through different policy levers. Governments cannot control the Chicago or global benchmark price, but they can influence the exchange-rate environment, border policy, reserve management, and domestic competition conditions.24

Second, food CPI and headline CPI are not interchangeable endpoints. Food CPI is designed to capture food-price movements directly; headline CPI spreads the same shock across a much broader basket that includes housing, transport, clothing, and services. The literature therefore finds stronger pass-through into food-specific indices than into headline inflation.13 That matters especially here: Ghana is the only country in this set for which a monthly food CPI series was available, while Nigeria, Kenya, and South Africa are assessed mainly through headline CPI.1314

The literature also gives a time profile. Pass-through is usually not immediate. It tends to emerge over 3 to 12 months, depending on import dependence, inventory turnover, milling and retail concentration, and whether governments use tariffs, subsidies, or export controls to buffer consumers.134 In southern Africa, maize is a politically sensitive staple, and policy efforts to stabilize local prices can delay or distort transmission rather than eliminate it.568

What this means: A country can be highly exposed to global maize shocks without showing a strong maize-to-headline-CPI relationship in the monthly data. The shock may be absorbed by the exchange rate, offset by policy, delayed in supply chains, or visible only in food-specific prices rather than in the full CPI basket.134

2. Historical Context: How the Transmission Corridor Works Over Time

The literature and the country evidence point to three broad phases in how global food shocks typically move through African economies.

2.1 Phase One: External commodity shock

The global benchmark itself is volatile. As Figure 5 shows, the monthly global maize benchmark runs from January 1992 to May 2026, providing a long external price anchor for all four country comparisons.11

Figure 5. Global Maize Benchmark Price (USD per metric ton)

What this means: The world price supplies the common external shock, but it is only the first step. A shared global benchmark does not imply a shared domestic inflation outcome.11

2.2 Phase Two: Currency translation into local costs

The exchange rate is the key conversion mechanism. A dollar maize shock becomes a local-currency import-cost shock only after the exchange rate is applied. The literature for Sub-Saharan Africa consistently emphasizes exchange-rate pass-through as a central amplifier or dampener of imported food inflation.234

2.3 Phase Three: Domestic price formation

Once the shock reaches the domestic economy, local factors take over: food-basket weights, milling concentration, transport costs, storage, trade restrictions, and subsidy regimes. South African evidence, for example, stresses that food-market structure can materially affect the speed and extent of pass-through.910 That is why the same world-price movement can produce a visible food-price response in one country and almost no detectable headline-CPI response in another.14

What this means: The transmission corridor is best understood as a chain, not a single coefficient. World price, exchange rate, and domestic price formation each matter, and weakness at any one link can break the measured pass-through.

3. Empirical Results by Country

3.1 Comparative summary

Table 1 brings the country results together. The key distinction is between Ghana’s food CPI endpoint and the headline CPI endpoints used elsewhere.

Before Table 1, note the scope point that matters most for interpretation: Ghana is assessed against food CPI, while Nigeria, Kenya, and South Africa are assessed mainly against headline CPI because no confirmed monthly national maize-price or food-specific series was established for those countries in the assembled evidence.1314

Table 1. Transmission corridor summary by country

Source: 1112131415

What this means: The clearest corridor is not “global maize directly raises CPI everywhere.” It is “exchange-rate movements matter most where the domestic endpoint is food-specific and the sample is deep enough to detect it.” Ghana fits that description; the others do not.

3.2 Ghana: the clearest exchange-rate corridor

Drawing on the Ghana Statistical Service’s monthly exchange-rate series and monthly food CPI, together with the global maize benchmark, the Ghana sample aligns to 337 monthly observations from January 1998 to January 2026.1213 This is the deepest and most decision-useful country panel in the set.

As Figure 1 shows, when Ghana’s food CPI and the global maize benchmark are both indexed to 100 at their first common month, domestic food prices climb far faster and more persistently than the dollar maize benchmark — the gap that the exchange-rate channel, discussed below, helps explain.1213

Figure 1. Ghana: Global Maize Price vs Domestic Food CPI (indexed, 100 = first common month)

The simple association supports that visual read. The peak cross-correlation between the exchange rate and food CPI is about 0.98, versus only about 0.33 between global maize prices and food CPI. In plain terms, food prices in Ghana move much more closely with cedi depreciation than with the dollar maize benchmark itself.

