A week-by-week look at four Saturdays in June and July shows the same story every time: Kenya’s currency market is shut, and the chart still finds a number to plot. That number is not the shilling. Instead, it appears to be a synthetic value inserted by an upstream data feed, one that repeatedly creates the illusion of a weekend rally in a market that never traded. Compare those Saturdays against independent historical exchange-rate records and the pattern vanishes: Friday, Saturday and Monday sit in the same narrow range, with no evidence of the sharp swings shown on Google Finance. The episode is a reminder that financial charts are not direct windows into markets but products of decisions made deep inside data pipelines. When exchanges or over-the-counter markets close, data providers must decide whether to carry forward the last quoted price, interpolate a value or rely on thinner secondary inputs. Most of the time those choices pass unnoticed. But in Kenya’s relatively illiquid, bank-driven foreign exchange market, they can produce a phantom rally convincing enough to mislead anyone relying on a quick glance at a chart rather than the market itself.
Look closely enough at the USD/KES chart on Google Finance in the past month and a rhythm emerges. Roughly once a week, the line drops, then recovers within a day or two, as if the shilling gathers strength every Saturday before giving it back. Chart the same window against a second, independently sourced rate feed, and the rhythm disappears. The Saturday numbers sit quietly inside the same narrow band as the Friday before and the Monday after. Nothing moved. The market most likely never opened.
This is not a story about the shilling. It is a story about what happens when a data provider has to fill in a day that the underlying market has left blank, and about how easily that gap-filling can be mistaken, at a glance, for a trend.
The market that goes dark
Kenya’s shilling is not fixed by decree. It floats, priced hour to hour by the commercial banks and licensed forex bureaus that make a market in it.
The Central Bank of Kenya says as much on its own site: it does not set the exchange rate, leaving that to competition among banks and bureaus, and simply compiling the resulting indicative figure once trading has happened.
That compilation happens on business days only. The bank’s own published archive of indicative rates has no entries for Saturdays or Sundays at all, a gap that repeats every single week of the year, not just the five examined here.
Five Saturdays, tested against the record
To see whether the shilling really was moving on its own schedule, each of the five Saturdays inside the past month’s chart window was checked against an independent historical-rate archive, one that logs a daily reference close for every calendar day, weekends included, and against the Friday and Monday closes on either side. If Saturdays were genuinely different, the reference feed’s own weekend numbers should show it too.
They do not. Each chart below plots that Saturday’s real, sourced close next to its Friday and Monday neighbors, alongside the value the Google Finance chart appeared to show for the same date.
Why the two sources disagree
Google Finance does not price currencies itself. Its USD/KES feed is licensed from Morningstar, which, like any aggregator, has to produce a number for every calendar day regardless of whether the underlying market traded. The independent archive checked here appears to carry forward the last available close over the weekend, which is why its Saturday figures sit so close to Friday’s. Morningstar’s feed, for whatever upstream reason, does something different on at least some Saturdays, producing a reading that swings several shillings away from the rest of the week before correcting itself by Monday.
That description was written about global currency data generally, not Kenya specifically, but it fits. When a market has no real trading to report, any feed that still needs to plot a line is choosing, deliberately or not, between two honest options: hold the last known price steady, or fall back to some thinner secondary source and accept whatever noise comes with it. The reference archive checked here appears to take the first path. Google’s feed, on at least four of the five Saturdays examined, behaves as though it is taking the second.
Brokers who do offer genuine weekend trading warn their own clients about exactly this kind of distortion, just applied to the ordinary weekly open rather than the weekend itself. Forex.com cautions traders about the potential for illiquid market conditions particularly at the open of the trading week, since thin order books let prices swing on very little volume. Kenya’s shilling market does not even have a thin order book on Saturdays. It has no book at all.
Why this looks new
Nothing in how Nairobi’s banks or bureaus set the shilling has changed. What plausibly shifted is upstream, in whatever secondary or synthetic source Morningstar’s feed leans on when Nairobi’s own market is closed. A few shillings of weekly wobble, repeating because the same mechanism fires every weekend, reads as a rhythm once it is compressed into a single monthly chart. Zoomed out far enough, noise starts to look like a signal.
