The Parallel Rate Was the Honest Rate: What the Black Market Revealed

The premium as a confession

If you wanted to know what was actually happening in the Ethiopian economy at any moment from approximately 1977 to 2024, you did not look at the official exchange rate posted by the National Bank of Ethiopia. You looked at what dollars were trading for in the back rooms of Mercato, on the curbside in Bole, in the offices of traders moving khat and gold across the borders into Djibouti and Kenya. The number on the wall of the NBE was a political artefact. The number whispered in the back of the shop was the price.

This article argues that the parallel exchange rate — what nearly everyone called the “black market” rate — is the single most important data series in modern Ethiopian monetary history, more revealing than any official statistic, because it is the one number no Ethiopian government could control. The official rate was a policy declaration. The parallel rate was a price. The gap between them, the parallel premium, was a continuous, unfalsifiable measure of the distance between what the regime claimed and what was true. To follow the disease this series is tracking, you follow the premium.

The argument has a corollary that may sound provocative but is, on the evidence, correct: the participants in the parallel market — the diaspora sending remittances, the trader holding dollars, the foreign-exchange dealer in Mercato, the Eritrean and Somali hawala networks moving currency across borders — were not criminals defrauding the state. They were the system’s accountants, providing the price-discovery service the state had refused to provide and that the economy could not function without.

What the premium measured

The textbook definition of a parallel exchange rate is the rate at which currency is traded outside official channels, in violation of regulation. The textbook definition of the parallel premium is the percentage by which this rate exceeds the official rate. The Ethiopian parallel premium, across the entire period this article covers, was the percentage by which the open market valued dollars more highly than the NBE was willing to admit.

What does this premium measure, in economic terms? Several things at once, in any country with a fixed or managed exchange rate. It measures excess demand for foreign currency at the official rate — the gap between the dollars people want to buy at the official price and the dollars actually available at that price. It measures expected future devaluation — the discount the market is applying to the official rate in anticipation of an eventual adjustment. It measures the cost of the controls themselves — the deadweight loss from rationing, paperwork, and capital controls. And, in countries with significant monetary financing of deficits, it measures the implicit inflation that the official statistics are not reporting. All four of these were features of the Ethiopian situation across the entire period.

The premium widens when any of these factors worsens. A government that increases its deficit and finances it with central-bank advances increases the money supply, which raises the underlying inflation, which means the official exchange rate is more overvalued in real terms, which raises the parallel rate. A government that imposes new capital controls makes the official channels harder to use, which raises demand for the parallel channels, which raises the parallel rate. A government that escalates a war or absorbs a sanctions hit loses access to official foreign-exchange inflows, which tightens the rationing on the official side and pushes more demand to the parallel side, which raises the parallel rate. The premium is, in this sense, a diagnostic dashboard: when it widens, something in the monetary or fiscal stance has gotten worse, and the data tell you to go look at the underlying.

The history in numbers

The Derg-era parallel premium is difficult to document precisely because the parallel market was, by definition, not officially reported. The reconstructions available rely on Mercato dealer surveys, on the records of foreign embassies whose staff used parallel markets to manage their own currency, and on the Bank of Eritrea’s records of the cross-border trade with Sudan and other neighbours. The composite picture, assembled by Alemayehu Geda and others, is that the Derg-era premium fluctuated between roughly 100 percent and 250 percent across the 1980s.¹ That is to say: at the height of the Derg’s monetary disorder, a dollar that the NBE officially valued at 2.07 birr could be bought on Mercato for somewhere between 4 and 8 birr.

This number contained, in real time, the information the official statistics were suppressing. The official inflation rate in 1985 was reported by the Central Statistical Authority as 16.5 percent.² What the parallel premium was saying, by widening sharply in the same period, was that the true rate of monetary erosion was significantly higher — that the official inflation number was being held down by price controls that were not actually clearing markets. The true inflation rate, by any honest measure that used the prices Ethiopians actually paid in the open market for the goods they actually bought, was probably between 25 and 35 percent in the worst Derg years.³ The parallel premium was the data point that revealed this.