The dynamic result points the same way. The Granger-causality specification is a reduced-form monthly system on stationary transformations with four lags, of the form:

$\begin{aligned} \Delta P^{food}_t &= \alpha + \sum_{i=1}^{4}\beta_i \Delta P^{food}_{t-i} + \sum_{i=1}^{4}\gamma_i \Delta FX_{t-i} + \sum_{i=1}^{4}\delta_i \Delta M_{t-i} + \varepsilon_t \end{aligned}$

where $\Delta P^{food}_t$ is the monthly change in Ghana food CPI, $\Delta FX_t$ is the monthly change in the GHS/USD exchange rate, and $\Delta M_t$ is the monthly change in the global maize benchmark. Using that specification, the exchange rate Granger-causes food CPI at the 5% level (F = 1.80, p = 0.048), while global maize prices do not.1213 Granger-causality here means predictive precedence: past exchange-rate moves contain information about future food-CPI moves. It does not mean the exchange rate is the sole structural cause.

The impulse-response evidence is more restrained. In the differenced VAR(4), shocks from either global maize prices or the exchange rate produce only very small cumulative food-CPI responses over a 12-month horizon. That means the predictive link exists, but the measured average short-run effect in this system is economically modest.

Table 2. Ghana monthly transmission results

Source: 111213

What this means: Ghana shows the clearest transmission corridor in this four-country comparison, but it is an exchange-rate-to-food-price corridor, not a strong world-maize-to-food-price corridor. For policy, that is a crucial distinction: cedi stability matters more for near-term food inflation than the global maize benchmark alone.1213

3.3 Nigeria: no robust maize-to-headline-CPI corridor

Nigeria is assessed using headline CPI, not food CPI.14 The monthly series align over 96 observations from March 2016 to February 2024 for the exchange-rate window used in the system.1415

As Figure 2 shows, when Nigeria’s headline CPI and the global maize benchmark are both indexed to 100 at their first common month, headline CPI rises far more steeply than the dollar maize benchmark — but, as the monthly-change analysis below makes clear, that shared upward drift is not the same as transmission.1415

Figure 2. Nigeria: Global Maize Price vs Domestic Headline CPI (indexed, 100 = first common month)

The monthly-change correlations are effectively zero: 0.015 for maize vs NGN/USD, -0.095 for NGN/USD vs headline CPI, and -0.009 for maize vs headline CPI.111415 In practical terms, month-to-month moves in these series do not line up in a way that supports a clean transmission claim.

A cointegration test suggested one long-run relationship, but the underlying series properties weaken that result materially, so it should not be used as a decision-grade estimate. No robust pass-through coefficient is therefore reported for Nigeria.

Table 3. Nigeria monthly transmission results

Source: 111415

What this means: Nigeria’s inflation story is clearly severe, but this evidence does not support the narrower claim that global maize shocks are passing through cleanly into headline CPI via the naira over the monthly sample used here.1415 That is an important analytical boundary.

3.4 Kenya: weak evidence into headline CPI

Kenya is also assessed using headline CPI, because no national maize-price or food-specific CPI series was confirmed in the evidence assembled for this report. The monthly CPI series is long, but the relevant multivariate result remains weak at the headline level.

As Figure 3 shows, when Kenya’s headline CPI and the global maize benchmark are both indexed to 100 at their first common month, the two series follow visibly different paths, with no tight common pattern that would imply a strong transmission corridor.

Figure 3. Kenya: Global Maize Price vs Domestic Headline CPI (indexed, 100 = first common month)

The Granger-causality results are not strong enough to support a short-run pass-through claim: p = 0.42 for global maize → CPI and p = 0.10 for exchange rate → CPI. The impulse response is also small: a one-standard-deviation maize shock produces a peak response of about +0.0019 in the first month and a 12-month cumulative response of about -0.0026 index points.

