A bit more than two months ago in early December, there was much fuss about the ECB’s secret Anfa (Agreement on Net Financial Assets) and how the Central Banks of France and Italy allegedly misused it for additional monetary easing to the tune of roughly half a trillion euros, by buying sovereign debt. The public outrage over a secret agreement and using it to circumvent the prohibition of monetary financing, prompted the ECB to first release an explainer, that Anfa was quite the opposite of what was speculated in the media – but nobody believed it and politicians and economists (including myself) demanded the agreement to be made public. And on February 5th, the ECB did publish the text, along with the signatures of the NCB governors, the technical appendix and an extended explainer. Unfortunately, the agreement is a rather unreadable document, but nevertheless, I read through it and came to the conclusion that Anfa is actually bad, but not as bad as we may have thought. Anfa actually does what the ECB claims, namely to limit the amount of assets the NCBs may hold for non-monetary policy purposes. But there are several catches.
The biggest one is the notion that there could be “non-monetary policy” purchases of assets by a central bank, without sterilisation, which Anfa does not explicitly stipulate. It does not matter for whatever purpose a central bank buys securities, it is always monetary policy, because it has a liquidity effect, except for when there is a counteracting sterilisation, i.e. a fine-tuning operation in the parlance of the ECB, that extracts the liquidity from the markets that was injected by an Anfa-operation.
Second, and related to the first one, is the notion that the NCBs can financially operate without affecting the monetary union. Any financial operation of any NCB affects the liquidity amount of the whole of the euro area – it is a monetary union after all.
Third, the sum of NFA stood at roughly €470bn on December 4th, using the consolidated balance sheet of the Eurosystem used in the explainer, which amounts to about 17% of the total. This would be ridiculously high, even in normal times, but since the bulk of the balance sheet is actually made up of QE-purchases (€1.9tr or 69% of the balance sheet), compared to “normal” central bank activities, the NFAs are even higher. Granted, NFAs also include gold and foreign reserves (can someone explain to me how gold and reserves would qualify as being a held for a non-monetary policy purpose?) but still, the amount of NFAs is very high. Mario Draghi himself said the purchases of the NCBs are hard to understand and that he cannot rule out that some of these purchases might actually be forbidden monetary financing.Which brings me to the next catch.
Anfa stipulates on quite a few occasions that the ECB’s governing council controls the orderly use of Anfa and that it will hold misconducting NCBs responsible.
Only, that it doesn’t. Anfa says only that the control can be exerted, but there is no clear-cut procedure of how to make sure that NCBs not only adhere to the letter of the agreement but also to the spirit. In fact, this “control” is about as fubar as the “control” of the Maastricht criteria by the EU council. It is the governors themselves who sit as judge and jury over the misbehaviours of their peers, requiring an unanimous vote to decide whether a single NCB has misused Anfa – no wonder that the ECB explainer states reassuringly that this has never been the case.
Another catch is the secrecy. The ECB is not only a public institution that serves some 350 million Europeans but it is – surprise! – also a central bank, as such its most important asset is trust of the public. Secret agreements with the possibility of misuse do not serve the public and doing this in secret undermines the already stretched trust of the public in the ECB. Anfa explicitly states in Article 7.4 that it should be confidential. The only reason for this can be that the ECB and the NCBs knew beforehand that the whole thing is a little fishy and would not withstand public scrutiny. A sensible solution would have been to either scrap Anfa altogether or to come up with an agreement that would be acceptable for the public. Instead, the ECB and the NCBs chose the path of secrecy, in the hope not to get caught. Now they are caught and it is embarrassing. Lucky for them that Europeans are more concerned with other problems right now, than the dirty nitty-gritty of some obscure agreement.
Last and what troubles me the most is that Anfa bears the signature of Bundesbank president Jens Weidmann, along with all the other NCB governors of course. The troubling question is, why on earth a downright conservative (in Barro’s sense) central banker such as Mr Weidmann who knows and endorses the concept that a modern central bank’s fundament is trust of the public, would sign a secret agreement, let alone a secret agreement formulated with such weak and vague supervision requirements. All of the ECB’s directorate and governing council should know better, not only Mr Weidmann.
There are several consequences that must follow from revealing Anfa. First, if Anfa is to stay, we need stronger supervision that the NCBs adhere to it. The ECB or the NCBs are unable to control themselves, so it should be a committee of the Europan Parliament alongside economists and lawyers from all Euro Area members, who report to the public on say an annual basis.
Second, it has become clear that the construct of the Eurosystem where the NCBs still have national duties and are the integral part of conducting monetary policy cannot work. The set-up of America’s Fed is different insofar that the regional Feds are not responsible for single states, but for many of them, sometimes just parts of a state. This is more in line with the suggestion by Burda (2013) to reorganise the Euro Area. However it still misses the point, that in the US single states can very well default on their debt, not only theoretically but also with historical precedent.
This is not the case in the Euro Area since the Maastricht treaty is just a paper tiger, due to the same problems of control and enforcement that plagues Anfa. Hence, a Burda-like reorganisation of the Euro Area would make things better, but until we have either enforceable fiscal separation (like Maastricht stipulates but is unable to implement) or a fiscal union (which will not happen in the foreseeable future), abolishing the NCBs would be best (but terribly unfeasible). Only then monetary policy is guaranteed to be identical across the Euro Area. Since abolishing the NCBs is not going to happen any time soon, it boils down to a vamped-up Maastricht that really is able to enforce the No-Bailout, No-Exit stipulation required for the Euro Area’s future and a control committee for Anfa (or actually for the ECB and all of the NCBs) to ensure that the common monetary policy is not distorted by unilateral actions of the NCBs. Such an amendment of the current set-up is still politically hard to come by, but far easier than reorganise the Euro Area along Burda’s suggestion or create a fiscal union, although the latter two would be much more desirable.