Central bank profits and multiple equilibria
The theoretical and empirical realisation that economies are subject to multiple equilibria can help understand a number of phenomena that unique equilibrium models cannot explain: jumps in financial variables; self fulfilling expectations; the critical role of apparently unrelated variables, so called sunspots, for economic developments; the excessive volatility of financial variables with respect to that of fundamental variables; the connected excessive correlation of financial variables across countries with respect to the correlation of fundamentals; non-linear effects; booms and busts.
There are two prices to be paid, however, for this richness of results, one is conceptual, the other practical. The conceptual price is that models become, so to say, too rich to the point of becoming indeterminate. The practical cost is that economic policy authorities can use the concept of multiple equilibria to make “speculators” or “bad luck” the scapegoat for crises, forgetting that multiple equilibria are only possible when fundamentals have been brought, often by faulty policies, into a zone of vulnerability. 1)Paul R. Masson, Multiple Equilibria, Contagion, and the Emerging Market Crisis,IMF Working Paper WP/99/164, November 1999. Just to refer to the recent €-crisis, it is not by chance that it was not Germany that was subject to a shift from a good to a bad equilibrium but the countries in the €-periphery, which had accumulated acute macroeconomic disequilibria.
The two prices just mentioned are worth paying, however, to have a framework which allows understanding phenomena that are clearly under our eyes in the real world.
The concept of multiple equilibria can also specifically deal with two issues of particular importance for central banks:
- Solve the apparent contradiction between the Bagehot principle (in a crisis lend at penalty rate to troubled banks) and what central banks have done during the Great Recession (where they lent at less than the market rate). The solution is that central banks should lend at a cost higher than the “good equilibrium” interest rate (and in this sense at a penalty rate) but lower than the “bad equilibrium”, current rate;
- Explain why moving from a bad to a good equilibrium can result in large financial gains and why only an institution with infinite (or very large) financial means and a long horizon (which is a good description of a central bank) can do this. Of course, this works only as long as fundamentals are consistent with the good equilibrium. In this respect the matching of, on one hand, central bank financial action and, on the other hand, real adjustment to establish stronger fundamentals, is critical.
This post is dedicated to the second point above, to estimate specifically whether the ECB has achieved large financial gains during the crisis, as it would be consistent with having engineered a move from a bad to a good equilibrium. Such a result would also be consistent with the famous Friedman maxim that central bank interventions (he referred to those on the foreign exchange market, but the argument can be generalised to all kind of interventions) are only stabilising, and therefore macro-economically useful, if they are profitable. Indeed interventions that move the economy from a bad to a good equilibrium are both financially profitable and macro macroeconomically stabilising.
Specifically, this post presents estimates of the “intermediation gains” achieved by the ECB 2)Here I use the term ECB as synonymous of the more pedantic Eurosystem, that includes the ECB and the 18 national central banks of the countries having adopted the €. during the crisis. The way I use the concept of “intermediation gains” can be understood looking at table 1, which is the update of a table originally presented in a chapter of a book that my colleague Tuomas VälimäkiI and I wrote. 3) See: “The Concrete Euro: Implementing Monetary Policy in the Euro Area“.
Table 1: Change in euro money market turnover and increase in Eurosystem balance sheet (2008 – 2011)
|Reduction in unsecured turnover (bn)||Increase in secured turnover (bn)||Net reduction of turnover (bn)||Increase in Eurosystem balance sheet (bn)||Substitution between Eurosystem and market intermediation (%)|
|(1)||(2)||(3) = (1) – (2)||(4)||(5) = (4) / (3)|
This table puts together, and explains, two salient features of the crisis: the drying up of private intermediation (in column (1) the fall of the unsecured market is reported, which is partially compensated by the increase in the repo market in column (2)) and the extraordinary increase in the size of the ECB balance sheet (reported in Column (4)). The sequence of events underlying this table is that a shock hit private intermediation and the ECB offered to carry out on its book that part of intermediation that the private sector was no longer able to do in appropriate quantity and with an affordable price.
This intermediation took two forms.
First, since October 2008 the ECB lent any amount the banks needed, especially in the periphery, by moving from fixed-quantity to fixed-interest-rate refinancing operations, and extended its lending from short term € refinancing to lending also in foreign currency (mostly in $ thanks to the swaps with the Federal Reserve) and medium term (lengthening the maturity of the Longer Term Refinancing Operations from 3 months to 1 year and then to 3 and even 4 years under the forthcoming Targeted Longer Term Refinancing Operation), while “borrowing” in the market the liquidity that banks, mostly in the core, were only willing to invest with the central bank.