The comparison across five separate Saturdays makes the point more firmly than any single week could. If the shilling were actually gaining every Saturday, an independent, weekend-inclusive rate archive should show the same gains. It does not, on four of the five dates checked here, which points toward a feed-specific quirk rather than a market-wide event.
What a phantom rally costs
A chart artifact sounds like a harmless curiosity until someone acts on it. Kenyans abroad timing a remittance around what looks like a weekly dip in the dollar, small importers who glance at a phone before confirming a supplier invoice, or a junior reporter on a Sunday news desk pulling a quick currency line for a business brief, all have reason to trust a number that Google presents with the same clean confidence whether the underlying market traded or not. None of them are told, anywhere on the page, that Saturday’s figure may not reflect a single actual transaction in Nairobi.
The risk compounds because the shilling’s real weekday moves are themselves modest. Over the same stretch covered here, the currency traded in a band of roughly one and a half shillings to the dollar on days when Kenyan banks were actually quoting. A weekend swing several times that size, even if it reverses by Monday morning, is large enough to look like genuine news rather than a rounding artifact, especially to anyone who checks a rate once and does not come back to watch it correct itself.
Method and its limits
The comparison here is deliberately narrow and says so. The reference figures for each Friday, Saturday, and Monday come from a historical-rate table that logs a daily reference close for every calendar day of the year, cross-checked across two related archives that publish the same underlying dataset. Those numbers are exact, dated, and independently retrievable. The Google Finance figures are not: Google’s interface does not offer a downloable export of its daily series, so the value attributed to each Saturday here was read visually off the one-month chart the person supplied, a method that carries real margin for error and is labeled as such on every chart in this piece.
That asymmetry is itself part of the story. A reader who wants to check a currency chart against a second opinion should be able to pull an exact number from more than one place. For Kenya’s shilling on a Saturday, one of the two major sources checked here makes that easy, and the other, the one most people actually open on their phones, does not make it possible at all.
The takeaway for anyone reading the chart
For a currency with no formal weekend session, the safest habit is to check whether a given source is quoting an actual functioning market on the day being read, and to treat any single-provider weekend print with real skepticism until a second source confirms it. For the shilling, on the Saturdays examined here, the honest answer is that Kenya’s market was closed, and the chart still drew a line.
None of this required a conspiracy, a central bank decision, or a shift in investor sentiment toward East Africa. It required one aggregator filling a gap the market itself left empty, five weeks in a row, in a way that a five-minute cross-check against a second source could catch every time.
Faustine Ngila is the AI Editor at Impact Newswire, based in Nairobi, Kenya. He is an award-winning journalist specializing in artificial intelligence, blockchain, and emerging technologies.
He previously worked as a global technology reporter at Quartz in New York and Digital Frontier in London, where he covered innovation, startups, and the global digital economy.
With years of experience reporting on cutting-edge technologies, Faustine focuses on AI developments, industry trends, and the impact of technology on society.
Khadija Sharife (South Africa/International)
Most Impactful Work: Her investigations for the Organized Crime and Corruption Reporting Project (OCCRP) and contributions to books like “Tax Us If You Can: Why Africa Should Stand Up for Tax Justice.”
Short Bio: An award-winning investigative journalist and senior editor for Africa at OCCRP, Khadija Sharife specializes in illicit financial flows, political economy, and environmental predation. Her work often uncovers complex international corruption networks.
Dangle
Laduma! – he scores! [laa-doo-maa] This term is usually screamed out when a soccer team scores a goal. Extend any of the syllables for full effect.Example: “Laaaaaaduuuuumaaaa!”
Murder Matter
Fun South African Fact: The waters surrounding the 2,500km long coastline are teeming with life. You’ll find great white sharks, African penguins, dolphins and much much more in South Africa’s oceans. Many travellers visit the coast to watch the annual whale migrations. In Hermanus, in the Western Cape, you can see hundreds of whales from June to late November. Simon’s Town, another coastal Western Cape town, is home to Boulder’s Beach. Here you can swim alongside colonies of African Penguins!