With inflation declining and constantly undershooting forecasts, and the economies gripped by fiscal austerity, zero-lower bounded interest rates, and gloomy expectations, the Eurozone may have found itself caught into what Caballero and Fahri (2103) call a ‘safety trap’, i.e., an excess demand for safe assets (money and government bonds). Essentially, agents in the Eurozone have responded to extreme risk aversion by developing a strong inclination for holding safe assets. To accumulate more of these assets, they have reduced consumption and investment, thus depressing aggregate demand.  In a safety trap, the economy suffers from cumulative disinflation, increasing real interest rates, and depression, and policies aimed at stimulating aggregate demand by boosting generic wealth, such as forward guidance, have less traction than in conventional liquidity traps. All these symptoms are very much suggestive of the “new secular stagnation hypothesis” – evoked by Larry Summers in his famous speech at the IMF in 2013 and further elaborated  – as being germane to the situation in the Eurozone.
The president of the European Central Bank (ECB) Mario Draghi has once again bought precious time to the Eurozone. The first time was in 2012, when he avoided major disaster with his “whatever it takes” policy insurance pronouncement, which involved the ECB pledging unlimited purchases of sovereign debt in secondary markets for applicant countries, subject to fiscal conditionality as judged appropriate. And he has done it again more recently, with the much suffered decision to go for a quantitative easing (QE) action plan, whereby the ECB has started injecting new money into the system by purchasing securities, in particular government bonds. Draghi is testing the limit of what monetary policy alone can do to let a large economy as the Eurozone out of a safety trap.
Will the Draghi bazooka be powerful enough to do the job?
Diego Valiante (2015) on this blog has offered an interesting analysis of the potential effects of QE in the Eurozone, and has raised interesting questions.
Let me try to approach the issue from a more analytical perspective, and assess the policy options available for Eurozone countries to address a safety trap.
Modeling secular stagnation
I have recently worked out a model of secular stagnation, which relies on a strong form of liquidity preference and could easily be extended to the case of a safety trap. I have used a DSGE model where the upsurge of pessimistic expectations causes high liquidity preference to become the source of a persistent drop in demand. In my model:
- Agents draw utility from current consumption and asset holdings
- The utility of assets is formalized as deriving from their being vehicles to future (uncertain) consumption at different speed (liquidation cost) and power (store-of-value capacity)
- Rational expectations interact with changing market sentiment, and
- Agent expected incomes incorporate anticipations of labor market developments. Secular stagnation follows from the agents increasing their demand for liquidity as a significant and persistent deterioration in their expectations raises their utility from being ultra liquid.
The model explains the forces that may have kept the equilibrium interest rate below zero in the Eurozone since after the debt crisis, and bears implications for the policy options available to escape stagnation.
Stagnation and policy options
Negative interest rates
Would breaching the ZLB through negative interest rates (NIR) really help the economy exit the trap? Theoretically, it would; in practice, it is much less certain. Assume the central bank resolves the issues concerning the application of NIR to cash the way, say, Willem Buiter or Miles Kimball suggest, and succeeds in moving borrowing and lending rates into negative territory. This would push agents to search for alternative safe assets earning higher returns (e.g., a foreign reserve currency or government bond). Any such asset would deliver a higher marginal utility to holders (nothing could be done to compress its marginal utility edge), and become the newly preferred liquid asset; it would then supplant in agent portfolios those liquid assets whose liquidity premiums the monetary authorities had sought to neutralize through NIR. With liquidity preference dominating the agents’ attitudes, NIR signals fail to stimulate consumption and to reach across the whole spectrum of interest rates, since the agents are willing to absorb any amount of the “new” liquid asset available, also induced to do so by the demand for it driving its price (and marginal utility edge) further up. Under strong liquidity preference, a new safety trap emerges, the NIR stimulus stops short of incentivizing agents to move into higher risk taking (that is, investment in productive capital), and the economy remains in secular stagnation.
For similar reasons, quantitative easing (QE) – whereby the central bank buys specified amounts of financial assets from the private sector, thus raising their prices (lowering their yield) and increasing the monetary base – is ineffective to boost aggregate demand. In fact, while QE may succeed in raising asset prices (contrary to anticipations of Wallace neutrality, and in line with recent empirical findings), under liquidity preference dominance it fails to stimulate consumption and investment.