Second, the ECB intervened to “repair” dysfunctioning markets: first, in 2009 and then in 2011, with the covered bond market, then in 2010 and 2011 in the sovereign markets of peripheral countries. The Outright Monetary Transaction Program, announced in the summer of 2012 but never actually implemented, was, in a way, the most efficient form of these “repairing” interventions 4)If one measures efficiency in terms of results per units of effort, the efficiency of the OMT was infinite as the “effort” was zero, given that the program did not need to be implemented..
The profits from these two forms of intermediation are estimated below. The estimates are approximate because the information that would be needed for more precise quantification is not always available (for instance the ECB does not publish information on the amounts lent under Emergency Liquidity Assistance, nor on the cost of this sort of refinancing) but also because the purpose here is not to measure precisely these profits but just to see whether their approximate size is consistent with the idea that they corresponded to the re-establishment by the ECB of a good equilibrium. Furthermore, no attempt is made to reproduce the complicated accounting rules of the ECB, as the analysis is based on economic more than accounting concepts.
The bulk of profits from refinancing operations, i.e. lending, is due to the fact that there was an “intermediation margin”, a mark-up, that the ECB applied to its lending, corresponding to the spread between the composite weighted rate on the Main Refinancing Operations, the Marginal Lending Facility and the Emergency Liquidity Assistance, on one hand, and the composite weighted rate on the Deposit Facility and on the Fixed Term Deposits, on the other hand. In addition, the ECB realised coupon revenue on its stock of covered bank bonds, purchased in the two Covered Bond Purchase Programs (CBPP), and on sovereign paper, purchased in the Securities Market Program (SMP). A more precise description of the estimation methodology is given in the appendix.
Table 2 gives our estimates of the remuneration from lending, starting in 2006, taken as the last year before the crisis, and in all years since. The estimate for 2014 is limited to the first six months of the year. The total revenue between January 2007 and June 2014 is thus estimated at 101 billion, while if the remuneration prevailing in 2006 had been prolonged for the following 7.5 years only about 52.5 billion of revenue would have been generated.
Table 2: Estimated revenues
|Year||Remuneration per Year ( EUR Bn)|
Table 2 gives our estimates of the cost of “ECB borrowing”. This was close to zero in 2006, and in earlier years, since at that time there were practically no deposits with the ECB, while the total cost of ECB borrowing between January 2007 and the first six months of 2014 was around 3.5 billion. Also in the last year the cost was nil as the deposit facility rate was zero. Indeed, at the end of the period, the ECB was in the enviable position to earn money both on its lending and on its borrowing, as banks had to pay the ECB for the favour it extended to them by accepting their money!
Table 3: Estimated costs
|Year||Cost per Year ( EUR Bn)|
The bottom line about the profits from the first form of intermediation for the ECB during the crisis was that additional gains of around 45 billions were made: net profits of 97.5 (101-3.5) billion were realized during the crisis while only about 52.5 billion would have been realized if the profits in 2006 would have been maintained for the following 7.5 years.
The intuitive way to understand the origin of these profits is to realise that the enormously expanded size of ECB intermediation more than compensated the lowering of the unitary intermediation margin, approximately represented by the distance between the Main Refinancing Operation and the Deposit Facility rates.
The profits from the second form of intermediation, i.e. purchases on the covered bond and the peripheral sovereign bonds are due to the coupon income, included in the intermediation margin estimated above, and to the capital gains deriving from the increase in price of the purchased bonds. A colleague at Bruegel and I had estimated in 2013 the latter component for the Securities Market Program in 5)Carlos De Sousa,Francesco Papadia, Has the European Central Bank transformed itself into a hedge fund?, Bruegel, March 2013 . Of course since then two things have happened: first, part of the capital gains have taken the form of coupon income; second, the price of the purchased bonds has gone further up. The net result of these two developments in terms of capital gains on the sovereign bonds is estimated below together with the capital gain on the securities bought under the Covered Bond Purchase Programmes.
Using the methodology reported in the Appendix, derived from that used in the post mentioned above, the estimated capital gains are 16.7 billion for the SMP at the end of June 2014 and 5.7 billion for the CBPP.
Overall, the additional profits from the ECB from conducting its extraordinary form of intermediation during the crisis can be estimated at more than 67 billion (45+16.7+5.7). Hefty gains have also been reported for the FED`s action during the crisis, indicating that the phenomenon is not limited to the ECB.
