A monthly, forward-looking read on Cyprus tourism demand — by source market, by district, with every forecast scored in public the following month. Built on official CYSTAT data, airline advance bookings and Tharro's own occupancy and visibility research.
June printed 489,965 arrivals, −1.7% year on year (CYSTAT) — the smallest decline of 2026, and the basis of the official "return to stability" narrative. Decompose it, and the stability disappears:
Israel's June 2025 base was crushed by the 12-day war and its flight suspensions — June 2026 compares against a hole. Strip Israel out and June was −12.5%. More telling still: excluding Israel, the market deteriorated from −7.3% in May to −12.5% in June — the underlying trend got worse in exactly the month the headline improved.
| Month | 2025 base | of which Israel | Point | Band (YoY) | Driver |
|---|---|---|---|---|---|
| July | 589,116 | 76,557 | ~545k (−7.5%) | −11% to −5% | Israel's base had already recovered by July 2025 — the tailwind is gone |
| August | 602,026 | 105,597 | ~547k (−9%) | −12% to −6% | The August 2025 Israeli base was a post-war catch-up peak — Israel turns headwind |
| September | 570,635 | 80,115 | ~531k (−7%) | −10% to −4% | Hardest comparison of the year (Sep-25 was +12%) |
The call in one sentence: −1.7% was the best print Cyprus will see this summer. The Israel base effect that manufactured June's stability expires in July and inverts in August — Israel would need over 105,000 August arrivals just to be flat. We expect the July release (mid-August) to show a decline of roughly −5% to −11%, and the recovery narrative to reverse. This barometer says it a month in advance — and scores it, in public, next edition.
Risk to the bands: Israel's forward flight book to Cyprus is deeply negative (section 05) — base-distorted, but the downside within our July–September ranges sits with Israel, not Europe.
| Market | Jun 2026 | Jun YoY | H1 YoY | The read |
|---|---|---|---|---|
| United Kingdom | 161,913 | −10.8% | −13.8% | One third of the market, down double digits all year. The Briton is still travelling — demand observed elsewhere in the Med grew strongly this season — but is choosing between destinations more actively than ever, and Cyprus is losing that choice in the actuals. The counterweights: Cyprus holds the least-bad UK forward book in the basin, and airlines kept UK seat capacity essentially flat. Winnable — but only by fighting for the choice. Still the biggest commercial exposure on the island. |
| Israel | 80,343 | +165.6% | −8.6% | June's surge is a base artifact — but the level is real: it matches the recovered September-2025 peak. Now market #2 at full run-rate. H1 still negative: March 2026 collapsed to just 1,537 arrivals during the regional escalation. |
| Poland | 35,871 | −2.0% | +4.8% | The only major market growing in H1. Held through both shocks. Quietly the most reliable volume market Cyprus has. |
| Sweden | 26,884 | +13.2% | −3.0% | The surprise of the table — a record June, fully recovered from April's −24%. Forward bookings sit 7pp above Sweden's Med median: the Nordic swing toward Cyprus is holding. |
| Germany | 21,587 | −19.0% | −6.9% | 2025's growth star has reversed hard — and unlike the UK, this one is Cyprus-specific: Germany's forward book to Cyprus is 22pp below its own Med median. Losing share, worsening. Top of the watch list. |
| Greece | 13,910 | −12.8% | −7.5% | Soft all year. |
| Romania | 13,308 | −21.1% | −14.4% | Sharp reversal from 2025 growth. |
| Hungary | 7,301 | +10.9% | +18.5% | Small, but the fastest-growing market in the table. |
The Western-Europe decline nobody is reporting: June year-on-year — Austria −33%, France −34%, Italy −34%, Netherlands −36%, Switzerland −18%, Denmark −22%; the US −19.5% across H1. This is broad-based, and it isn't a war echo — paired with falling per-tourist spend (−10.3% in April) it behaves like a price-competitiveness problem. The market is consolidating around the UK (shrinking), Israel (volatile), Poland (growing) and the Nordics (recovering).
Allocating June's market swings through stay distribution: Israel's +50k June swing lands ~+11k in Larnaca (22.5% of Israelis stay there) and ~+10k in Limassol (19.9%), while at any plausible UK stay-share Paphos loses 8–10k visitors from the UK decline alone. In arrivals volume, the island split in two: east up on Israel, west down on the UK.
