VoxEU Column Macroeconomic policy Monetary Policy

Why is euro inflation so low?

Eurozone inflation has been persistently declining for almost a year, and constantly undershooting forecasts. Building on existing research, this column explores the conjecture that low inflation in the Eurozone results from an excess demand for safe assets. If true, this conjecture would have definite policy implications. Getting out of such a ‘safety trap’ would necessitate fiscal or non-conventional monetary policies tailored to temporarily take risk away from private balance sheets.

Inflation in the Eurozone stood at 0.4% (year on year) in November. It has been persistently declining for almost a year, and constantly undershooting forecasts. The Eurozone is now clearly diverging from many advanced economies, where inflation is either on the rise – albeit at moderate levels – as in the US, or, when falling, still remaining close to target, as the UK.

Inflation has kept falling even in ‘core’ economies, where employment remains strong and the output gap has been reduced, such as Germany. Recent data show little prospect for improvement. According to the October World Economic Outlook (IMF 2014b), under current policies, the Eurozone inflation rate “is expected to remain substantially below the ECB’s price stability objective through at least 2019”.

This column explores one specific conjecture, that underpinning those evolutions, including the lack of growth, is one single and common factor – an excess demand for ‘safe assets’. Essentially, economic agents develop a strong inclination for holding money and government bonds, in preference to any other financial or real assets; and, doing do, are induced to cut down on their expenditures. This is both a sign and a consequence of extreme risk aversion. As a result, Europe may be caught into a ‘safety trap’ from which there is no escape other than a temporary transfer of risk from private to public balance sheets.

If true, this conjecture would have definite policy implications. The column brings no new insights. Rather, it builds on existing research (Caballero and Fahri 2013), to analyse the situation in the Eurozone.

The current situation

Disinflation in the Eurozone has been accompanied by symptoms commonly associated with a lack of aggregate demand: stagnant or negative growth, and increasing unemployment, especially in peripheral economies. The causation from demand to inflation may seem obvious, as the output gap has simultaneously increased in 2013 (from 2.5% to 3%).

However, it seems difficult to relate the recent downward shift in inflation to pure demand factors. Nominal wage growth is still robust, with compensation per employee rising by 1.1% in annual terms for 2014Q2. While fiscal policy has been a sizeable drag on the economy in 2012–2013, it is not the case anymore, and the Eurozone structural fiscal balance is stabilising. Finally, according to the IMF, the Eurozone output gap (at approximately 3%) is smaller than in the US (3.5%).

On the supply side, the appreciation of the euro could account for the decline in inflation in 2013. But that pass-through is now over and the effective exchange rate has depreciated by 5% in the last seven months. Import prices have been stable over the last six months, and may be expected to increase following the recent depreciation.

In sum, demand and supply factors could explain the low inflation prevailing up to the first quarter of 2014. They cannot account for the constant, and unexpected, fall in inflation that has occurred since.

Something deeper may be at work. Inflation is a process – while impacted by temporary shocks, it is fundamentally driven by endogenous dynamics and expectations. There are many signs of such deteriorating dynamics in the Eurozone. Core inflation (excluding food and energy), is now at 0.7%. Most significantly, inflation expectations are dropping fast. At a short horizon, they now stand below 0.7% for 2016. Longer-term expectations, broadly stable until June, are now falling too. For the first time in 15 years, the ‘5 years in 5 years’ swap stands below 2%, at 1.84%, down from 2.4% at the beginning of the year, prompting the ECB President to note this in his speech at Jackson Hole in August 2014. This is dangerous – as Japan’s experience has shown, long run inflation expectations are very rigid and, once destabilised, may move irreversibly down.

Safe assets shortage: How does it work?

The intuition is straightforward. In a safety trap, there is a strong demand for precautionary balances that cannot be satisfied under prevailing economic and financial conditions. In order to accumulate more safe assets, economic agents have to reduce their consumption or investment, thus depressing aggregate demand for goods. There is a close analogy with a basic monetarist model in which an excess demand for money translates, for given income and wealth, into an excess supply of goods. There is also some analogy with a liquidity trap, where the demand for money (one specific safe asset) becomes infinite.

In developed contemporary models (e.g. Caballero and Fahri 2013), the safe asset shortage works through the interest rate. An excess demand for safe assets drives down the risk-free natural interest rate, possibly to negative levels in real terms. When inflation is already low and the economy hits the zero lower bound, it becomes impossible to reach the necessary (negative) real interest rate. The economy falls into a ‘trap’ with cumulative disinflation and the possibility of an ever-deeper recession. Just like a liquidity trap, there is there is a self-perpetuating aspect in the safety trap because lower inflation increases real interest rates which, in turn, leads to lower inflation. In a sense, however, a safety trap is more serious. Increasing preferences for safety can constantly push the equilibrium interest rate further down, thus widening the gap between effective and equilibrium interest rates. An increased demand for safe assets therefore acts as an endogenous tightening of monetary conditions (Fisher 2013).

One important insight is that in a safe asset trap, the lack of demand is not the ultimate cause of disinflation. Rather, both low demand and low inflation are joint manifestations of an underlying disequilibrium in asset markets. This disequilibrium may persist for a long time, as real interest rates cannot adjust.