The EPRDF period after 1991 reduced but did not eliminate the premium. The 1992 devaluation (article 9) closed the gap substantially in the immediate aftermath; the premium widened again through the 1990s and 2000s as monetary policy remained loose and the official rate was maintained through periodic discrete devaluations rather than continuous adjustment. By the early 2000s, the premium fluctuated in the 10–25 percent range — better than the Derg years, far worse than what a healthy regime would tolerate.⁴

The premium widened sharply in the GTP era (2010–2020) and then catastrophically in the period 2020–24. By the time of the July 2024 float, the parallel rate had reached roughly 120 birr per dollar against an official rate of approximately 57, a premium of over 100 percent.⁵ This is the premium the float was designed to close, and largely did: by mid-2026, the NBE-published indicative rate was approximately 157 birr per dollar, with a parallel premium reported by the IMF and Reporter at approximately 12 percent.⁶

The participants

The standard framing of “black markets” in Ethiopian official rhetoric was that they were the work of criminals — speculators, smugglers, foreign-currency hoarders motivated by greed and indifference to national interest. The framing was politically convenient because it located the problem in the participants rather than in the policy that created the participants’ opportunity. The framing was also, on examination, largely false.

The principal participants in the Ethiopian parallel foreign-exchange market across the period were the following groups, each acting on entirely defensible motivations.

The diaspora — Ethiopians abroad sending remittances home — increasingly preferred parallel channels because the parallel rate was substantially better than the official rate. A diaspora worker in Saudi Arabia or the United States sending $500 to family in Addis would, at the official rate, deliver perhaps 30,000 birr in 2022. Through a hawala intermediary using the parallel rate, the same $500 would deliver more than twice that. The diaspora chose, rationally, to maximise the support delivered to their families. The choice was not anti-Ethiopian; it was pro-family.

Traders and importers needing dollars to pay foreign suppliers found that the official allocation system was slow, opaque, and inadequate. Documented waits for letters of credit in the 2020–24 period ran from six to eighteen months.⁷ The parallel market provided same-day execution. Traders who used the parallel market were not undermining the economy; they were keeping their businesses operational against an administrative system that was, by its own admission, unable to clear demand at the official rate.

Exporters — particularly coffee, sesame, and gold exporters — faced surrender requirements that forced them to sell a percentage of their foreign-currency earnings to the NBE at the official rate, which was below the parallel rate. Predictably, some portion of export earnings was diverted to parallel channels, by under-invoicing exports, holding earnings in foreign accounts, or smuggling. The Ethiopian government estimated that gold smuggling alone may have cost the country billions of dollars in lost export earnings over the GTP years.⁸ The exporters were not being unpatriotic. They were being taxed by surrender requirements that exceeded what an honest market would have imposed.

Hawala networks, operating largely through Somali and Eritrean diaspora communities with deep cross-border presences, provided the actual infrastructure of the parallel market. They moved money across the borders to Djibouti, Mogadishu, Hargeisa, Sanaa, and back. They settled accounts informally and reliably. They charged margins, but the margins were lower than the official banking system’s effective cost when waiting times and rationing were accounted for. The NBE has periodically denounced the hawala networks as money-laundering vehicles, and there is no doubt that hawala can be used for that purpose. But the bulk of hawala traffic in the Ethiopian case was ordinary remittance and trade settlement, conducted at the parallel rate because the official rate was a fiction.⁹

What the regime would not see

There is a deep Sowellian point worth making explicit here, because it goes to the heart of why every Ethiopian regime has misdiagnosed its parallel market.

The official explanation of the parallel market, in regime after regime, has been that it is a speculative attack on the currency — that traders are pushing the parallel rate higher in order to profit from the spread, that the rate does not reflect underlying conditions but rather the manipulations of market participants. The diagnosis points to a solution: enforcement. Crack down on the parallel traders, arrest the dealers, freeze their accounts, criminalise the practice. Each Ethiopian regime, in its turn, has pursued some version of this enforcement strategy. Each has failed.

Why the strategy fails is Sowell’s point. The premium is not generated by the traders; the premium is generated by the disequilibrium in the official market. The traders are responding to the disequilibrium, not creating it. If you arrest the traders, the disequilibrium is still there — the demand for dollars at the official rate still exceeds supply — and new traders will appear to meet the demand, because the spread between official and parallel rates makes it profitable for them to take the risk of being arrested. The only ways to actually close the parallel market are to eliminate the disequilibrium: either by devaluing the official rate to the parallel level (which is what the 2024 float did), or by sufficiently tightening monetary and fiscal policy that the parallel rate falls to the official level (which would require the genuine reform that no Ethiopian regime has implemented).