The reason is the following. As the marginal utility of liquid assets (say, money) raises above that of current consumption and other less liquid assets, and the ZLB binds interest rates from reaching their (negative) equilibrium level, QE amounts to the central bank buying less-liquid long-term assets in exchange for reserve money, supposedly bringing the marginal utility of the former into equality with that of liquid assets and below that of riskier assets. QE, thus, pressurizes safe asset prices as it seeks to raise the relative attractiveness of riskier assets. Yet, under liquidity preference dominance and binding ZLB, agents absorb any amounts of reserve money created and hold on to them without changing their consumption and investment plans: the policy-induced reduction in the nominal interest rates on less-liquid long-term assets is not (and cannot be) large enough to prop up the marginal utility from holding risky assets beyond that of money and consumption. In a safety trap, given strong preference for safe assets, even larger doses of QE would be necessary – ceteris paribus – to bring their marginal utility at par with that of money and consumption; yet, even in the best of circumstances, this would still be above that of riskier assets (unless QE would be able to affect inflation expectations, as discussed next). For the same reasons, and in line with Caballero’s and Fahri’s conclusion, no forward guidance policy commitment would be effective in raising the value of risky assets.
Assume the central bank commits to being “irresponsible” (Krugman 1998) and to doing “whatever it takes” to drive the economy out of stagnation. Under ZLB, it undertakes to execute QE to inject additional reserve money balances into the economy with a view to increasing the marginal utility of consumption and capital asset holdings by raising inflation expectations. The question is: what are the channels through which QE can influence inflation expectations?
- First, to the extent that the central bank purchases massive amounts of illiquid assets, and commits to holding onto them perpetually (eventually rolling over all maturing debt assets), QE policy actually becomes “helicopter money” policy and can impact spending decisions through the fiscal lever (see below).
- Second, short of this twist in policy, the central bank is left without effective channels. Take consumption first: as nominal interest rates are pushed down to their lower bounds, intertemporal consumption cannot be affected by them. The alternative would be for the central bank to affect expected sales and sales prices, but this would require it being able to affect consumption: a vicious circle. Look next at investment: in order to stimulate investment demand at the ZLB, the central bank should seek to raise the marginal utility of capital assets relative to other assets. But this, too, requires future expected sales and sale prices to go up. The central bank should therefore stimulate intertemporal consumption; yet, as just discussed, it is unable to do so.
In conclusion: at the ZLB and with strong liquidity preference, since long-term debt and liquidity are perfect substitutes at zero interest rates (and their marginal utility exceeds that of riskier assets), QE can’t work unless it is supported by fiscal policy, as discussed next.
As Nick Rowe provocatively puts it, the only way to ensure QE success would be for the government-owned central bank to “move towards communism”, where the state purchases and owns all the assets in the economy…
On the wake of Bernanke (2002), various authors have defended helicopter money (HM) as the most effective macro policy tool for economies in deep recessions. HM drops can provide newly created money directly to households and private businesses, or the government, without generating new (public or private) debt, in order to stimulate spending. Since central banks have generally no mandate to give money away (they can only exchange one asset for another), HM drops need to be backed by the budget-approval process and must essentially involve fiscal policymaking decisions (Grenville 2013). The advantage of this monetary cum fiscal policy instrument is to provide new purchasing power directly to (private or public) agents who are best placed to spend it immediately.
Today, the concept of HM refers to money creation operations intended to support aggregate demand by financing public spending or tax reduction programs. Buiter (2014) evaluates HM effectiveness through a formal model, and identifies the conditions under which it boosts aggregate demand. One of these conditions is the irreversibility of the new money base stock creation, which constitutes a permanent addition to the total net wealth of the economy. This is made possible by the (‘fiat’) money base constituting an asset for the holder but not a liability for the issuer. Operationally, irreversibility can be attained if HM drops are executed either by: having the government issue interest bearing debt, which the central bank would buy and hold in perpetuity, rolling over into new government debt when the existing debt on its balance sheet reaches maturity. In this case, the government would face a debt interest servicing cost, but the central bank would make an exactly matching profit from the difference between the interest rate it receives on its debt and the zero cost of its money liabilities, and would return this profit to the government, or by having the central bank purchase special government securities that are explicitly non-interest bearing and never redeemable. In terms of the fundamentals of money creation and government finance, the choice of these two routes would make no difference (Turner 2013).
The impact of HM and the importance of the irreversibility condition can both be appreciated by considering the intertemporal budget constraints of the individual agents and the government. As the part of the deficit is irreversibly funded through money creation, the government budget constraint is permanently relaxed by an equivalent amount, implying that the disposable income of individual agents can be equally raised. Since irreversibility preempts Ricardian equivalence effects, consumption increases permanently. Importantly, in my model HM acts also through the expectations channel: a large and sustained monetary-cum-fiscal stimulus increases both agents’ expected income and the marginal efficiency of capital, thus raising the marginal utility of consumption and risky assets relative to liquid assets, and stimulating the demand for them.