The reduced-form specification is:

$\begin{aligned} \Delta CPI_t &= \alpha + \sum_{i=1}^{3}\beta_i \Delta CPI_{t-i} + \sum_{i=1}^{3}\gamma_i \Delta FX_{t-i} + \sum_{i=1}^{3}\delta_i \Delta M_{t-i} + \varepsilon_t \end{aligned}$

with three lags selected by AIC. The system was stable, but the substantive result is still weak: there is no robust evidence that either the global maize benchmark or the KES/USD exchange rate predicts headline CPI movements over the monthly horizon used here.

What this means: Kenya may well experience maize-price pressure in specific food markets, but this evidence does not show a strong pass-through into the broad consumer basket. Headline CPI is too diluted an endpoint to reveal a clear maize corridor in this case.

3.5 South Africa: descriptive evidence and literature, not a full dynamic estimate

South Africa’s overlapping monthly exchange-rate sample is only 26 observations, which is too short for a reliable VAR or impulse-response estimate. That is why no decision-grade dynamic pass-through coefficient is reported here.

As Figure 4 shows, when South Africa’s headline CPI and the global maize benchmark are both indexed to 100 at their first common month, headline CPI rises more gradually than in the other three economies; the series are useful descriptively, but the overlapping monthly exchange-rate sample is not deep enough jointly for a full monthly system estimate.

Figure 4. South Africa: Global Maize Price vs Domestic Headline CPI (indexed, 100 = first common month)

The literature is more informative than the short sample. South African work emphasizes that exchange-rate pass-through to food prices exists but is heterogeneous, and that maize and food-price effects are often more visible in food categories than in headline CPI.459 BFAP and competition-policy evidence also stress that trade controls, market concentration, and value-chain frictions shape the speed and extent of pass-through.5910

What this means: South Africa almost certainly has a maize-price transmission story, but the right place to look is food prices and local maize markets, not a short monthly headline-CPI system. The current evidence supports a qualitative conclusion, not a precise dynamic estimate.459

4. Cross-Country Interpretation: Which Corridor Is Clearest, and Over What Horizon?

The cross-country comparison is less about ranking “sensitivity” mechanically and more about identifying where the transmission chain is actually visible in the available data.

Table 4. Cross-country assessment of the transmission corridor

Source: 1112131415

What this means: Ghana stands out because the domestic endpoint is the right one—food CPI—and the monthly history is long enough to detect an exchange-rate channel.1213 Nigeria and Kenya are assessed through headline CPI, which is structurally less sensitive to maize shocks.13 South Africa likely has meaningful food-market transmission, but the monthly overlap available here is too short to quantify it reliably.

The horizon is also clearer in Ghana than elsewhere. The literature expects pass-through over 3 to 12 months.134 Ghana’s predictive exchange-rate link fits that short-run window, while the impulse responses suggest that the average effect is not large even when it is statistically detectable. For the other three countries, no comparable horizon can be stated with confidence from the evidence assembled here.

Data Sources and Methodology

This report combines institutional literature with monthly time-series evidence. The theoretical framework draws on World Bank, IMF, Banque de France, BFAP, and South African competition and policy work on commodity-price transmission, exchange-rate pass-through, and food-market structure.12345910

The empirical country work uses:

• the global maize benchmark price in USD per metric ton for January 1992 to May 2026;11
• Ghana’s exchange-rate series from the Ghana Statistical Service and Ghana’s food CPI from the Ghana Statistical Service;1213
• Nigeria’s monthly headline CPI from the IMF and NGN/USD exchange-rate data from World Bank-linked exchange-rate sources;1415
• Kenya’s monthly headline CPI and monthly exchange-rate series used in the assembled evidence;
• South Africa’s monthly headline CPI and monthly ZAR/USD exchange-rate series used in the assembled evidence.

