There is a particular kind of confidence a Finance ministry displays when it starts turning down free money. Not literally free, of course because the Bank of Uganda (BoU) still pays for every shilling it borrows, but when a government is offered nearly nine times more cash than it asked for and still says no to the majority of it, that is a signal worth reading carefully.
July 2026, the opening month of Uganda’s new financial year, offered four such signals in a row. To understand July, it helps to know what came before it. The 2025/2026 fiscal year (FY) was an election year, and Uganda’s bond market behaved the way election-year bond markets tend to: nervously, then expensively. Yields on the long end of the curve touched 17.7 to17.9 percent at auction and briefly cleared 18 percent on the secondary market. Investors were, in effect, being paid a hefty political-risk premium to hold 15, 20 and 25-year government paper through a contested electoral cycle.
A slow unwind
The Government of Uganda (GoU), meanwhile, was in no position to argue: its domestic financing needs were large (a Shs72 trillion budget, later topped up with a roughly Shs10 trillion supplementary), donor grants were declining, and it needed the market to show up. It did. Auctions were oversubscribed almost without exception, and the GoU accepted, on average, roughly 13 percent more than it had originally offered across the year. That appetite front-loaded itself.
Uganda issued heavily in the first half of the fiscal year and then, tellingly, skipped a bond auction altogether in December and pulled back acceptance sharply through January and March, as coupon payments and maturities came due and the Treasury managed its own cash-flow pressure rather than the market’s. Then came a wrinkle nobody had priced for. By January 2026, with the election behind it and roughly 90 percent of its financing needs already met, the government’s cost of borrowing began falling fast.
The 20-year touched 14.7 percent in January, a startling six-month drop from the 2025 peaks. Then the Iran war escalated in February and March, the dollar strengthened, offshore investors, who had built their Ugandan sovereign holdings back up to nearly 15 percent of the market by December, began heading for the exits, and yields snapped back up into the 15 to16 percent range, where they largely stayed through the first half of 2026. As Janet Anayo, an investment analyst at Old Mutual Investment Group Uganda, put it: ‘That Iran war escalated, in February, March… you’d see that there was a bit of volatility in the fixed income market. Part of that was that these offshore investors were exiting some of these positions.’
July, dissected
Against that backdrop, July’s four auctions that include two, five and 15-year bonds on the 1st; Treasury bills on the 8th; three, 10 and 20-year bonds on the 15th; and two, five, 15 and 25-year bonds again on the 29th, read more like a government pressing its advantage. The numbers, taken together, show that across the month, the BoU offered a combined Shs3.73 trillion of paper. Investors tendered Shs8.79 trillion, 2.36 times over-subscribed. The GoU accepted just Shs3.85 trillion, meaning it turned away roughly Shs4.93 trillion that investors were actively trying to hand it.
More interesting than the aggregate is the shape of the curve within the month. Three tenors were reopened twice in July, and all three priced lower the second time. The two-year fell from 12.800 percent (1 July) to 12.500 percent (29 July), down 30 basis points. The five-year fell from 14.700 percent to 14.250 percent, down 45 basis points, even as its tender book nearly doubled, from Shs824 billion to Shs1.06 trillion. The 15-year fell from 15.750 percent to 15.650 percent, down 10 basis points The short and belly of the curve, in other words, are still compressing, and doing so with demand accelerating rather than tapering off, a signature of a market conditioned to expect further easing.
The long end also tells something. The 25-year bond, the same 16 percent-coupon, 2050-maturity paper first floated in August 2025, cleared at exactly 16.000 percent on July 29, having traded, according to secondary-market levels in June, around 15.93 percent. That is a small but real reversal: seven basis points higher, not lower, after weeks of apparent softening. It is also barely above where the 20-year cleared two weeks earlier, at 15.950 percent, an unusually flat spread for 500 basis points of extra duration risk. Meanwhile the 10-year, at 15.450 percent, is essentially unchanged from where it was trading roughly six weeks earlier at 15.44 percent.
This means that Uganda’s front end is still easing decisively, its belly has stabilised, and its very long end has stopped falling and shown its first flicker of resistance. That divergence matters for how one reads the ‘taper.’ Government’s declared preference, as Susan Namaganda, a fixed income expert noted, is unambiguous. At borrowing costs of 16 to17 percent, ‘that’s really expensive debt for the government of Uganda as they’re running the country,’ and the state ‘wants to start tapering… maybe we can go down to the 14 percent ranges, 13 percent ranges.’ The July data shows that ambition is being realised at the short end and stalling, for now, at the long end, precisely where offshore investors, still skittish after February’s exodus, exert the most pricing power.
The architecture
None of this is accidental. The 2026/2027 auction calendar, published just before the fiscal year began, was engineered with exactly this dynamic in mind. Treasury bill auctions were cut from twice a month to once, a change Ms Namaganda linked directly to the State’s declining appetite for short-dated rollover risk. ‘Now for this particular calendar the government is saying, no, we no longer want that much of the short money, we want longer-term money,’ she noted. Several long-dated benchmarks are also being deliberately retired. The three-year bond used in July’s auction, which is the 15.550 percent paper maturing July 2028, is scheduled to stop being reissued around August, replaced by a fresh 3-year benchmark from September.
The 15-year (15.800 percent, maturing 2039) and 20-year (15.000 percent, maturing 2043) bonds sold in July face the same fate around August and September. Ms Namaganda’s explanation was that: ‘Since these particular bonds, especially the 20 and 15-year bond, ever since they issued them out, they have been appearing on the calendars non-stop… the government has also accumulated an outstanding debt on those particular [maturities].’ A bond switch is also scheduled for March 2027, giving holders of the August 2029 bond the option, not obligation, to roll into fresh paper, a tool Ms Namaganda described as designed to relieve exactly the kind of maturity-clustering pressure that a large single redemption date creates: ‘It is not a must to switch your bond.’
There is also a longer-term financing story sitting underneath all of this. Ms Anayo flagged Uganda’s approaching oil production, expected to commence in the second half of 2026, as a structural reason for the government to want cheaper domestic debt now. A new external revenue stream reduces the urgency to keep paying elevated coupons to attract Treasury demand. ‘The government wants to diversify away from expensive shilling-denominated borrowing to be able to reduce the rate at which they are borrowing,’ Ms Namaganda noted. That, combined with public debt sitting at roughly 53 percent of GDP and a Fitch rating of B+ with a stable outlook, figures which Ms Anayo cited as the credit backdrop investors should be weighing against the yields. This gives the government both the incentive and, so far, the credibility to keep rejecting bids rather than chase the market up.
What to watch
Owing to this, two things are worth tracking into August and beyond.
First, whether the roughly Shs900 billion rejected at the 25-year auction alone, plus the hundreds of billions turned away across the rest of July, resurfaces, as expected, at the mid-August auction covering the 20-year, 10-year and 3-year bonds.
A further oversubscription there would confirm that July’s discipline is sustainable rather than a one-month anomaly.
Second, whether the long end’s slight uptick on July 29 was noise or the start of a genuine re-steepening, as the market prices in the retirement of the current 15- and 20-year benchmarks and waits to see what coupon the replacement bonds carry in September.
If the pattern of the last financial year is any guide, heavy front-loaded issuance early, followed by a deliberate pullback once financing needs are largely met, investors would not assume July’s rejection rates persist unchanged all year.
One month in, the government has already cut yields on three bonds it didn’t have to reopen, and the long end hasn’t punished it for trying.
In July 2026, at least, Uganda’s Treasury was setting the terms, not taking them.