II. Insurance Against What? Shocks and Their Costs
- Paolo Mauro, Torbjorn Becker, Jonathan Ostry, Romain Ranciere, and Olivier Jeanne
- Published Date:
- April 2007
In analyzing the types of events that countries may wish to insure against, this paper will focus on drops in per capita GDP as a practical and widely available proxy for the economic costs that countries incur when adverse disturbances occur.2 This approach requires a number of simplifications, such as abstracting from distributional effects. Nevertheless, drops in per capita GDP would seem to capture many empirically relevant features of welfare declines, as well as being a highly visible measure that the public and country authorities are worried about when they consider the costs of economic crises.
The events analyzed are defined as starting in the first year of a decline in per capita GDP and ending when per capita GDP returns to its pre-event level. Yearly losses are measured relative to pre-event GDP per capita and are cumulated over the duration of the event. (See the shaded area in Figure 2.1.)3 Two further conditions are imposed to filter out events that might result from measurement error or temporary growth spurts: (i) the duration of the event must be at least two years; and (ii) the total output loss must be at least 5 percent of pre-event per capita GDP. If an event is completely observed within the sample period, it is called a concluded event (Figure 2.1); this seems to correspond to the notion of a temporary, though costly crisis. The analysis also, however, includes ongoing events where per capita GDP has failed to recover to pre-event levels by the end of the sample period. Many ongoing events in the sample are extremely long lasting and associated with severe output losses and prolonged growth slowdowns. (Several of these started in the 1970s and 1980s and relate to emerging and developing countries that experienced major domestic crises in the wake of civil wars, oil price increases, interest rate hikes, or adverse terms of trade developments from which they had not fully recovered by the end of the sample period.)4 Finally, a subevent is defined as a new event starting before the end of a previous event.
Figure 2.1.A “Concluded” Output Event
Note: The shaded area is the cumulative output loss, and the event duration in this example is four years.
Empirical Features of Output Drops
Output events are more frequent, long-lasting, and costly for emerging market and developing countries than for advanced economies (Table 2.1).5 On average, both emerging market and developing countries have output events starting about every 16 years (or approximately twice during the three decades considered in the sample period); the events last for six years in emerging markets and twice as long in developing countries. The median cumulative output loss over the event (for concluded events) is equivalent to 15 and 38 percentage points of GDP for emerging markets and developing countries, respectively. (To illustrate, a total cumulative output loss of 15 percentage points of GDP per capita would correspond to the hypothetical case of a country whose output per capita fell by 5 percentage points, remained stable for 3 years, and then jumped back up to its initial level.) For both emerging markets and developing countries, the frequency, duration, and especially median loss of concluded output events is significantly lower than for all events. Output losses are two to three times larger (for any type of event) in developing countries than they are in emerging markets.6
|Advanced Economies||Emerging Markets||Developing Countries|
|(in percent of country-years)|
|(in percent of pre-event GDP per capita)|
|Median cumulative loss|
Taxonomy of Shocks and Their Cost
A systematic analysis of the types of shock that are associated with output drops may help countries prioritize among different forms of country insurance. The shocks analyzed here include the following:7
financial and macroeconomic—currency crises, banking crises, debt crises, and sudden stops in financial flows;8
country-specific external—terms of trade shocks and disasters;
sociopolitical—wars and political turbulence;
global—large increases in international interest rates and oil prices; and
boom-bust cycles—the end of lending booms and growth booms.9
Two simplifying assumptions bear highlighting. First, the analysis does not address the causes of shocks; in particular, it does not ask whether shocks cause declines in output rather than the other way around. (Inspection of WEO forecasts for 1990–2001, however, suggests that output events have been largely unexpected.) Second, the analysis does not seek to separate the effects of individual shocks for those events that are associated with more than one shock. (One-third of output events are associated with more than one shock; for example, various types of financial crises—currency, debt, or banking crises, or sudden stops—often occur in combination.) Moreover, the frequency and nature of the shocks may result from underlying factors—including domestic policies and institutions. Nevertheless, the associations between shocks and output drops may provide a useful gauge of the relative importance of the various types of shock for different country groups, which may help countries to find strategies for mitigating output costs through country insurance.
