VoxEU Column Financial Markets

The fundamentals of safe assets

What makes government bonds a safe asset? This column shows that the low political and institutional risk of issuing countries and the relative size of the debt market foster a safe asset status, with the latter factor – size – reflecting the special role of the US in providing a large, deep and liquid market for government bonds. Inertia – whether the bond behaved as a safe asset in the past – is also important. Notably, the drivers of safe asset status are heterogeneous within advanced and emerging markets, with external sustainability in particular being relevant for the latter group of countries. 

One of the legacies of the global financial crisis is a relative scarcity of safe assets coupled with stronger demand for safety. Safe assets have stable nominal payoffs, are highly liquid, and carry minimal credit risk. They are particularly valuable during periods of stress in financial markets, as they maintain their nominal value while the value of other assets typically falls. 

The demand for safe assets has increased since the crisis as the probability that investors assign to macroeconomic tail risk is higher than before. At the same time, the supply of safe assets has fallen as the size of publicly and privately issued securities with safe asset status has shrunk considerably after the crisis due to rating downgrades (Caballero and Farhi 2018, Farhi and Gourio 2018).  As a result, yields on safe assets have fallen, as investors are ready to pay a large premium to hold liquid and safe securities: the so called ‘convenience yield’, a term usually used with reference to US Treasuries (Krishnamurthy and Vissing-Jorgensen 2012).

High demand for safe assets has important macroeconomic consequences. The equilibrium safe real interest rate may in fact decline well below zero and below the actual rate, as nominal rates hit the zero lower bound and central banks find it difficult to further decrease them. A ‘safety trap’ equilibrium (Caballero et al. 2015) may emerge in which, due to the zero lower bound and safe asset scarcity, adjustment takes place through a contraction of global demand (Caballero et al. 2017). Unconventional monetary policies have so far averted the most harmful consequences of ‘safety trap’ scenarios, but they need to be accompanied by appropriate, supervisory prudential and fiscal policies to avoid possible side effects (CGFS, 2019). 

The fundamentals of safe assets 

Despite the growing academic and policy interest in ‘safe assets’, not much research has been done in pinning down their features and determinants. The empirical work so far has largely been limited to US debt securities. One exception to the strict focus on US liabilities is Du et al. (2018), who quantify the difference in the ‘convenience yield’ of US Treasuries and government bonds of other advanced countries and document a secular decline in this premium. Another exception is Habib and Stracca (2012), focusing though on safe haven currencies, not debt instruments.

In a recent article (Habib et al. 2020), we fill this gap analysing empirically the economic fundamentals that make government bonds safe, using a sample of monthly changes in government bond yields in 40 advanced and emerging countries between 1990 and 2018. In our study, the fundamentals of safe assets are country characteristics that robustly predict whether a government bond appreciates in times of stress, as measured by spikes in the implied volatility in US equity markets (a proxy for global ‘risk off’ episodes). The geographical breadth of our sample ensures that the analysis is less US-centric than the rest of the literature, offering a truly global perspective on the determinants of a safe asset. The length of the sample ensures that the study cuts through a number of episodes of global financial turbulence, including the speculative attacks on currencies in the European Exchange Rate Mechanism in the early 1990s, the Asian and Russian crises in the late 1990s, the burst of the dot com bubble in the early 2000s, and the Great Recession.

Our findings suggest that only a handful of characteristics robustly determine the safe asset status of government bonds. These include: 

  • a measure of the political risk rating, an indicator of the quality of institutions of the issuing country; 
  • the relative size of the debt market, reflecting the special role of the US in providing a large, deep and liquid market for government bonds;
  • intriguingly, an inertial term too – whether the asset behaved like a safe asset in the past.

Other fundamentals may also matter to identify a safe haven status, such as strong economic growth, a low level of public debt and a better current account position, even though these findings are less robust across periods and country samples. In fact, the relevant fundamentals are different between advanced and emerging countries. The political risk and institutional quality rating and the size of the debt market are important for advanced countries only, and inertia, real GDP and external sustainability (measured by the current account) for emerging markets only. The US has a special role, but no single fundamental summarises it. For instance, both the size of its debt market as well as the dominance of the US dollar in foreign currency reserves capture somewhat the safe haven role of the US financial market in times of stress. Finally, the results are largely independent on the type of shock that generated financial turbulence.

