The reference rate has stood at 3.75% since the 5 March cut. Eurostat's harmonised measure put Poland's July inflation at 3.1% year on year, against 3% for the EU as a whole. That comparison is largely decorative for a domestic inflation-targeting central bank; what matters to the Council is the NBP's own target band of 2.5% plus or minus one percentage point, and the March projection already placed the CPI path above the 3.5% upper edge of that band through all of 2026, with a return to the 2.5% point target only in mid-2027.
The projection pack in hand is stale by its own admission. It predates the June fuel-cap expiry and the Middle East supply shock, and bank research previewed by the desk expects the coming July round to show a materially higher CPI path through mid-2027, not a lower one. If that preview is right, the projection due out will push the return-to-target date later than March assumed, not earlier. That is the opposite of what a sector pricing further cuts needs to see.
Warsaw's banks are not behaving like a sector waiting nervously for that verdict. WIG-BANKI's 2.27% gain on 19 August 2026 outpaced WIG20's own 1.72% rise the same session, and the five-session read, up 2.26% for the banks against 0.48% for the broader index, shows the sector leading rather than following. PKO Bank Polski trades at a P/E of 13.3 with a 5.3% dividend yield; Bank Pekao at 10.5 times earnings yields 7.6%. Neither multiple screams a market bracing for margin compression from rate cuts arriving faster than the March path assumed.
A bank index that gains faster than the broader market on the same day the projection behind its rate-cut premise is expected to get less dovish is pricing a policy path the data has not yet delivered.
That is the split this desk has run since 3 August: WIG-BANKI's rally rests on an NBP easing case, and every new data point, the July HICP print, the stale March projection, the fuel-cap and Middle East complications, cuts against that case rather than for it. The 13 and 14 August sessions briefly widened this gap, when WIG20 fell 1.44% on stock-specific weakness in KGHM and Orlen while WIG-BANKI slipped only 0.26%; that episode is now overtaken by the banks' own five-session gain into 19 August, so the divergence this desk flagged has not resolved either way, it has simply moved.
The industrial production and PPI release due 20 August 2026 will not settle this on its own. The test that matters is the NBP's own July projection round, still unpublished. If it pushes the return-to-target date to 2027 Q4 or later, as bank previews suggest, that would confirm the bank sector's rate story is running ahead of the central bank's own data, not alongside it. If instead the round holds the mid-2027 date from March despite the fuel-cap and supply-shock complications, the rally's premise survives its first real test. Either way, the projection's publication is the event this reading is built to be judged against, not another CPI print.
