Yesterday's Bloomberg "trial balloon" that the ECB would cut its inflation forecasts while boosting GDP expectations was confirmed moments ago, when during his prepared remarks Draghi increased the governing council's Eurozone GDP expectations as follows:
- 2017: from 1.8% to 1.9%
- 2018: from 1.7% to 1.8%
- 2019: from 1.6% to 1.7%
However, what the market was focusing on was the ECB's inflation expectations, and it was here that the ECB confirmed that the central bank with the world's biggest balance sheet has once again failed to stimulate inflation, slashing its forecast for 2017-2019 inflation across the board, blaming it on commodity prices. The new forecast is as follows:
- 2017: from 1.7% to 1.5%
- 2018: from 1.6% to 1.3%
- 2019: from 1.7% to 1.6%
WIth that cut, the ECB is taking a big gamble because as Bloomberg strategist Tanvir Sidhu writes, while downward revisions to ECB’s 2017 and 2018 forecasts are plausible to roughly around 1.5% given current peak inflation dynamics, "the central bank risks its credibility if cutting longer-term expectations to pre-QE projection levels."
He adds that lower-for-longer and Japanification of the bund curve will continue to play out if the 2019 forecast does come out in line with a report yesterday saying that HICP will be around 1.5%, about projection levels before QE. And while Draghi cut 2018 to only 1.3%, the lowest forecast in the series history, he still expects 2019 inflation to remain just above the 1.6% bogey.
# As Bloomberg concludes, cutting the longer-term forecast would make Draghi’s communication more challenging given the ECB’s 33% issue limit means it would have to end QE earlier than inflation dynamics warrant.
As a result, the fate of ECB credibility now remains in the handover of HICP inflation from 1.3% in 2018 to 1.6% in 2019. Any further cuts to the outer years in the coming meets, the ECB will find itself in deep trouble, as it is rapidly running out of German bonds to monetize, and has little other recourse on how to stimulate Eurozone inflation....