Our opinion is that:
The direct cost of ECB policy rates for the banking system is low and mainly paid by weakly capitalized banks from the periphery. A further reduction of the deposit rate with no change in the so-called “corridor” would not change substantially that cost, nor would it increase excess reserves. These are mainly the consequence of liquidity injections by the ECB (LTRO, MRO, QE, ELA) and have no significant impact on credit growth, which is mainly the consequence of supply (GDP growth, deviations from the Taylor rule, market sentiment) and demand (capital requirements, risk appetite, available liquidity), but can impact negatively the leverage ratio of large wholesale banks. In a highly regulated world where ROE is driven by regulatory formulas, credit growth is mostly relevant for the few capital rich banks with limited payouts and for banks with high NPLs ratios which need to be replaced by performing loans. Lower rates impact bank profitability through different channels, from NII dynamics to balance sheet management (AFS securities, pension liabilities) or business mix (asset management vs. lending, etc.), but banks are generally negatively geared to lower rates. As it is the case for rates themselves (down before QE, up after) the market consensus tends to front-run the negative impact of QE on bank profitability before QE is actually announced. Crucially, QE also tends to sharply reduce risk premium and discount rates applicable to share valuation, leading to multiple expansion that often more than compensate the reduced consensus expectations. Post QE, consensus tends to bounce back and multiples to stabilize.