Risk management in Treasury:

We have already looked at the use of AI in treasury several times in the Corporate Treasury Newsletter, and related articles are linked accordingly.

In times of increasing economic uncertainty, the relevance of risk management is constantly growing. In particular, efficient financial risk management can be a decisive factor in securing the continued existence of a company by identifying financial risks at an early stage and taking appropriate countermeasures.

In an almost completely globalised world and a diverse intermeshing of effects and counter-effects, the marginalising phrase "A sack of rice has fallen over in China" is increasingly losing its real economic significance. This became clear, for example, with the accident of a single ship, the "Ever Given", which blocked the Suez Canal for six days and thus had a significant impact on the global economy. Even though this example is an operational risk that is not the focus of financial risk management, its consequences have a direct impact on financial risk management. In this case, the consequences had a direct impact on the "classic" risks of financial risk management such as market risk, credit risk and liquidity risk.

The ongoing development of technologies, in particular artificial intelligence (AI), is opening up new opportunities to identify these correlations more quickly, making it easier to assess and manage the resulting risks. AI technologies are transforming treasury and risk management in many ways and offer companies the opportunity to optimise processes and make them future-proof.

AI offers significant potential for financial risk management, particularly in the following areas:

1. Identification of risks

AI can and should analyse large amounts of data from various sources in order to identify risks more quickly and precisely. By using machine learning and data mining, patterns and anomalies in the data that indicate potential risks can be recognised. This enables more objective, faster and more accurate risk identification, which traditional methods are often unable to achieve. This applies in particular to risks that are hidden behind complex correlations.

• Credit risks: AI can help to assess the creditworthiness of business partners and customers by analysing historical payment data, socio-demographic information and other relevant data.
• Market risks: AI models can analyse market trends and behaviour to identify potential market risks such as price volatility and market liquidity.
• Fraud risks: By analysing transaction patterns, AI can detect unusual activities that indicate fraud or other unethical behaviour.
• Operational risks: AI can help to identify weak points in processes by recognising patterns in data that indicate inefficient or error-prone processes.
• Liquidity risks: AI can analyse cash flows and liquidity requirements in order to predict potential liquidity bottlenecks.

2. Assessment and prediction of risks

Another significant advantage of AI in financial risk management lies in its ability to predict potential future risks and assess the probability of occurrence and the extent of the potential consequences for a company. Predictive analytics uses historical data from various sources to forecast future events with potential risks. This enables companies to take proactive measures before the risk materialises. Risk reduction measures can also be implemented and reviewed in a more targeted manner based on the forecasts.

3. Automation and increased efficiency

AI can also help to increase efficiency in risk management by automating routine and time-consuming tasks. This allows risk managers to focus on more complex and strategic aspects of risk. For example, automating data collection and analysis can help risk managers to adapt more quickly to changing conditions and make informed decisions.

4. Compliance and regulatory requirements

Compliance with legal, regulatory and financing requirements is an essential part of financial risk management. AI can help minimise compliance risks by continuously monitoring compliance and automatically generating reports. This reduces the risk of non-compliance and the associated financial and reputational damage.

Although the use of artificial intelligence in financial risk management offers considerable advantages, it also brings with it specific challenges. These challenges also highlight the limits of the use of AI.  

1. Data quality and availability

AI systems require large amounts of high quality data to be effective. However, it can be difficult to gain access to clean, well-structured and comprehensive data. Incomplete or erroneous data can lead to inaccurate predictions and analyses. Any use of AI models is therefore limited to data that is available in sufficient quality. The costs of increasing quality and obtaining data should be weighed up against the potential benefits before use.

2. Complexity of the models and explainability

AI models, especially those based on deep learning, can be extremely complex. This complexity not only makes it difficult to understand how decisions are made, but can also complicate the maintenance and updating of the models. The decisions made by AI systems are often not transparent, which makes them difficult to understand. In the area of risk management, however, it is crucial that the decision-making processes are clear and comprehensible. If the causes and characteristics of a risk cannot be explained, it becomes difficult to react in a targeted manner.  

3. Overfitting

AI models can be overly "trained" on the nuances of the training data and therefore perform poorly with new or changing data. This is particularly problematic in financial risk management, as market conditions can fluctuate rapidly.

4. Dependence on the technology

An over-reliance on AI systems can lead to human experts being less involved in the decision-making processes, which increases the risk of errors that can arise due to a lack of human judgement. This can be particularly problematic in the case of risks that are subject to human irrationality, as this is where the mechanics used reach their limits. This includes geopolitical risks, for example.

5. Data protection and data ethics

As the analyses in financial risk management are always based on highly sensitive data, it must be ensured that the data used is protected against unauthorised access at all times. It must be ensured that the IT systems and applications used fulfil the compliance requirements as early as the consolidation and storage of the data used. If personal data is included in the AI in addition to the company's own data, it must also be ensured that the requirements of the GDPR are met. It is also important to check whether the AI fulfils the ethical requirements of EU AI law and to what extent there is a reporting obligation.

The use of AI in financial risk management is promising, as the relevant risks are subject to certain logics and a good data basis is often available thanks to modern treasury management systems. It is therefore possible to demonstrate a good business case for the use of AI in treasury. However, AI cannot replace the human risk manager - nor should this be the goal. Rather, the aim should be to provide risk managers with a tool that enables them to hand over routine activities, develop scenarios and run through them. An efficient division of labour between humans and AI enables better results to be generated and also allows correlations to be considered that were previously ignored due to time restrictions.

Source: KPMG Corporate Treasury News, Issue 147, September
Authors:
Nils Bothe, Partner, Finance and Treasury Management, Corporate Treasury Advisory, KPMG AG
Tobias Riehle, Manager, Finance and Treasury Management, Corporate Treasury Advisory, KPMG AG