The safe asset hypothesis therefore explains the persistence of low inflation and no growth despite extremely easy financial conditions. It also helps to clarify some of the puzzles currently affecting advanced – and more specifically, European – economies.

Three puzzles

Take, first, the financial behaviour of corporate firms. Considered in aggregate, they have been simultaneously issuing debt (most often with high yields and light covenants), hoarding cash (in the order of 18% of GDP in France and Germany) and buying back their shares. Admittedly, this is not purely a European characteristic, and such behaviours may result from tax arbitrages. The phenomenon, however, shows that low inflation and low demand cannot fully be explained by a ‘balance sheet recession’ (Koo 2009) triggered by the deleveraging occurring in the private sector. Actually, outside the banking sector, little deleveraging is taking place in aggregate. In the Eurozone, corporate debt is up by 7% of GDP as compared to 2007 (Hannoun 2014). There is strong issuance of new debt, but proceeds are used to finance precautionary cash hoarding. Policy debates may be excessively focusing on debt. For the Eurozone, at least, the main economic problem lies on the asset – not the liability – side of corporate balance sheets.

Indeed, a second puzzle relates to physical investment, which remains well below pre-crisis levels (by an order of 2% of GDP) despite extremely easy financial conditions. The safe asset hypothesis brings some elements of clarification. In a safe asset trap, risk-free assets are abnormally attractive since they offer an excess return (as compared to equilibrium). This creates a disincentive to invest – high risk-free rates compress the spreads with other assets to levels insufficient to compensate for perceived economic risk.

Still, while reluctant to take ‘economic’ risk, investors show a great appetite for ‘financial’ risk. This is a third puzzle, recently illustrated by the IMF in its latest Global Financial Stability Report (IMF 2014a).

Here, it is useful to identify the two main characteristics of a safe asset. First they are ‘information insensitive’ – they keep the same value (and therefore protect their owner’s wealth) in all possible states of the world (this is the definition of safety). And, second, they are liquid and can be exchanged for money without loss of value at any single moment in time. It has been shown (Dang et al. 2012) that those two properties are closely related. But they are not identical. They provide a matrix through which one can look at the investment universe.

Physical investments are the exact opposite of safe assets. They are both illiquid and fully exposed to economic risk (and, therefore, sensitive to uncertainty). Risky financial instruments – such as high yield debt or equity – are in an intermediary situation. Their value is information sensitive. But, under the assumption of market liquidity, they can be transformed into a safe asset (and money). Therefore they do not suffer from the irreversibility attached to physical assets. Easy monetary policies strengthen the perception of liquidity and increase substitutability between safe and risky financial assets. They do not affect, however, the illiquidity of physical assets. Hence the currently observed disconnect between economic and financial risk.

Why could there be an excess demand for safe assets?

Many causes can explain a shortage of safe assets – some global, some more specific to the Eurozone.

In the Eurozone, like in the rest of the world, the demand for safe assets is mechanically bound to increase, as they are needed to meet new regulatory and market requirements. Banks will hold more Government debt and other ‘high quality’ assets to meet the liquidity standards set up by the Basel Committee. Market transactions, whether or not intermediated by central clearing counterparties, will increasingly depend on the provision of high quality collateral.

At the same time, and more specific to the Eurozone, the supply is shrinking. The Eurozone crisis has resulted in a large chunk of government debt losing its ‘safe asset’ status, as markets have become increasingly sensitive to credit and liquidity risks. This annihilation of previously safe government debt amounts to several trillions of euros. It is conceivable that it will produce macroeconomic effects on a large scale and for a long time while investors adjust their behaviour to a new, and unforeseen, environment.

Finally, and more relevant to recent evolutions, economic uncertainty and expectations of deflation generate their own preference for safety. An excess demand for safe assets is just one consequence and manifestation of strong aversion to risk.

Policy implications

If the conjecture proves correct, any policy response must necessarily increase the net supply of safe assets. Measures that would simply substitute one safe asset for another would not work. It turns out that some monetary policy actions, including non-conventional one, fall into that last category. This requirement – increasing the net supply of safe assets – therefore provides a useful criterion through which to assess various policy initiatives.

In the Eurozone, the excess demand for safe assets has been, until recently, satisfied by an expansion of central bank liabilities – notably through the ECB’s long-term refinancing operations (LTROs). This is a useful but imperfect solution. Access to the central bank’s balance sheet, while much broader in the Eurozone than in the US, is still limited to financial institutions. If households and firms need to hold additional safe assets, they depend on banks to ‘transform’ that central bank liquidity into other safe assets, i.e. to issue money through credit. As is well known, this process is currently severely impaired in the Eurozone. Broad money aggregates are growing at an annual pace of 2.5%, but external counterparties play a major role and domestic credit is shrinking (-1.5% year on year in October).

