Stress tests have become an indispensable tool in risk management. They enable financial institutions to assess their resilience in the face of unfavorable scenarios. Their importance continues to grow in an increasingly complex and uncertain economic and geopolitical environment. 

These stress tests can be defined by individual institutions or by regulators. Results of the Fed's stress tests were made public at the end of June 2024, and the EBA has just launched a consultation with the banking industry on the methodologies for the 2025 stress tests exercise.

Types of Stress Tests

There are several types of stress tests, each with its own specific features and applications:

Historical stress tests

An historical stress test is a valuation method that uses past data and events to simulate the impact of previous financial or economic crises on the institution's current situation. This type of stress test enables us to understand how specific historical events would affect the institution if they occurred again today.

The components of an historical stress test are as follows: 

• Selection of historical events: Significant past events are chosen as scenarios. Examples include the 2008 financial crisis and the 2010 European sovereign debt crisis. These events can go back a long way, such as the bursting of the internet bubble in 2000 or the crash of 1987.
• Analysis of past data: economic and financial data associated with these events are collected and analyzed. This includes variables such as major market movements (interest rates, exchange rates, stock market indices, commodities, etc.), default data on credit portfolios, and behavior on real estate deposits and loans. One of the difficulties lies in estimating and applying these variations to the current parameters.
• Simulation of impacts: The effects of these variations are simulated on the balance sheets of financial institutions. The results are used to project impacts on financial performance indicators such as financial results, capital ratios, credit losses, and liquidity.

Hypothetical stress tests 

Unlike historical stress tests, which are based on past events, hypothetical stress tests are built on possible but as yet unrealized scenarios. For example: a global recession, a commercial real estate crisis, the exit of a country from the eurozone, a massive terrorist attack  (dirty nuclear bomb…) , the default of a GSIB, an oil supply shortage for developed countries, etc.

The components of a hypothetical stress test are as follows:

• Scenario definition: The economic and risk management teams work together to precisely define these scenarios and their macroeconomic consequences.
• Parameter breakdown: the need to detail the consequences of these scenarios on market, credit, and liquidity parameters. Given the scope of the parameters to be defined, this can be a complex and resource-intensive task. ‍
• Impact simulations: as with historical scenarios, we need to analyze the impact of these scenarios on financial results, capital ratios, credit losses, and liquidity.

Specific scenarios

These scenarios can be either historical or hypothetical, but only within a particular "specific" perimeter of the institution. For example: a fall in dividends, a de-peg between two currencies, a short squeeze on a stock or commodity.

How to apply stress testing outputs to risk management

Once the stess test scenarios have been defined and produced, all that remains is to analyze their results and variations. A solution like Opensee offers many advantages. Opensee is already used by numerous financial institutions on both the sell-side and buy-side for stress test analysis. Benefits include:

• Complete scalability in terms of storage
• Enabling access to the most granular data and unlimited historical data 
• Ability to carry out adjustments and simulations while keeping all audit trails
• AI tools for analyzing results

Let's illustrate this need in different areas:

Market risk

To fully understand which instruments have the greatest positive or negative contribution to the various stress tests, it is necessary to have access to data at the most granular level (portfolio, instrument). Opensee has all of this data at its disposal, and its AI modules can be used to perform analyses explaining, for example, the origin of variations between two dates, and the instruments with the greatest contributions. 

For banks, such an analysis has been used to identify the sensitivity to stress tests of "Krach Puts" products; these instruments hedge a portfolio if the underlying index moves in one day above a predefined level. If the underlying index exceeds this level, the stress test will be strongly impacted, but if it ends just below this level, the contribution of these products will be nil. 

Stress tests will therefore vary considerably if these levels are reached or not. Analysis at the most granular level is essential to understand this mechanism and has led some institutions to limit this type of product.

Counterparty risk on market transactions

One of the best ways of assessing risk for a financial institution is to simulate a large number of "specific" stress tests on the portfolio of transactions carried out with the counterparty. 

Based on this data, and thanks to its AI modules, Opensee identifies the riskiest transactions for the institution and proposes a dashboard with the type of transactions to be limited and favored to reduce these risks. Thanks to these simulation functions, the financial institution can immediately see the impact of potential new transactions.

Reverse stress testing

A new type of stress test is emerging: reverse stress testing. This is an approach in which the institution examines possible worst-case scenarios, both in terms of risk of loss and impact on the income statement. Although often based on the expertise of professionals within the institution, this method raises the question of the sufficiency of expert judgments alone. The integration of artificial intelligence enables the Risk teams to improve this analysis by identifying areas for attention, taking into account all the stress test results available.

Once again, Opensee's granular data access and AI modules make this type of analysis a breeze.

‍

Stress testing is essential for navigating the modern financial landscape. By understanding and exploiting the different types of stress tests and having access to accurate, real-time data, financial institutions can better anticipate and manage potential risks, ensuring their stability and resilience in the face of crises.

The ability to understand and explain the nature and variations in stress test results is crucial. This highly complex process is greatly facilitated by a solution like Opensee, which is able to produce all the necessary analyses using the most granular information. Reach out to us to learn more.