The economic significance of fundamentals in explaining safe asset status

The relative importance of the factors driving the safe asset status is depicted in Figure 1, where we report the differential change in government bond yields in three groups of countries with respect to the US following a large shock to the VIX. The diamonds report the actual differential change in yields following large global risk shocks. For instance, in the case of emerging markets government bond yields tend to rise on average by almost ten basis points (+8 basis points) following large shocks, whereas in the United States yields decline by ten basis points (-10 basis points); therefore the differential is around 20 basis points. The stacked bars instead report the value predicted by the model, accounting for the relative strength of the fundamentals of safe assets with respect to the US. The size of the debt market, measured as the countries’ share of world public debt, clearly explains most of the divergence in yields between the US and other advanced economies (see the green bars in Figure 1). The political risk rating, instead, appears to be a key variable for investors to discriminate between advanced and emerging markets (see the orange bar in Figure 1).

Figure 1 What makes government bonds safe? Size of the market, political risk and inertia

Changes in bond yields following large financial shocks: differential relative to the US (basis points)

Notes: The chart reports the average change in government bond yields for three groups of countries with respect to the US following a large (two standard deviations) shock to the VIX. The “Actual” value is the unconditional average across all the episodes of large financial shocks. The stacked bars report the value “Predicted” by the model accounting for the relative strength of the fundamentals and drivers of safe assets with respect to the US. “Size” of the debt market is the countries’ share of world public debt. “Political risk” is a rating index from the International Country Risk Group, a synthetic index measuring variables such as political unrest and the presence of conflicts, government stability, the investment climate, corruption, the rule of law and the quality of bureaucracy. “Inertia” is the recursive correlation of the change in yields and the change in VIX in the past. Note that the contributions are not necessarily orthogonal; hence the sum of the predicted factors, the asterisk in the chart, can be different from the actual developments. Safe haven advanced economies include Germany, Japan, Switzerland and the United Kingdom. Source: authors’ calculations

Policy implications

The policy implications of our work are that country fundamentals do play some role in explaining the safe asset status of government bonds. Crucially, some of these fundamentals can be influenced by policy makers, which could improve the institutional risk profile or lower the level of public debt of their economy to foster the safe haven status of their government debt. Moreover, other characteristics, such as the relative size of the debt market, indicate that only a limited number of actors can act as safe haven. This in turn underpins, with quantitative evidence, the discussion on the evolution of the international monetary system. In particular, our findings speak to the debate over the creation of a euro area safe asset (Bini Smaghi and Marcussen 2018), if this could be designed in a manner to adequately deal with moral hazard and potentially expand the overall supply of highly rated government bonds. 

Authors’ note: The views expressed in this column belong to the authors and do not necessarily represent the views of the ECB.


Bini Smaghi, L and M Marcussen (2018), “Delivering a safe asset for the euro area: A proposal for a Purple bond transition”,, 19 July.

Caballero, R J, E Farhi and P-O Gourinchas (2015), "On the global ZLB economy",, 5 November.

Caballero, R J, E Farhi and P-O Gourinchas (2017) “The safe assets shortage conundrum”, Journal of Economic Perspectives 31(3): 29–46.

Caballero, R J and E Farhi (2018) “The Safety Trap”, Review of Economic Studies 85(1): 223-274.

Farhi, E, and F Gourio (2018), “Accounting for Macro-Finance Trends: Market Power, Intangibles, and Risk Premia”, Brookings Papers on Economic Activity Fall: 147-250.

CGFS (2019) “Unconventional monetary policy tools: a cross-country analysis”, Committee on the Global Financial System Paper No. 63, Bank for International Settlements, October.

Habib, M M and L Stracca (2012) “Getting beyond carry trade: What makes a safe haven currency?”, Journal of International Economics 87: 50–64 (summarised on Vox).

Habib, M M, F Venditti and L Stracca (2020), “The fundamentals of safe assets”, Journal of International Money and Finance 102 (forthcoming in April 2020).

Krishnamurthy, Arvind and Annette Vissing-Jorgensen (2012) “The aggregate demand for treasury debt,” Journal of Political Economy 120: 233–267.

3,360 Reads