But arrivals aren't rooms. Our occupancy model — visitor-nights (arrivals × official per-market stay-lengths) measured against registered hotel capacity and calibrated on official occupancy anchors — puts national hotel occupancy at roughly 67% in May (vs ~77% in May 2025) and ~74% in June (vs ~84%): on the order of ten occupancy points below last year, in months whose arrivals headlines read −4.9% and −1.7%. A model estimate, published with that label — and scored against the official statistics when they print. The district-level occupancy split follows in Edition 2, once the model is extended with per-district capacity from the official hotel register.
The two live windows this summer: UK late booking for August–September stays (the only path to softening the UK gap this season), and the entire October book. Both are decided in the next eight weeks.
Advance flight bookings on the books at end-June for travel still to come, across nine Mediterranean destinations. Raw levels overstate weakness everywhere — 86% of all origin-destination pairs in the basin are negative, and Cyprus is a late-booking market whose book fills close-in. The reliable signal is the differential: each market's Cyprus bookings versus that market's median across the other eight destinations.
| Market | CY book | vs Med median |
|---|---|---|
| Czech Republic | −46.7% | −34pp |
| Israel | −51.7% | −27pp |
| Germany | −49.3% | −22pp |
| Greece | −39.1% | −17pp |
| USA | −36.8% | −17pp |
| Market | CY book | vs Med median |
|---|---|---|
| Denmark | +36.9% | +38pp |
| Switzerland | +2.0% | +34pp |
| Ireland | +5.4% | +31pp |
| United Kingdom | −12.9% | +11pp |
| Poland | +1.3% | +9pp |
Bookings measure demand intent; scheduled seats measure supply commitment. For August–October, airlines have kept Cyprus capacity essentially flat: 2.23 million seats, −0.6% against last autumn — even as the outstanding book runs deeply negative. Airlines, who watch these booking curves daily, are betting on the late window. Crossing the two signals sorts every market into one of three buckets:
| Bucket | Markets (seats YoY) | What it means |
|---|---|---|
| Capacity-backed growth | Denmark +30.5% · Croatia +34.1% · Slovakia +40.4% · Poland +12.0% · Greece +8.7% · Switzerland +5.4% | Seats added and (for Denmark, Poland, Switzerland) bookings above their Med median — the swing markets are real and airline-underwritten. Greece is the odd one: seats up while its Cyprus book lags — expect aggressive fares on that corridor. |
| Held capacity, weak book — the conversion fights | UK −1.7% · Israel +0.9% · Germany −5.4% | The UK keeps 634k autumn seats, near-flat, despite the soft book: full access, late-window fill expected — the fight is conversion, not connectivity. Germany's held capacity against a collapsing book points to fare pressure now, winter cuts later. Israel's flat seats at the recovered level make our September flat-Israel assumption capacity-consistent. |
| Structural retreat | France −40.0% · Czechia −28.6% · Austria −25.1% · UAE −20.7% · Hungary −13.1% · Bulgaria −11.6% · Italy −10.6% · Romania −10.5% | Seats withdrawn — these markets cannot recover this autumn regardless of marketing, because the aircraft aren't coming. France and Czechia now show the full triple: falling arrivals, worst-tier bookings, deep capacity cuts. |
Implication for the section-1 bands: flat seats cap the downside — the bottom of the August range would require load factors to fall hard across held capacity. We keep the bands unchanged and let the data judge, but the seat commitment shifts probability toward the upper half of each range.
What the discovery surface looks like for Cyprus, from Tharro's AI-visibility research:
Next edition adds the per-market cut this month's data begs for: who owns the answer surface for the growing markets — Israel, Poland, Sweden — versus the market everyone needs to win back.
The common thread: the island's demand mix is rotating faster than any annual strategy can track. Which market, which window, which channel — it changes month to month this year. That's the argument for reading this barometer monthly, and running the capture side continuously rather than at budget season.
| Call (locked before the June release) | Predicted | Actual | Verdict |
|---|---|---|---|
| Cyprus June arrivals, YoY | −2% to −7% | −1.7% | Missed high by 0.3pp. Direction right — we underweighted the Israel base effect. This edition models Israel's base explicitly, month by month. |
| UK June arrivals | 163.7k–174.7k | 161,913 | Missed low by 1.1%. The UK is running below even our pessimistic band — which is exactly why it's the structural watch item. |
Method, in plain terms. Forecasts are a momentum ensemble on official CYSTAT arrivals series (national × source market), with the Israel base effect modelled explicitly and published as bands, not points of false precision. Occupancy figures are Tharro model estimates — visitor-nights against registered hotel capacity, calibrated on official occupancy anchors — and carry that label until scored against the official statistics. Every forward call in this edition gets scored, publicly, in the next one.