In a way, this result can be seen as the empirical, ex-post confirmation of a Bagehot dictum recently quoted and formalized by Ulrich Bindseil: 6)Ulrich Bindseil and Julius Jablecki, Central Bank Liquidity provision, Risk-Taking and Economic Efficiency, European Central Bank Economic Working Papers Series, N. 1542, May 2013 “…making no loans as we have seen will ruin it [the Bank of England]; making large loans and stopping, as we have also seen, will also ruin it. The only safe plan for the Bank [of England] is the brave plan, to lend in a panic on every kind of current security, or every sort on which money is ordinarily and usually lent. This policy may not save the Bank; but if it does not, nothing will save it.” Contrary to the fears of many, who thought that the central bank would jeopardize its financial soundness by its extraordinary lending during the crisis, Bagehot saw that as the only way to avoid “being ruined”. Consistently with that view, the ECB even made money following the “brave plan”.
Of course an assessment of the consequences for the profit of the ECB from the Great Recession should take into account that the lower interest rates inevitably had a negative effect on seignorage profits. But combining this observation with the intermediation gains estimated above gives rise to an interesting interpretation. If the ECB had limited itself to combating the crisis with “conventional” tools, thus lowering rates, inevitably it would have reduced its profits. However, complementing the action on rates with “unconventional measures” such as the unlimited provision of liquidity, the CBPP, the SMP and even the Outright Monetary transaction program, it more than offset the negative effect on its profits.
In conclusion, the fact that the ECB realised significant extra-profits, so far more than 67 billion, on its unconventional monetary policy measures during the Great Recession is consistent with the hypothesis that it managed to bring back the €-area economy from a bad to a good equilibrium. It would, however, be a mistake to conclude that it did this on its own. Its action matched the macroeconomic adjustment and governance reform conducted by the €-area governments, which improved the fundamentals of the €-area economy and thus reinforced the shift back to a good equilibrium. It would also be a mistake to conclude that the shift to a good equilibrium is permanent: the anaemic growth situation casts a dark shadow over the €-area economy.
Estimation of intermediate gains ( monthly series):
A. Estimation of Revenues
(1). [(MRO+MLF+LTROs )- Reserve requirements]*[MRO rate/12]
(2). [SMP outstanding] * [5 yr IT and DE government bond average Index]/12
(3). [CBPP1+CBPP2 outstanding] *[ iBoxx EUR Covered Index/12]
(4). [ELA estimates]*[MLF rate +100bps]/12
Estimation of ELA was obtained by considering the net change since April 2012 of the ” Other Claims” account of the Eurosystem. Prior to this date ELA was reportedly included in the ” Other Assets” account and has been derived from this pivotal point.
Total revenues: (1)+(2)+(3)+(4)
B.Estimation of Costs
(1). [Deposit facility]*[Deposit facility rate/12]
(2). [Fixed Term Deposits] * [Average allotment rate]*7/365
Total costs: (1)+(2)
Total profits from lending to banks: A-B
Estimation of capital gains:
Using the average Italy and Germany 5 year generic bond index for the Securities Markets Programme and the iBoxx EUR Covered index for the two Covered Bond Purchase Programme, the Zero Coupon price was derived using the formula Price =1/(1+yieldt)t , with maturity t considered 5 years. As an estimate of the initial price, a simple average was used for the time that the purchase was conducted on each programme. The estimation of the gain is: Profit=Outstanding amount at End June 2014 *( Current derived ZC price- Average purchase price )/Average Purchase Price.
*** Madalina Norocea gave assistance in preparing this post.
|↩1||Paul R. Masson, Multiple Equilibria, Contagion, and the Emerging Market Crisis,IMF Working Paper WP/99/164, November 1999|
|↩2||Here I use the term ECB as synonymous of the more pedantic Eurosystem, that includes the ECB and the 18 national central banks of the countries having adopted the €.|
|↩3||See: “The Concrete Euro: Implementing Monetary Policy in the Euro Area“|
|↩4||If one measures efficiency in terms of results per units of effort, the efficiency of the OMT was infinite as the “effort” was zero, given that the program did not need to be implemented.|
|↩5||Carlos De Sousa,Francesco Papadia, Has the European Central Bank transformed itself into a hedge fund?, Bruegel, March 2013|
|↩6||Ulrich Bindseil and Julius Jablecki, Central Bank Liquidity provision, Risk-Taking and Economic Efficiency, European Central Bank Economic Working Papers Series, N. 1542, May 2013|