Enforcement, in the absence of one of these two underlying fixes, treats the symptom and not the disease. The Derg’s enforcement, the EPRDF’s enforcement, and the post-2018 government’s enforcement all failed for the same reason: they were attacking the messengers carrying the news that the policy was wrong. As long as the policy was wrong, the news kept arriving.

The parallel rate as honest accounting

If this article has a single point, it is that the parallel exchange rate was not a problem to be solved by enforcement. It was an information system, generated spontaneously by the market in response to the absence of accurate official prices, that performed precisely the function Hayek’s framework predicts the price system performs: it transmitted dispersed knowledge about supply and demand to anyone willing to look.

Diaspora households reading the parallel rate in their WhatsApp groups knew, more accurately than the World Bank’s macroeconomic models, how much purchasing power their remittance would deliver. Traders watching the parallel rate move knew, more accurately than the NBE’s forex rationing committee, what dollars were actually worth that week. Foreign investors watching the spread between official and parallel rates knew, more accurately than the Ministry of Finance’s debt-sustainability analyses, what the real-exchange-rate adjustment would have to be eventually. The parallel rate was the honest rate. The official rate was the political rate. Every Ethiopian regime, while it lasted, tried to suppress the first and protect the second. Every regime failed, because the first kept transmitting what the second was hiding.

The implication, which the series will return to in articles 19 and 27, is that the right monetary regime for Ethiopia is not one that suppresses the parallel market through enforcement. It is one that eliminates the parallel market by making the official rate equal to the parallel rate, and then keeps it equal by maintaining the monetary and fiscal discipline that would prevent a new gap from opening. The 2024 float, whatever its defects, was the first serious attempt at the first half of this. The second half — keeping the gap closed — is the question of whether the post-2018 regime is structurally different from its predecessors, or only momentarily so. Articles 21 and 22 examine that question. The next article turns to the EPRDF’s 1991–92 transition and the partial reform that failed to close this gap the last time it was attempted.

---

Footnotes

1. Alemayehu Geda, “The Macroeconomic Performance of the Ethiopian Economy,” in Befekadu Degefe and Berhanu Nega, eds., Annual Report on the Ethiopian Economy, Vol. 1 1999/2000 (Ethiopian Economic Association, 2000); see also International Monetary Fund, Ethiopia: Recent Economic Developments, IMF Country Report 96/130 (December 1996). ↩

2. Central Statistical Authority of Ethiopia, Consumer Price Indices for Selected Towns, various issues 1980–1991, summarised in IMF, Ethiopia: Recent Economic Developments, 1996. ↩

3. Alemayehu Geda, “The Macroeconomic Performance,” 84–89. ↩

4. Alemayehu Geda and Befekadu Degefe, “Explaining African Economic Growth Performance: The Case of Ethiopia,” AERC Research Paper, 2002. ↩

5. PWC Kenya, “Ethiopian Birr Devaluation,” August 2024; “Ethiopia’s IMF backed reforms prompts currency devaluation by 30%,” Eastern Africa Association, August 2024. ↩

6. National Bank of Ethiopia, “Indicative Daily Exchange Rates,” https://nbe.gov.et/exchange/; Capital Market Ethiopia, “National Bank of Ethiopia Exchange Rate Today,” April 2026, citing the indicative rate of 156.83 birr/USD as of 20 April 2026, with parallel premium narrowed to approximately 12 percent. ↩

7. International Monetary Fund, “Ethiopia: 2025 Article IV Consultation,” July 2025, IMF Country Report 25/188, on pre-reform forex backlogs. ↩

8. “Ethiopia targets Somali-owned remittance firms in U.S. over money laundering claims,” Hiiraan Online, August 2025. ↩

9. Tewodros Aragie Kebede, “Should Ethiopia legalise its informal currency exchange markets?”, ENACT Africa, https://enactafrica.org/enact-observer/should-ethiopia-legalise-its-informal-currency-exchange-markets. ↩