The argument is often presented by some in the following form : « banks would refuse to sell sovereign securities if the reserves they obtain in return were not remunerated, or were taxed at a negative rate. » Thus, the negative rate imposed by the ECB on cash held by banks (see our insight here) would be an obstacle to the effectiveness of the large-scale asset purchase program (QE) that will soon be implemented by the ECB. This post argues that this argument is unfounded.
Let’s imagine that banks’ cash holdings are remunerated at 1% instead of the current -0.20%. The ECB, launching its Quantitative Easing program, wants to buy securities from banks [0]. In this context, the ECB will try to find a selling bank on the market. Some banks want to sell long-term debt securities with a yield of 2% for a price of €1,000, while others do not want to sell these same securities for this price. If the ECB wants to buy a large quantity of these securities, it will have to convince the banks that do not want to sell their securities at a yield of 2%: it will then simply offer a higher price to buy these securities [1]. We can therefore see things in this way: banks will agree to sell their securities when the gain they make on the sale of these securities is large enough to offset the loss of yield caused by the sale of a (long-term) security at 2% for cash remunerated at 1%. In other words, they will agree to sell when the capital gain sufficiently offsets the decline in interest income.
With a negative rate on bank cash, the reasoning is the same. To convince reluctant banks, the ECB will offer higher prices to buy securities. An important and perhaps counterintuitive point at first glance is that there is no plausible reason why there should be more reluctant banks when cash is remunerated at -0.20% than when it is remunerated at 1%. The reason is simple: yields on debt securities will also have fallen, and the difference between the return on cash and the interest rate on the debt security previously used as an example will be the same [2]. Exchanging an asset yielding 2% for cash yielding 1% yields the same profit as exchanging an asset with the same characteristics yielding 0.80% for cash yielding -0.20%. The opportunity cost remains the same, and in this context, the ECB will not be any more constrained in its asset purchases.
Thus, the argument that banks would not want to sell their securities to the ECB because of the negative return on cash overlooks the fact that the reasoning in such a case must be based on opportunity cost rather than nominal cost. The negative rate is not a constraint on the effectiveness of the ECB’s asset purchase program [3] [4].
J.P.
Notes:
[0] We simplify here by assuming that the ECB only buys from banks.
[1] We ignore here « price-insensitive » investors for whom the supply of assets is not sensitive to price (see [4]).
[2] We assume a constant term premium, which does not alter the relevance of the reasoning proposed here (as the term premium is not part of the argument we are discussing in this post).
[3] Note that the negative rate may become a constraint in itself in a different context. In this context—which we will discuss later—the ECB could decide, with QE, to set a threshold rate on liquidity subject to the negative rate, as Switzerland did recently.
[4] See F. Ducrozet’s analysis of the potential limits to the development of QE (notably the presence of « price-insensitive » investors).
