The Sixty-Day Peace

In one minute:

A time-limited deal is a clock. What matters is the chance the clock runs out into renewed closure.

Turning an annual risk into a 60-day risk takes two lines: λ = −ln(1 − p), then P = 1 − e^(−λt). An annual 30% becomes about 5.7% over 60 days; an annual 50% becomes about 10.8%.

In June, Brent slid from about $101 to about $75 while the deal's expiry was written into the text. The clock was the signal.

1 June: A Barrel Near $96

On the first day of June, a barrel of Brent was trading at about $96. Picture a matatu driver in Nairobi doing sums at the pump, or a clinic manager trying to keep a generator, a vehicle and a fridge running on one budget. The number on the screen was the surface. Underneath it were a strait, a ceasefire and a calendar.

What was visible

A disciplined forecaster reading the June sources would have seen three signals at once. In hindsight, all three were on the table before the price moved.

The first was inflation. US CPI rose 4.2% year on year in May, the highest since April 2023, with energy up 23.5% (CNBC). Central banks noticed. On 11 June the ECB raised its deposit facility rate to 2.25%, citing Middle East war inflation pressure (ECB).

The second was the oil premium itself. Brent printed about $101.36 on 3 June, the highest June reading the research found (Fortune). A price that high carries a fear, and the fear has to be priced somewhere.

The third was the deadline. On 14 June, the US and Iran agreed a framework extending their ceasefire by 60 days and reopening the Strait of Hormuz (Axios). Three days later, a memorandum signed by Trump and Iranian President Pezeshkian described safe passage for commercial vessels as lasting "60 days only" (Britannica). The clock was not hidden in the fine print. It was the agreement's own number.

What happened

Brent fell. By 30 June it was about $75.02, after readings of about $79.25 on 22 June and $73.74 on 26 June (Fortune). Most of these price readings come from daily market pages rated medium confidence, so treat them as snapshots, not settlements.

Kenya felt the same wave at the pump, with a cushion. EPRA's June cycle cut Nairobi super petrol by KSh0.22 to KSh214.03 a litre and diesel by KSh10 to KSh222.86, with a record diesel subsidy of KSh34.07 a litre cited (Kenyans.co.ke). The cycle before had pushed diesel to a record KSh242.92 (Tech-ish). Meanwhile the Central Bank of Kenya held its rate at 8.75% with May inflation at 6.7% (CBK).

Table sources: Fortune, CNBC, NOAA, Axios, Britannica, Fortune.

The math

A hazard rate is the chance per unit of time that something ends. It converts a long-run risk into a short-run one, and the conversion is not a straight line. Three steps:

Step 1. Turn the annual probability p into a constant rate: λ = −ln(1 − p). Step 2. Express the window as a fraction of a year: t = 60 / 365 ≈ 0.164. Step 3. Compute the chance of breakdown inside the window: P = 1 − e^(−λt).

Case A, annual risk 30%: λ = −ln(0.70) ≈ 0.3567 λt ≈ 0.3567 × 0.164 ≈ 0.0586 P = 1 − e^(−0.0586) ≈ 0.057, or about 5.7%.

Case B, annual risk 50%: λ = −ln(0.50) ≈ 0.6931 λt ≈ 0.6931 × 0.164 ≈ 0.1140 P = 1 − e^(−0.1140) ≈ 0.108, or about 10.8%.

A straight-line shortcut would give 4.9% and 8.2%. The gap widens as the annual risk rises, because the exponential form compounds the chance of an ending over time.

Two lessons follow. The first is that a 60-day window is a slice of the year, and the slice can be computed. Counted from 14 and 17 June, the clock ran out in mid-August. The second is that the constant-rate assumption is itself a judgment. The 30% and 50% cases are illustrations, not estimates drawn from the sources. We have criticised a rival report for assuming constant annual risk with nothing to compare against. The same caution applies here. A forecaster's job is to pick the annual number honestly, write it down, and update it when the news changes.

The Pawanax lens

For a Kenyan household, the pump is the transmission belt. A litre of diesel at KSh222.86, even after a subsidy of KSh34.07 a litre, is a cost someone is carrying. For a clinic, that cost lands on referral transport, generator fuel and the movement of stock. The useful question is not "is there peace?" but "how many weeks of budget does this clinic have if the price returns?"

The health system faced its own clock. The WHO's Bundibugyo Ebola outbreak had 1,155 confirmed cases and 304 confirmed deaths in DRC by 26 June, with 20 confirmed cases and 2 deaths in Uganda (WHO). A ceasefire on one front does not pause an outbreak on another. Health aid was also expected to fall, with the sharpest cuts projected for sub-Saharan Africa and health aid possibly returning to 2008 levels (Devpolicy).

Families and clinics share one lesson. When prices can jump on a headline, a budget needs a buffer sized to a clock, not to a mood. At Pawanax, we build for clinics that must plan around prices that can move in a week.

The next 90 days

This chapter's sources end in June. The two forecasts below are about what comes next, not about what the June sources showed. Both were published on 16 October 2026 and will be scored with the Brier score on their resolution dates.

F1: 0.65. Brent front-month settles above $90 on 15 January 2027. The June sources ended near $75, so F1 is a bet that the premium comes back above $90 by then. If it does, the score is (0.65 − 1)² = 0.12. If it does not, the score is 0.65² = 0.42. An always-50% forecaster scores 0.25, so F1 counts as skill only if we land clearly below that.

F12: 0.45. Kenyan authorities report at least 50 flood-related deaths in the October–December 2026 short rains. The earlier season was already serious. Wikipedia listed 112 deaths as of 3 April, and a relief directory cited at least 169 deaths and over 150,000 displaced by April (Wikipedia). Our model links the El Niño to October–December flooding, and NOAA's 63% chance of a very strong event is the input behind that link. If the reported toll reaches 50, the score is (0.45 − 1)² = 0.30. If not, it is 0.45² = 0.20.

Both forecasts resolve on 15 January 2027, one day after the ninety-day mark. We will publish the scores either way.

Part 6 of The Signal Ledger. Read the series: [The Signal Ledger](/blog/signal-ledger-2026-read-backwards).

Cover image: Strait of Hormuz.png — ESA; contains modified Copernicus Sentinel data 2020 · CC BY-SA 3.0 IGO. ESA satellite image dated 14 May 2021, showing the Strait of Hormuz near Qeshm Island. Some clouds were removed in this retouched version. Geographic illustration only; it does not show current shipping, passage conditions, or 2026 events. Resized for web display.