I had a delightful lunch the other day with Vaidas Urba, a NGDP targeting supporter from Lithuania. Lars Christensen once let him do a post, and I agreed to publish this one. Before turning it over to Vaidas, however, let me anticipate one objection. He talks about “competitive central banking” where others might simply refer to “free banking.” If you object to his terminology, I’d encourage you to look past that superficial issue into the substance of what he has to say. I can see links with the free banking approach of White/Selgin/Woolsey, etc, but also with the Woodfordian interest rate targeting cashless economy. See where you think it fits in:
Free Central Banking
On December 16, 2010, the ECB has decided to increase its capital from €5.76 billion to €10.76 billion, citing increased risks and volatility. [ http://www.ecb.int/press/pr/date/2010/html/pr101216_2.en.html ] Did insufficient capital position of central banks contribute to the depth of the Great Recession? On September 29, 2008, markets have crashed after the TARP was not approved by the House of Representatives. Did the TARP alleviate the shortage of central bank capital, or did it work through the regular fiscal channels? Is modern central banking compatible with the principles of free market? To answer these questions, a model of free competitive central banking is presented here. To enable competition, this model avoids discretion in central banking. It uses Taylor-like rules and NGDP level targeting as its building blocks. The key advantage of free central banking is the automatic inflow of capital into the central banking sector when risk-adjusted return on central banking business is expected to be elevated.
During the Great Recession, many central banks have earned above-normal return on capital, and with hindsight it is clear that extra capital could have been profitably deployed in central banking. During the most dangerous episodes of the crisis, central banks wanted to expand the quantity of capital devoted to macroeconomic stabilization. On September 17, 2008, Bernanke, Paulson and Geithner have reached the consensus that the Fed was not able to solve the crisis on its own, and that there was a need to ask the Congress for TARP funds. On the other side of Atlantic, policymakers were comfortable with the capital position of the European Central Bank. On October 8, 2008, the ECB decided to provide unlimited euro liquidity against a wide range of acceptable collateral, and after a few days the ECB started providing unlimited dollar liquidity, this was a decisive step that eventually restored the functioning of dollar money markets. [ http://www.ecb.europa.eu/press/pr/date/2008/html/pr081015.en.html ] It was only later, during the euro sovereign debt crisis, when the ECB decided that the additional capital was needed to preserve macroeconomic stability, and then the fiscal authorities have established EFSF and EFSM programs. TARP, EFSF and EFSM share a common feature – these capital pools reduce money demand, thus softening the blow when risk management considerations prevent central banks from expanding the money supply sufficiently.
Free central banking makes the process of attracting new capital automatic. Suppose there are multiple competing private central banks in a single country. All liabilities of these central banks have the status of a legal tender, and all equally and interchangeably serve as a medium of account and exchange. All the central banks have to follow the same exact rules, and no discretion is allowed. The only place where discretion exists is on the level of investors, they have to decide on the optimal use of their capital. If they believe the expected risk-adjusted return is attractive, they can establish a new central bank, or they can purchase the shares of existing central banks on the market.
All the central banks in this model have to pay the interest on reserves according to the same variant of Taylor rule that uses the deviation from the nominal GDP level target as the key element. The leading contender in the rules versus discretion debate is the Taylor rule, the most promising idea for monetary stability is the nominal GDP level targeting, this free central banking model combines them both.
Additional rules are needed to regulate the risk of central banking. Central bank asset risk needs to be limited in a manner consistent with macroprudential requirements. Central banks must be well capitalized, extraordinarily high capital requirements need to be imposed to mimic the current low risk profile of main central banks who do not really need explicit capital buffers as their financial strength is enhanced by the monopoly rents they will earn in the future. Three forces shape the rules that regulate the central bank asset purchases: the need to avoid discretion, optimization of risk exposure, and systemic macroprudential considerations. Government bonds are not good assets for central banks to hold as their risks are hard to measure and to control, and when money demand is elevated, the term risk premium of government bonds is likely to disappear. On March 1, 2013, Bernanke presented a chart, according to which the 10-year treasury term premium is negative, in other words, on average the Fed is expected to lose money from treasury purchases. Bank of Japan is allowed to purchase equity index ETFs, however equities are too risky to comprise a major portion of central bank balance sheets. Taking ECB’s three year LTROs as an inspiration, safer assets can be constructed from equity index baskets by using them as a collateral for non-recourse three year loans, while applying large haircuts to reduce risk. The interest for such overcollateralized three year loans is the interest paid on reserves plus a margin that is not regulated, but that rather is determined by markets. Haircut size is determined according to a fixed formula that is based on stock-market index/GDP ratio at the inception of the loan, so the macroprudential considerations are taken into account in this model.
Would free central banking have made any difference before the crisis? In my opinion, no. The demand for non-cash base money was very low before the crisis, central banking was very unattractive as a business (with the exception of cash which is ignored here), and the regular commercial and shadow banks have competed with central banks by supplying lots of better and cheaper substitutes for base money. We can say that before the crisis, central banks have emulated the actions of a competitive free central banking sector. By the way, I am aware of a quasi-Austrian argument that central banks have in effect employed too much capital before the crisis by supplying underpriced implicit bailout guarantees to the commercial and shadow banking sectors. If this argument is correct, free central banking would have constrained the unsustainable credit growth boom before the crisis.
The special character of free central banking would only become visible during the crisis. When the credit crunch started, the outlook for the non-central bank part of the financial sector deteriorated, while simultaneously higher demand for central bank money created new profit opportunities in the expansion of monetary base. The disruption that was caused by the credit crunch would have been softened by the automatic reallocation of capital to the central banking sector. At the same time, the aggregate demand would have remained anchored by the NGDP level targeting that is encoded in the Taylor-like rule for calculating the interest on reserves.
The free central banking model discussed here is presented as an illustration only. It is not intended to serve as an actual policy proposal. Centuries of additional data and huge volumes of new research are needed before we could set a particular Taylor-like rule in stone. Indeed, many market monetarists do not believe at all that it is possible to produce a robust Taylor rule. In Scott Sumner’s NGDP futures targeting model the setting for the policy instrument is produced as an outcome of profit-maximizing market process. Free central banking model is different, as it uses the profit-maximizing market process for a different purpose – to allocate capital. Correspondingly, Scott Sumner’s proposal would result in a suboptimal quantity of capital deployed in the activities of central banks, while free central banking would result in a higher and possibly suboptimal volatility of NGDP-linked bonds.
Some central banks, especially those with publicly listed equity, could use some of the elements of free central banking in practice. Taking Switzerland as an example, we can argue that by issuing a new class of callable equity instruments to the public, the Swiss central bank could have established an exchange rate ceiling at a level that is more conductive to macroeconomic stabilization with a lower risk to the central bank itself.
Alan Greenspan once remarked that history has not dealt kindly with the aftermath of protracted periods of low risk premiums. When the next such period ends and a new credit crunch arrives, let us hope that central banks will employ optimal levels of capital.