The Global Crisis and the numerous episodes of bank restructuring or recapitalisation over the past years have had a major toll on public finances, making clear the large implicit guarantees that governments tend to give to the financial sector (Amoglobeli et al. 2015, IMF 2015). Often times, debt increases significantly because an unforeseen obligation materialises. These ‘contingent liabilities’, as they are known in the economic jargon, can have significant economic and fiscal costs. In fact, on many occasions large and unexpected increases in debt across the world were due to the materialisation of contingent liabilities (Cebotari et al. 2009, Weber 2012). During the Asian and Latin American crises, for example, these fiscal costs amounted to up to 50% of GDP (Honohan and Klingebiel 2000). That is why contingent liabilities are often called hidden deficits. Understanding their origin and size, and the triggers for their materialisation, is a crucial step towards better management of these obligations and related shocks.
More specifically, a contingent fiscal liability is a potential obligation for the government, which depends on a possible future event. For example, if a government provides a guarantee for a public or private company’s loan and that company fails to make payments, the lender will call the guarantee. As a consequence, the government will have to take the loan on its books. These liabilities can be embedded into explicit contracts like a loan guarantee, but they can also be implicit unwritten obligations. Political pressure, public opinion, or moral obligations can lead governments to recapitalise struggling banks, but also to cover the debts of local authorities or state-owned enterprises when these are unable to pay. Countries like Australia and the UK record, monitor, disclose, and manage these potential liabilities to minimise risks to their budgets. However, in most cases, taxpayers only know about them when it’s too late; that is, when these liabilities are no longer ‘potential’ but ‘actual’.
So, how often do these potential liabilities materialise, creating fiscal costs for the government? When such events do occur, how large are the fiscal costs? Are some countries more likely than others to experience such events? And are there ways to mitigate the occurrence and impact of contingent liability shocks?
A new dataset
To gain a better understanding of these liabilities, in a recent paper (Bova et al. 2016) we created a new dataset documenting over 200 episodes (covering 80 countries over the period 1990–2014) involving the materialisation of contingent liabilities. The study encompasses a broad range of contingent liabilities, from financial ones to those originating from subnational governments, public-private partnerships (PPPs), legal cases, state-owned enterprises, and private enterprises.
The key finding is that the fiscal cost of these liabilities is large.
- The fiscal cost averages about 6% of GDP and in some cases exceeds 20% of GDP (Figure 1).
While a truly large event is fairly rare, it can potentially cause substantial damage to a country’s debt sustainability when it occurs.
Figure 1. Distribution of contingent liability realisations, 1990-2014
How frequent are these contingent liability shocks?
- We find that on average, a country experienced a contingent liability realisation every 12 years, implying expected costs of slightly over half a percent of GDP per year.
However, this average masks important differences between countries. Not all countries are equally likely to experience these costs. For example, in Brazil it happened on average every 5-6 years and cost 8.3% of GDP. The incidence was much lower in countries like Canada, Hong Kong, and Israel.
Digging deeper, we find that support to the financial sector, including bailouts, accounts for the largest share of contingent liability costs – think of Indonesia, Thailand, and Korea during the Asian Crisis; or think Iceland and Ireland during the recent Global Crisis (Figure 2). Subnational government bailouts, support to state-owned enterprises, and legal liabilities also give rise to substantial costs. Examples of these include Argentina (2001-2004), Greece (2007-2010), and Macedonia (starting in 1999).
Additionally, our analysis shows that liabilities tend to materialise after periods of high growth and coincide with low-growth periods and banking crises. And they tend to happen at the same time, putting considerable strain on government finances. From a fiscal point of view, when it rains, it pours. The Asian Crisis and the Global Crisis are prime examples, as Figure 2 illustrates.
Figure 2. Contingent liability realisations by type and year
The analysis suggests that it takes a long time to recover from one of these budgetary surprises. Figure 3 illustrates the development of key fiscal variables around the time of contingent liability realisations. Contingent liabilities tend to coincide with already worsening public finances, which they amplify. On average, the overall balance deteriorates by 2 percentage points and takes a number of years before reverting to the pre-event level. At the same time, debt increases by around 15% of GDP and stabilises at that higher level afterwards. This indicates the difficulty governments face in reducing debt levels again after sustaining such an important shock.
Figure 3. Contingent liability realisations and the macroeconomy
What are the most important determinants of contingent liability shocks?
- We have learned that countries with stronger institutions and lower growth volatility tend to suffer less from contingent liability realisations.
Indeed, countries with strong public institutions could face a cost which is 30% lower than the average.
How concerning are contingent liabilities in the current macroeconomic environment? In an environment of weak growth, where commodity prices are low and most currencies are depreciating, contingent liabilities are likely to remain a major source of fiscal risk. High levels of corporate debt in emerging markets might be one source of concern (Chow 2015). But also in advanced economies, stocks of explicit contingent liabilities are important. Figure 4 uses Eurostat data to show the stock of government guarantees, PPPs, non-performing loans on government assets and liabilities of public corporations in the EU28. The average of these explicit contingent liabilities is about 50% of GDP and can be substantially higher in certain cases.
Figure 4. Stock of explicit contingent liabilities in EU28
Summary and policy implications
What are the lessons for policymakers? What can countries do to prevent or mitigate costly shocks to public finances?
Countries should build strong institutions to avoid the emergence of these contingent liabilities in the first place. This means better governance at the local government level, in state owned-enterprises, and public-private partnerships. It also means stronger supervision and better resolution regimes for financial sector institutions to avoid the taxpayer having to pay for financial sector bailouts, the most costly source of liabilities.
Lastly, governments need to put in place more transparent and better arrangements for disclosing and monitoring their hidden liabilities. In particular, fiscal frameworks should be strengthened to enhance discipline and limit the excessive growth of contingent liabilities. This is particularly important because the evidence shows that these hidden deficits tend to emerge in times of financial distress when public finances are the weakest.
Authors' note: The views expressed herein are those of the authors and should not be attributed to the IMF, its Executive Board, or its management.
References
Amaglobeli, D, N End, M Jarmuzek and G Palomba (2015), “From Systemic Banking Crises to Fiscal Costs: Risk Factors”, IMF Working Paper WP/15/166.
Bova, E, M Ruiz-Arranz, F Toscani, and H E Ture (2016), “The Fiscal Costs of Contingent Liabilities: A New Dataset”, IMF Working Paper WP/16/14.
Chow, J T S (2015), “Stress Testing Corporate Balance Sheets in Emerging Economies,” IMF Working Paper WP/15/216.
Eurostat (2016), “What is the extent of contingent liabilities and non-performing loans in the EU Member States?”, Press Release (27 January).
Cebotari, A, J Davis, L Lusinyan, A Mati, P Mauro, M Petrie and R Velloso (2009), “Fiscal Risks: Sources, Disclosure and Management”, IMF Fiscal Affairs Department.
Honohan, P, and D Klingebiel (2000), “Controlling the Fiscal Costs of Banking Crises”, World Bank Policy Research Working Paper No. 2441.
IMF (2015), “From Banking to Sovereign Stress: Implications for Public Debt”, Policy Paper.
Weber, A (2012), “Stock Flow Adjustments and Fiscal Transparency: A Cross Country Comparison”, IMF Working Paper WP/12/39.