Stress Tests



In this station, we will discuss how to determine the sensitivity of projected debt burden indicators to changes in the assumptions. To assess this, the LIC DSF calculates the impact of temporary shocks on the evolution of debt burden indicators in both the external and public DSAs.

Why does the DSF need stress scenarios?

  • The baseline scenario should represent the most realistic macroeconomic and debt portfolio assumptions based on the most current and best available knowledge.
  • However, future realizations may differ from baseline projections due to the uncertainty surrounding the baseline scenario.
  • One way to gauge the impact of such uncertainty is through the use of stress scenarios.

Stress Tests and Risk Rating


Recall: Risk ratings are based on several debt ratios breaching their respective thresholds. Breaches of thresholds under stress tests indicate debt vulnerabilities to shocks and affect a country's risk rating.

The most extreme stress test informs the calculation of the mechanical risk signal (defined as the test that yields the highest level of debt on or before the tenth year of projection, which has a greatest chance to breach the threshold among stress tests). In addition, where the test, rather than the baseline, leads to a breach of any of the thresholds, the mechanical risk signal switches from “low” to “moderate".

Stress Tests in the LIC DSF

The LIC DSF template automatically applies stress tests to gauge the sensitivity of the projected debt burden indicators to changes in the underlying assumptions. It calculates the impact of temporary shocks on the evolution of debt burden indicators in both the external and public DSAs.

Types of Stress Tests

  1. Standardized Stress Tests

  2. Tailored Stress Tests

  3. Fully Customized Stress Tests


    

Standardized Stress Tests



Standardized Stress Tests are universally applied to both public and external DSAs in all LIC DSF countries. They consist of the historical scenario, as well as various scenarios with temporary shocks to real growth, primary balance, exports, other flows (current transfers, foreign direct investment), and domestic currency depreciation. Click on each shock for more details.


Shock Scenarios

Shock Design

The following macroeconomic variables are set to their respective 10-year historical averages: real GDP growth, primary balance-to-GDP ratio, GDP deflator, non-interest current account, and net FDI flows. This scenario is used as a realism check for the baseline scenario and does not inform mechanical risk signals.

Shock Design

In the second and third projection periods, real GDP growth set to its historical average minus one standard deviation, or the baseline projection minus one standard deviation, whichever is lower.

Shock Interactions among Variables

  • Inflation decreases with assumed elasticity to growth equal to 0.6.
  • Primary balance deteriorates due to the revenue-to-GDP ratio remaining the same as under the baseline, while the ratio of non-interest expenditures-to-GDP increases to maintain the baseline level of spending.

Shock Design

In the second and third projection periods, primary balance-to-GDP ratio set to its historical average minus one standard deviation, or the baseline projection minus one standard deviation, whichever is lower.

Shock Interactions among Variables

  • For a LIC with domestic market financing, domestic borrowing costs rise by 25 basis points for each 1 percent of GDP in primary balance deterioration.
  • For market-access LICs, external commercial borrowing costs increase by 100 basis points for each 1 percent of GDP in primary balance deterioration, or by 400 basis points, whichever is lower.

Shock Design

In the second and third projecion periods, nominal export growth (in USD) is set to its historical average minus one standard deviation, or the baseline projection minus one standard deviation, whichever is lower.

Shock Interactions among Variables

  • Real GDP growth rate is lowered with an elasticity to exports of 0.8

Shock Design

In the second and third projection periods, current transfers-to-GDP and FDI-to-GDP ratios are set to their historical average minus one standard deviation, or the baseline projection minus one standard deviation, whichever is lower.

Shock Interactions among Variables

  • No interactions.

Shock Design

In the second year of projections, there is a one-time depreciation of domestic currency by 30 percent, or the size needed to close the estimated real exchange rate overvaluation gap, whichever is larger.

Shock Interactions among Variables

  • Real net exports, as a percent of GDP, increase starting in the year following the shock assuming elasticity to real depreciation of 0.15.
  • In the year of the shock, pass-through to inflation is set to 0.3.

Shock Design

Apply all individual shocks above(except for the Historical Scenario) at half of each shock magnitude.

Shock Interactions among Variables

  • All the interactions from each individual shock scenarios.



    

Tailored Stress Tests



Tailored Stress Tests are related to so-called “tail” risks. They only apply to a country if relevant and likely, and they have implications for a country's risk rating. The exception is the Contingent Liabilities stress test, which applies to all countries.

1. Contingent Liabilities

2. Natural Disasters

3. Volatility of Commodity Exports

4. Market-Financing Shocks

For each tailored stress scenario, default parameters of the stress scenario are based on cross-country averages but are expected to be customized based on country-specific characteristics.

Contingent Liability Stress Test

The Contingent Liability (CL) Stress Test applies to all countries.

The Contingent Liabilities Tailored Stress Test is built to take into account elements of the broader public sector that were missing from the debt definition in the DSA.

The shock has a minimum starting value and a tailored value, depending on the vulnerabilities outside the public sector that are already covered by the DSA.