The characteristic of fiscal policy is that the government can use only debt and/or taxation to finance its budget: relaxing the budget constraint now in order to allow for current lower taxation (or higher public spending) requires larger government indebtedness, which in turn implies higher taxation and/or lower public spending at some future dates. Whether, and to what extent, this is going to affect current consumption decisions depends on various factors, such as, inter alia, the agents’ relevant time-horizon and factors binding their rationality, the state and sustainability of public finances, and the credibility of the fiscal and monetary authorities.
It should be noted that, with low nominal interest rates, running a front-loaded expansionary budget with a view to postponing fiscal adjustment to some future dates may seem to be a winning strategy to help the economy out of stagnation, especially when spending in public infrastructures yields high economic returns. However, the question is whether and the extent to which the issuance of new public debt affects the equilibrium interest rate, and the answer ultimately depends on the market assessment of future debt sustainability. In the case of a largely indebted country, for instance, the success of the fiscal stimulus rests on the markets trusting that the stimulus be successful in triggering output growth, thus improving debt sustainability. Such positive perception allows interest rates on public debt to remain low while the government stretches the budget to finance the stimulus. Perceptions could be negative, however, and multiple equilibria are possible depending on market beliefs, meaning that an element of uncertainty is inherent in the exclusive use of fiscal policy as a way out of secular stagnation. For this reason, fiscal policy ranks second to HM drops in terms of effectiveness: it can resolve safety traps by boosting demand without creating public or private debts.
Policy options for the Eurozone
The European Central Bank is now trying to resurrect the Eurozone through QE. Unfortunately, in line with theory predictions, QE experiments so far have shown to exert no significant impact on aggregate demand and consumer price inflation: by construction, QE has no way to funnel money to those who are most inclined to spend it. QE has delivered positive effects only when it has been implemented in conjunction with decisive fiscal stimulus, since it has counteracted the interest rate rises that deficit and debt growth would have otherwise caused. In Japan, QE has kept long-term interest rates low, but the already huge debt has grown further larger, raising the likelihood of future necessary fiscal corrections and increasing the uncertainty for private sector spending decisions. Debt, therefore, becomes a key consideration, and the following policy conclusions ensue:
- QE must be accompanied by fiscal expansion for policy to succeed in stimulating aggregate demand and inflation: only fiscal action could guarantee that money would be spent; yet,
- In economies with large public debt, fiscal expansion would worsen the debt situation, thus neutralizing the expansionary impact through higher interest rates and Ricardian equivalence effects. Therefore,
- A combination of i. and ii. is necessary to mimic some form of HM operation, whereby monetary and fiscal policy mutually interact at no extra cost to taxpayers. For this to be possible,
- The central bank must commit to holding permanently the public debt purchased, so as to ‘sterilize’ its corresponding government (and taxpayer) obligations indefinitely: the purchased debt would be monetized and (public and private) budget constraints permanently relieved.
These conclusions are especially pertinent to largely indebted Eurozone countries, and not just because of the binding 3% deficit-to-GDP ratio ceiling under the Fiscal Compact. In fact, even if this ceiling were to be interpreted flexibly, or if they were hypothetically removed, issuance of new debt by these countries (in the absence of central bank ‘sterilization’) would raise perceptions of debt unsustainability, triggering interest rate rises and raising fears of future tax adjustments. In this case, only the combination of actions above would deliver effective demand stimulus and drag the economies out of stagnation.