The econometric logic follows standard price-transmission practice in the literature: cointegration to test for a long-run relationship; Granger causality to test whether one variable helps forecast another; and vector autoregression (VAR) or impulse responses where the monthly sample is deep enough to support dynamic analysis.123 The generic system estimated is:

$\begin{aligned} \Delta P_t &= \alpha + \sum_{i=1}^{k}\beta_i \Delta P_{t-i} + \sum_{i=1}^{k}\gamma_i \Delta FX_{t-i} + \sum_{i=1}^{k}\delta_i \Delta M_{t-i} + \varepsilon_t \end{aligned}$

where $P_t$ is the domestic price index, $FX_t$ is the local-currency-per-US-dollar exchange rate, and $M_t$ is the global maize benchmark in USD. In plain language, the model asks whether past changes in the domestic price, the exchange rate, and the world maize price help explain the current month’s domestic price change.

Limitations

The main limitation is not conceptual but measurement-related: Ghana is the only country in this set with a confirmed monthly food CPI endpoint, while Nigeria, Kenya, and South Africa are assessed mainly through headline CPI, which is a broader and less sensitive measure of maize-price transmission.1314 South Africa’s overlapping monthly exchange-rate sample is also too short for a reliable dynamic system estimate, so the South African result is interpretive rather than coefficient-based.45

Policy Implications

For Government

Ghana: Focus food-inflation management on the exchange-rate corridor. The evidence shows the cedi/USD rate has predictive content for food CPI, while the direct global maize-price link is weak.1213 That argues for tighter FX liquidity management, clearer import-finance access for grain and milling firms, and faster customs clearance during external food shocks.

Nigeria and Kenya: Do not infer maize pass-through from headline CPI alone. The evidence does not support a clean monthly maize-to-headline-CPI corridor.111415 Governments should therefore monitor food-specific baskets—especially cereals and maize meal—rather than relying on aggregate CPI to gauge food-security pressure.

South Africa: Use food-market surveillance, not headline CPI alone, as the primary early-warning system. South African literature points to the importance of market concentration, trade controls, and value-chain frictions in shaping pass-through.5910 Competition and trade-policy tools matter as much as macro tools.

For Investors

Imported-food and consumer-staples exposure should be screened through the currency channel first. In this four-country set, the exchange rate is the more plausible transmission corridor than the global maize benchmark alone.24 Ghanaian consumer-facing firms are especially exposed to cedi weakness through food-input costs, even when the global maize benchmark is not moving one-for-one with domestic prices.1213

For Nigeria and Kenya, investors should avoid over-reading headline CPI as a direct measure of maize-price stress. A company with cereal exposure can face real margin pressure even when headline CPI does not show a strong statistical link.

For Development Partners

Support the publication of monthly food-specific and staple-specific price series. This report’s strongest result emerges precisely where the domestic endpoint is food CPI rather than headline CPI.13 Better food-price data would materially improve inflation surveillance, social-protection targeting, and shock-response design.

Development partners should also support integrated monitoring of world prices + exchange rates + domestic food baskets. That is the actual transmission chain identified in the literature and partly confirmed in Ghana.123 Monitoring only one link in that chain misses the policy-relevant signal.

References

1. International Commodity Prices Transmission to Consumer Prices in Africa [link]

2. International Commodity Prices Transmission to Consumer Prices in Africa [link]

3. International Commodity Prices Transmission to Consumer Prices in Africa | Publications [link]

4. Exchange rate pass-through to food prices in South Africa [link]

5. Impact of trade controls on price transmission between southern African maize markets — Bureau for Food and Agricultural Policy (BFAP) [link]

6. Food Inflation brief September 2022 — Bureau for Food and Agricultural Policy (BFAP) [link]

7. Full article: Measuring the pass-through effect of global food price volatility and South Africa’s CPI on the headline inflation of Zimbabwe [link]

8. Food Inflation Brief February 2022 — Bureau for Food and Agricultural Policy (BFAP) [link]

9. EFPM Report Sep 2023 — Competition Commission South Africa [link]

10. EFPM Report Aug 2022 — Competition Commission South Africa [link]

11. Global Maize Benchmark Price — International Monetary Fund [link]

12. Exchange Rates — Ghana Statistical Service [link]

13. Consumer Price Index (CPI) and Inflation — Ghana Statistical Service [link]

14. IMF IFS — Consumer Price Index, Nigeria (monthly) — International Monetary Fund [link]

15. Exchange Rates (IMF/fawazahmed0) (NGN_USD) — World Bank [link]