Poland: Weak industrial production hit by calendar effects, one-offs

In June, industrial production fell by 2.5% year-on-year compared to an increase of 7.7% YoY in May and 9.2% YoY in April. The data was significantly worse than consensus (+2.4% YoY) and our forecast (+2.9% YoY). We assumed that June would be weak due to extremely unfavourable calendar effects - April and May had one more business day (in YoY terms), while June had two fewer business days in YoY terms.

However, today's data is worse than the nominal difference in working days alone would suggest – the seasonally adjusted figure showed IP rising by 2.7% YoY in June, while the April-May average was 5.8% YoY.

Without a full sector breakdown, it is difficult to precisely explain the reasons for this slowdown. It may be a first sign of weakening domestic demand in Germany and the euro area. In previous months, stable domestic demand in Germany made Polish production very resilient to recession in the German manufacturing sector. This is because Poland's exports mainly land in the European domestic market, while only 14% are consumed outside Europe. So weaker production in June may signal that domestic demand softened in the eurozone.

However, we think one-offs may offer the best explanation (although warmer weather seems to be the new norm). In a very hot June, the local labour market watchdog recommended shortening work hours. We see a strong drop of production in energy-intensive sectors (production of metals, mineral, mining).

We also think the impact of calendar effects may be higher than implied by the nominal difference in working days. The school year ended earlier than usual, and this was followed by the long Corpus Christi weekend (this bank holiday was exceptionally late this year). This encouraged parents to start holidays early. So effectively, there were fewer working days than the calendar would imply.

Overall, we don’t expect the weakness in production to continue on this scale. There is strong impulse of pre-election social transfers in the pipeline. A one-time pension bonus was paid out in April and an extension of the 500+ child benefits will be gradually paid out in the coming months. That should boost household incomes by one third in 2Q and 2H19. This factor should support domestic demand and locally-oriented production, keeping consumption as the key GDP driver. Tomorrow's retail sales data should show how strongly the retirement bonus influenced consumer spending, although we worry that the data could also be distorted by unusual calendar effects.

The weak June production reading deducts about 0.3-0.4 percentage points from 2Q17 GDP estimates, calling for growth close to 4.6% year-on-year rather than the 5%YoY rate we mentioned before. But we are waiting for tomorrow's retail sales, which should provide evidence about the strength of the main GDP driver, consumption.

The data should not change the MPC's bias too much, rather it will provide justification for its aversion to tightening, which it discussed last month. The main message from the press conference was that rising core inflation and a consumption boom are not enough to consider tightening. Today's data (if indeed caused by real changes rather than calendar effects) supports the MPC's bias. We see some downside risk to our call for 2019 GDP at 4.7% YoY, but growth should still stay safely above 4%, providing rather limited risk for the fiscal side in the short term (ie 2019-2020).