Bossone, B. (2014), “Secular stagnation”, Economics, Discussion Paper No. 2014-47, November 19 http://www.economics-ejournal.org/economics/discussionpapers/2014-47
Bossone, B (2015) “QE + permanent debt purchases + fiscal expansion = helicopter money: recipes for the Eurozone”, The World Bank All About Finance, January 26
Bossone, B., T. Fazi, and R. Wood (2014) “Helicopter money: the best policy to address high public debt and ieflation”, VoxEu, 1 October
Buiter, W. H. (2009), “The wonderful world of negative nominal interest rates, again”, VoxEu, 4 June
Buiter, W. H. (2013) “Negative interest rates: when are they coming to a central bank near you?, Financial Times, Willem Buiter’s Maverecon, May 7
Buiter, W. H. (2014), “The simple analytics of helicopter money: why it works – always”, Economics, Vol. 8, 2014-28
Caballero, R. J., and E. Farhi (2013) “On the role of safe asset shortages in secular stagnation”, in C. Teulings and R. Baldwin (eds.) “Secular Stagnation: Facts, Causes and Cures”, CEPR and VoxEu, pp. 111-122
Draghi, M. (2012), Speech by Mario Draghi, President of the European Central Bank at the Global Investment Conference in London, 26 July http://www.ecb.europa.eu/press/key/date/2012/html/sp120726.en.html
Gambacorta, L., B. Hofmann, and G. Peersman (2012) “The effectiveness of unconventional monetary policy at the Zero Lower Bound: a cross-country analysis”, BIS Working Papers No 384, Bank for International Settlements, August
Grenville, S. (2013) “Helicopter Money”, VoxEu, 24 February
Joyce, M. A. S., M. R. Tong, and R. Woods (2011) “The economic impact of QE: lessons from the UK”, VoxEu, 1 November
Kimball, M. (2013) “A minimalist implementation of electronic money”, Confession of a Supply-Side Liberal, May 20
Krishnamurthy, A., and A. Vissing-Jorgensen (2013) “The ins and outs of LSAPs”, prepared for the Federal Reserve Bank of Kansas City’s Jackson Hole symposium, August 9
Krugman P (1998) “It’s baaack! Japan’s slump and the return of the liquidity trap”, Brookings Papers on Economic Activity 2
Landau, J. P. (2014) “Why is euro inflation so low?”, VoxEu, 2 December 2014
Summers, L. H. (2014a) “U.S. Economic Prospects: Secular Stagnation, Hysteresis, and the Zero Lower Bound”, Business Economics, Vol. 49, No. 2, National Association for Business Economics
Summers, L. H. (2014b) “Reflections on the new ‘Secular Stagnation hypothesis’”, VoxEu, 30 October
Turner, A. (2013) “Debt, money and Mephistopheles: how do we get out of this mess”, Cass Business School Lecture, 6 February
Valiante D. (2015) “Myths and realities of European QE”, EconoMonitor, March 11th http://www.economonitor.com/blog/2015/03/myths-and-realities-of-european-qe/#sthash.DTtkXcFn.dpuf
Wallace N (1981) “A Modigliani-Miller theorem for open-market operations”, American Economic Review, Vol.71, n.3, June, 267-274
Wen, Y. (2014) “Evaluating unconventional monetary policies ─why aren’t they more effective?”,
Working Paper 2013-028B http://research.stlouisfed.org/wp/2013/2013-028.pdf
 I wish to thank Larry Summers for his encouraging comments on a previous version of this article, and Coen Teulings for his very helpful suggestions. Of course, I am the only responsible for the arguments and opinions expressed in it.
 Landau (2014) has an interesting narrative of the safety-trap story as it may apply to the Eurozone.
 See Bossone (2014).
 See Buiter (2009, 2013), and Kimball (2013).
 An interesting real-world example of this is suggested by this comment by J. P. Koning (http://jpkoning.blogspot.ca/2015/01/the-zlb-and-impending-race-into-swiss.html)
 This issue of Wallace (1981) neutrality as relates to QE has recently become the subject of blogosphere debate. See, for instance, Richard H. Serlin, “The Intuition Behind Wallace Neutrality”, Economics, Finance, Personal Finance, Politics, and Other Subjects with a Focus on Intuition, Clarity, and Non-Misleading, August 10, 2014 http://richardhserlin.blogspot.it/2014/08/the-intuition-behind-wallace-neutrality.html.
 See, for instance, Krishnamurthy e Vissing-Jorgensen (2013), and the studies cited in “Evaluation of quantitative easing”, Econbrowser (http://econbrowser.com/archives/2014/11/evaluation-of-quantitative-easing).
 See Wen (2014).
 In the Euler setup, as interest rates move down to the lower bound, current consumption leaps upward but this is just one time effect since future period consumption adjust to lower levels.
 The reduction of interest rates, including via QE, can at best, equate the marginal utility of liquid assets with that of riskier assets. Stimulating the demand for the latter would require further interest rate reductions (below their lower bound) and/or an increase in the marginal utility of the riskier assets themselves.
 See Nick Rowe, “Fractional reserves, capital, communism, and the optimum quantity of money”, Worthwhile Canadian Initiative, 8 September, 2014
 See Bossone et al (2014) for references.
 A comment on irreversibility: the irreversibility condition has nothing to do with the fact that, at any future date, the central bank may decide to withdraw part or all of the liquidity injected in the system by selling bonds form its portfolio. In this case, the holders of liquidity would exchange money for the bonds sold by the central bank. Yet the total financial net worth of the economy would not change, only its composition would (shifting from more to less liquid assets): the addition to the economy’s financial net worth originally operated through the HM drop would not (and could not) be undone by any new open market operation. In this sense, and only in this sense, should irreversibility be understood.
 For a discussion of alternative policy options for the Eurozone, see Bossone (2015).