Would quantitative easing (QE) – i.e. purchase of government bonds by the central bank – work? In a first step, it would simply substitute one safe asset (money) with another (government debt), leaving the total net amount unchanged. What follows will depend on the degree of preference for safe assets. It is usually assumed that, with less government debt in their hands, investors will ‘rebalance’ their portfolios by purchasing additional risky assets, thereby pushing their price up and driving their yields down. This portfolio rebalancing is essential for QE to have an impact on the real economy, through a reduction in spreads and an increase in financial wealth.

Crucially, however, the process relies upon some (imperfect) substitutability between safe and risky financial assets. Investors must be willing to swap money and government debt for riskier corporate debt, real estate, or equity in order to get a higher return. A strong preference for safe assets would inhibit that arbitrage and significantly impair or paralyse portfolio rebalancing.

Things are very different when the central bank engages in direct and massive purchases of private (risky) assets. In that case, the net amount of safe assets does increase. This is exactly what the ECB is currently doing through its programme of asset-backed securities (ABS) and covered bond purchases. This action will potentially bring huge benefits to the Eurozone economy.

In the short run, it will increase the net volume of safe assets, although by limited amounts and mainly in the hands of the banking sector. Should, however, the demand be ‘satiated’, some portfolio rebalancing towards risky assets could occur, credit would increase again, and one could expect some resumption of consumption and growth, together with an increase in inflation.

In the longer run, the ECB can help to create a new class of safe assets in the Eurozone and significantly increase their supply. Plain vanilla (not structured) ABS are very safe (with an average default rate of 2%), and will remain so if strict rules are implemented and liquidity is guaranteed. ABS have no ‘nationality’ as they can be made of assets originated in different countries and be traded across borders. A vibrant ABS market will further delink sovereign, banking, and credit risk, offering investors a truly ‘European’ asset class. It will help to overcome the geographical segmentation of European capital markets and improve the transmission of monetary policy in different countries.

At present, however, the European ABS market is only one-fifth of the size of the US market. The ECB is now creating the conditions for it to grow. As the ultimate buyer, it can impose the conditions that ABS would have to meet and quickly standardise the market, thus bypassing the complicated and burdensome process of EU regulation. (see Bank of England and ECB 2014). Ideally, the ECB should set itself as a ‘market maker of last resort’, committing to sell and buy ABS and guaranteeing market liquidity for some period of time.

Admittedly, this is no ordinary job for a central bank. And it involves taking some risks. This may be unavoidable in a safe asset trap when the only remedy is for the public sector to temporarily take additional risk by itself, on its consolidated balance sheet. If the strategy is successful, such action will be costless as aggregate risk in the economy will diminish and so will the excess demand for safe assets.

While this remains extremely controversial inside the Eurosystem, the necessity to take risk away from private balance sheets now seems well internalised by policymakers. Through its recent communication, the ECB is signalling its willingness to significantly increase the size of its balance sheet. It implicitly aims at a certain level of leverage, and, therefore, seems to accept some additional and temporary risk-taking.

Finally, of course, issuing new government debt and spending the proceeds could increase the amount of safe assets. The private sector would then hold both money and newly issued bonds. The power to tax gives governments a comparative advantage in creating safe assets. This advantage, however, is fragile and heavily dependent on fiscal and debt sustainability. Preserving ‘fiscal capacity’ (Caballero and Fahri 2013, 2014) is essential. It raises difficult issues of fiscal policy governance and moral hazard. Fiscal rules increase the fiscal space to the extent that they credibly ensure debt sustainability in the long run. Disagreements on how they should be implemented in current circumstances have the opposite effect. Setting up an appropriate, robust, and commonly agreed fiscal governance becomes crucial at a moment when the Eurozone is just regaining some fiscal credibility.


Bank of England and European Central Bank (2014), “The case for a better functioning securitisation market in the European Union – A Discussion Paper”, May. 

Caballero, R and E Farhi (2013), “A Model of the Safe Asset mechanism (SAM): Safety Trap and Economic Policy”, NBER Working Paper 18737. 

Caballero, R and E Farhi (2014), “On the Role of Safe Asset Shortages in Secular Stagnation”, in C Teulings and R Baldwin (eds.) Secular Stagnation: Facts, Causes and Cures, VoxEU.org eBook. 

Dang, T V, G Gorton, and B Holmström (2012), “Ignorance, Debt and Financial Crises”, mimeo. 

De Paoli, B and P Zabcszyk (2014), “Risk Aversion and the Natural Interest Rate”, Liberty Street Economics, Federal Reserve Bank of New York, 16 July.

Fisher, P F (2013), “Reflections on the meaning of ‘risk free’”, BIS Paper 72. 

Kocherlakota, N (2014), “Low Real Interest Rates”, Speech, 5 June.

Koo, R (2009), The Holy Grail of Macroeconomics, John Wiley & Sons.

Hannoun, Hervé (2014), “Central Banks and the Global Debt Overhang”, Speech, 20 November. 

International Monetary Fund (2014a), “Improving the Balance Between Financial and economic Risk Taking”, Global Financial Stability Report, Chapter 1, October.

International Monetary Fund (2014b), World Economic Outlook, Chapter 1, October.

Landau, J-P (2013), “Macroprudential rebalancing”, VoxEU.org, 18 April. 

11,128 Reads