Shock Trigger Shock Size and Interactions
Financial sector (applicable to all)  One-off shock of 5 percent of GDP + tailored element (user specified) 
Incomplete coverage of the general government  Shock as defined by user (based on country history, and missing elements of general government) 
PPP capital stock above 3 percent of GDP  Debt increases by 35 percent of PPP capital stock in second year of projection 
Unaccounted or partially accounted SOEs' guaranteed external debt that has not been accounted for in the general government   2 percent of GDP increase in public debt from second year of projection 


Contingent Liability Shock

Users can customize the contingent liability shock by filling in the following table in the template:


Chosen coverage of public sector debt The central government plus social security, central bank, government-guaranteed debt
A Other element of the general government not covered above (default = 0 percent of GDP)
B SOE debt, guaranteed and not guaranteed by the government (default = 2 percent of GDP)
C Financial markets (default = 5 percent of GDP, which is the minimum)
D PPP - calculated size of the shock in percent of GDP (D = E*F)
E PPP - capital stock from the World Bank data base in percent of GDP
F PPP - size of the shock (default = 35 percent of the PPP capital stock)
G Total size of the contingent liability shock (G = A+B+C+D)

The default shock of 2 percent of GDP will be triggered for countries whose government-guaranteed debt is not fully captured under the country's public debt definition. If it is already included in the government debt, and risks associated with SOE debt not guaranteed by the government are assessed to be negligible, it may be reduced to 0 percent.

  Considerations

  • Are there known additional contingent risks in the financial sector? For example, those identified in an asset quality review?
  • Do the missing elements of general government have the potential to generate debt vulnerabilities? For example, do budgetary funds have a history of government bailouts?
  • Does a specific SOE pose a risk to debt sustainability? Can it cover its gross financing needs over the next several years?
  • Is the PPP capital stock larger than assumed? How risky is the PPP to the government? Are the likely costs greater than 35 percent of the nominal value? PPPs are typically related to infrastructure, but they could also be projects in other sectors of the economy, such as in the healthcare system.
  • Also consider any other shock relevant to your country, such as those based on specific historical circumstances or potential missing elements of the general government.

Natural Disasters

Shock Trigger Shock Size and Interactions
Small states vulnerable to natural disasters One-off shock to public debt of 10 percent of GDP in the second year of projection
Countries that meet the frequency (2 disasters every 3 years) and economic loss (greater than 5 percent of GDP/year) criteria Real GDP growth lowered by 1.5 percentage points in the year of the shock
Exports are lowered by 3.5 percentage points in the year of the shock

  Considerations

  • Does the baseline macro framework already assume an average impact from natural disasters? Ensure that double-counting is avoided.
  • Do you have empirical evidence that the size of the shock should differ from the default setting of 10 percent of GDP? Would that impact the ability to cover your gross financing needs over the next several years?
  • Do you have empirical evidence that the default setting of export and GDP growth interactions should be changed?

Volatility of Commodity Exports

Shock Trigger Shock Size and Interactions
Countries with commodity exports exceed 50 percent of exports of goods and services Commodity exports prices fall by 1 standard deviation, and gap closes over 6 years. Interactions include reduction in:
Commodity exports = fuel + non-fuel exports real GDP growth by 0.5 percentage points from the baseline
Based on the last three years shares (of fuel + nonfuel commodity exports) provided by you fiscal revenues-to-GDP by 0.75 percentage points from the baseline
GDP deflator by the impact of the commodity price gap in the first year of the shock, converging to the baseline in 6 years


  Considerations

This test can be adapted to take into account a contemporaneous fall in the price of a key import and can be netted out for the purposes of the shock as a “mitigating factor” in the template. For example, exporting crude oil and importing refined oil.



  • Is there a reason to model a bigger/smaller shock to commodity prices?
  • What is the historical evidence of the commodity/country/region during the commodity cycle? Is it different from the default setting?
  • Do you have empirical evidence that the default setting of export and GDP growth interactions should be changed?

External Market Financing Shock

For countries with market access, this scenario helps focus discussions on potential market stresses during upcoming rollovers.

Shock Trigger Shock Size and Interactions
Countries with at least one of the following: A combination shock, which includes:
Have outstanding Eurobonds, and/or have access to the international financial markets on a durable and substantial basis An increase in borrowing costs of market-based financing of 400 bps (sustained for 3 years) and shortening to 5-year maturity or shortening by 2/3 of the original maturity that is less than 5 years for newly issued market debt; and
One-off nominal depreciation equivalent to 15 percent

  Considerations

  • Is there empirical country-specific evidence suggesting the response to a market financing shock in borrowing costs/depreciation/maturity shortening may be different?

    

Customized Stress Tests



In addition, fully-customized stress test scenarios are relevant for specific risks not covered by the template.

Fully-Customized Scenarios

Examples include, but are not limited to, the following scenarios:

  • Idiosyncratic risks, such as an epidemic;
  • Large delays in investment projects; and
  • Policy slippage where different outcomes lead to very different debt paths.

Customized scenarios can also be used to investigate alternative financing strategies and their implications without affecting the risk rating, such as a scenario for achieving sustainable development goals, a debt restructuring scenario, or a higher grant scenario.

While fully flexible in their design, customized scenarios inform the risk ratings in the same way that other stress tests do.

There are two options of incorporating customized stress tests into the LIC DSF:

1. Use the “Customized Scenario” sheets in the template to design a stress test.
2. Create an alternative baseline and enter them in separate template from the original baseline.


Takeaways for Station 5

  • Stress test scenarios are critical for understanding the nature of risks to which a LIC is vulnerable and to inform the risk rating.
  • Tailored stress tests help provide insights on specific risks relevant for some LICs.


Travel to Station 6 for the next part of your journey



50% Complete
50% More To Go (success)