Ramadan doesn’t flatter

1“It’s Ramadan, that’s normal”

Composite example, not a specific client

The first days of Ramadan. The mall branches are empty at midday; the street branches explode half an hour before iftar, and again after taraweeh. On Sunday the finance manager posts the week’s sales in the group chat: three branches didn’t cover their rent and salaries.

A minute later a partner replies: “That’s normal, it’s Ramadan. Eid will make it up.”

It might. But the one number that tells the truth about those three branches was read, and the subject closed in a single line.

2The bank doesn’t say “it’ll make it up”

Since 2011 the Saudi Central Bank (SAMA) has bound banks to written rules on stress testing.¹ Every bank must test itself against “exceptional but plausible” shocks at three levels: mild, moderate and severe. It must also run a reverse stress test: which scenario would break me?

A bank has to imagine its shock, and nobody there reads a bad result and says “it’ll make it up”. Your shock arrives every year, on a known date, already recorded in your books: your weakest Ramadan week. You don’t have to invent it, only read it.

And Ramadan isn’t a bad month for restaurants; it’s an honest one. It takes some hours and gives back others, and it doesn’t flatter the branch built on the hours it takes.

3What the national numbers say

According to SAMA’s weekly point-of-sale data, as reported by Arab News, spending at restaurants and cafés in the week ending 21 February 2026, the early-Ramadan week, fell 28.3% on the previous week to SAR 1.24 billion. In the same week, spending on food and beverages, meaning groceries, rose 2.1%.²

And in the week after Eid al-Fitr 2025, ending 5 April, restaurants and cafés were the only category that rose: up 2.6% to SAR 2.23 billion.³

So Ramadan demand doesn’t disappear; it changes shape, hours and place. One branch catches the shift, another stands in the wrong place at the wrong hour.

One honest line: these are national figures for one week in each year, not a rule that repeats every Ramadan, and they aren’t broken down by city or branch type. There’s no published figure for Ramadan in Riyadh, or for mall versus street branches. Your map comes only from your own weeks.

And your weeks are within reach. Electronic payments were 79% of retail payments in 2024, according to SAMA.⁴ And ZATCA requires every business whose VAT-able revenue exceeded SAR 187,500 in any year from 2022 to 2025 to integrate with the Fatoora platform by 1 February 2027,⁵ just before Ramadan, which is expected in early February per the Umm al-Qura calendar (the moon sighting decides). Weekly sales by branch are no longer a data problem; they’re a decision.

4The number: weakest-week coverage

Rent and salaries don’t know the season. They’re paid in full in the weakest week as in the strongest. So the question isn’t “how far did the branch’s sales fall?” but “did what was left cover what doesn’t move?”

Fixed-cost coverage = the week’s contribution ÷ (the week’s rent + the week’s branch salaries)

The week’s contribution: what’s left of sales after cost of goods and delivery commission, before rent and salaries.

Illustrative example: two branches, each with SAR 25,000 a week in rent and salaries, and each keeping 60% of sales after goods and commission.

• The street branch: a normal week is SAR 80,000, leaving 48,000: coverage 1.92. In its weakest Ramadan week sales fell 40% to 48,000, leaving 28,800: coverage 1.15. It hurt, and it covered.
• The mall branch: a normal week is SAR 55,000, leaving 33,000: coverage 1.32, and it looks fine. In its weakest week sales also fell 40%, to 33,000, leaving 19,800: coverage 0.79.

The mall branch was SAR 5,200 short that week. Repeat it four times and that’s SAR 20,800 another branch paid on its behalf.

That’s SAMA’s reverse question in branch language: how many weeks like this can the branch take before it needs cash from the group? The mall branch can’t take a single one.

5Draw the map

Rows are your branches; columns are last Ramadan’s weeks plus Eid week. Put the coverage figure in each cell. Colour every cell below 1.0 red.

Illustrative example: a group of 14 branches. Three are red in all four Ramadan weeks. Two turn red only in the last ten nights. The other nine hurt, and cover.

The map doesn’t tell you to close anything; it tells you where to look. The two last-ten-nights branches need an hours-and-roster plan. The three that are red all month need a harder question.

6Where the money leaks

The good branches quietly fund the red ones. It doesn’t show, because the group report adds everything into one number. The doctrine puts it plainly: scale multiplies whatever the unit already is, profit or loss.⁶

The Ramadan campaign gets pushed into branches that can’t convert it. A branch without peak staff won’t profit from an ad that sends people to it at peak.⁷ It’s worse when the branch manager is paid on sales rather than contribution, so he asks for the campaign himself.⁸

And the exit decision gets postponed. “Let’s wait for Eid”, then “wait for summer”, then “wait for the school term”. Each wait sounds reasonable on its own.

7The Sunday-morning question

One question, one metric, one rule

In fairness: this threshold is my proposal, not the Doctrine’s. The Doctrine’s exit rule is quarterly: a mature branch whose contribution after rent stays negative for two consecutive quarters goes to the same review.⁹ The Ramadan test is an early warning that comes before it, not a substitute for it.

9Credit, and the next episode

Stress testing is the bank regulators’ idea, and SAMA wrote it into binding rules in 2011. I only carried it from the bank to the branch, and from an imagined shock to a week that comes every year.

And when you see your neighbour’s branch packed in Ramadan and decide to copy his idea… remember you are only looking at the branches that stayed open. That’s the next episode: “I copied my neighbour’s idea and forgot the ten that closed.”