The importance of a given type of shock may be summarized by the expected cost of the shock or, equivalently (for cases where insurance arrangements might be conceivable), the ex ante value of insurance (analogous to the value a risk-neutral homeowner would attach to fire insurance). Three inputs are needed, and are estimated on the basis of observed frequencies in 1970–2001 (Table 2.2): (i) the probability of the shock (how often a fire starts), (ii) the conditional probability that the shock will lead to a loss in output (the likelihood the house will burn down if a fire starts), and (iii) the output cost associated with the event (the cost of rebuilding the house).10
|Unconditional Frequency of Shocks||Frequency of Output Event Conditional on Shocks||Cumulative Output Loss Conditional on Shocks|
|Advanced economies||Emerging markets||Developing countries||Advanced economies||Emerging markets||Developing countries||Advanced economies||Emerging markets||Developing countries|
|(in percent of country-years)||(in percent)||(in percent of pre-event GDP per capita)|
|Financial and macroeconomic shocks|
|Sudden stop in capital flows||5.5||11.5||15.1||3.2||9.3||1.6||7||76||10|
|Country-specific external shocks|
|Terms of trade shock||5.7||14.3||21.4||2.2||7.3||2.9||27||14||64|
|Global interest rate hike||12.5||12.5||12.5||1.0||2.8||2.4||6||19||41|
|Oil price hike||12.5||12.5||12.5||0.0||1.4||2.1||…||…||38|
|End of booms|
|End of lending boom||2.1||3.6||3.5||0.0||…||0.0||…||…||…|
|End of growth boom||0.4||1.1||1.3||…||0.0||8.8||…||…||24|
Combining these three components (Table 2.2), the expected cost seems to be substantial for several types of shock (Figure 2.2).11 For emerging markets, the largest expected cost is for financial and macroeconomic shocks—especially sudden stops (¾ of 1 percent of GDP per capita annually based on concluded events) and currency crises. For developing countries, terms of trade shocks are the most costly (amounting to 2½ percent of per capita GDP annually when concluded and ongoing events are considered), followed by debt crises and global interest rate hikes. The expected cost refers to the impact of one type of shock (regardless of whether it occurs in combination with other shocks). Thus, for example, for emerging markets, the expected cost is 1 percentage point of GDP for currency crises and 0.8 percentage points of GDP for debt crises, but the expected cost of both shocks would be less than 1.8 percentage points of GDP, because some part of the cost is double counted when currency and debt crises occur simultaneously.
Figure 2.2.Expected Cost of Shocks
Notes: The expected cost of a given shock is the expected annual loss of output associated with each type of shock. It equals the unconditional probability of the shock (left panel of Table 2.2), times the probability of an output event given the shock (middle panel), times the median cost of the event when it occurs (right panel). The large differences in expected costs between the two panels of this figure, especially for developing countries, reflect costly ongoing events, including very long-lasting events.
The main results—including on the relative importance of various shocks—hold using drops in either income or consumption per capita as the proxy for the economic costs of shocks (Becker and Mauro, 2006). Consumption drops and output events are closely associated in most countries, though the association is somewhat weaker for countries that are highly integrated into international financial markets. Closer proxies for income (such as GNP rather than GDP) might also be used but present greater difficulties with respect to data availability; in most cases, however, the differences are small.
On the one hand, a potential concern is that this approach may produce a conservative measure of the cost of output events by abstracting from trend growth during the events. On the other hand, defining the start of events and associated costs with respect to pre-event GDP might lead to an overstatement of event cost if boom-bust cycles are prevalent. Conceivably, these two biases might offset one another. As a robustness check, an alternative approach is to define the start of an event as output falling relative to a (Hodrick-Prescott-filtered) trend, and to assess the end of the event and the associated loss relative to this trend: the main results are similar.
To compute the duration and output loss associated with ongoing events (for which the end date is unknown), it is assumed that the event ends in the first year after the end of the sample period. This produces a lower bound on the durations and costs associated with these events.
Throughout this paper, advanced countries are defined as in the IMF’s World Economic Outlook (WEO) database, except for the Republic of Korea, which for the purpose of the empirical analysis is classified as emerging, rather than advanced, to capture the experience of its 1997–98 crisis; emerging market countries are countries included in either the (stock market–based) International Finance Corporation’s Major Index (2005) or JPMorgan’s EMBI Global Index (2005) (which consists of countries that issue bonds on international markets), excluding countries classified as advanced by the WEO; remaining countries are classified as developing. The exact sample varies depending on data availability for each exercise. Real GDP is measured in purchasing power parity (PPP)-adjusted dollars. The end of the sample period (2001) is determined by the availability of comparable data. All results are similar using an alternative classification of countries according to their level of financial development (high, intermediate, or low). Data sources and definitions are reported in Appendix I.
Furthermore, the relative impact on consumption is exacerbated by the degree of economic and financial development: for a given output decline, consumption falls more in developing countries than in emerging markets. This may reflect either liquidity constraints or the events’ more pronounced impact on permanent income in developing countries.
Volatility owing to abrupt changes in aid flows, an important issue for developing countries, is not considered; but see Bulíř and Hamann (2003) and Gelb and Eifert (2005). For data sources and definitions of shocks, see Appendix I.
The present study defines a sudden stop as a worsening in the financial account balance by more than 5 percentage points of GDP compared with the previous year, though the main results hold using alternative numerical thresholds. Should sudden stops be attributed to volatile supply of international flows to emerging markets, or are they caused by worsening expectations regarding a country’s economic performance? Although this question cannot be answered definitively, the list of sudden stops seems to include few, if any, instances in which the stop was clearly triggered by worsening growth expectations. For the 1990s, this was confirmed by analyzing quarterly data on financial flows, identifying the first quarter when the sudden stop began, and checking that the immediately preceding World Economic Outlook did not forecast a slowdown in economic growth for the country in question.
The end of a lending or growth boom may fail to fit the usual definition of a shock; nevertheless, this type of episode is included in the analysis to capture output declines that might be part of a boom-bust cycle.
The expected cost will be substantially lower than the ex post cost of observed output events, because the relevant probabilities are much lower than one.
Figure 2.2 is based on contemporaneous correlations between shocks and output events. Similar results are obtained using lagged shocks. The figure omits advanced countries, because the value of insurance for this segment appears to be very low. This may be due to a better diversified production structure or more resilient financial systems and institutions. An additional factor, however, may be the focus on types of shock that seem to be more relevant for emerging and developing countries.