How to minimise the risk of bank runs in times of social media

Liquidity risks are reaching new dimensions for the banking sector in light of the digital transformation. Customers can withdraw their deposits via app much faster than before. Rumours that spread like wildfire on social media can therefore lead to bank runs within a very short space of time. The collapse of several financial institutions in 2023 showed just how dangerous this development can be.

One thing is clear: technological progress is making risk management more complex. It is important to make strategic adjustments and adequately incorporate social media as an influencing factor for corporate reputation on the market. In the white paper "Bank runs and liquidity contingency planning in the era of social media", we provide a compact and precise overview of the aspects that the financial sector should focus on for optimised, continuous liquidity protection in emergency situations.

According to our experts, three basic steps are essential to prevent a liquidity crisis fuelled by social media posts that can quickly escalate. The core content of the white paper at a glance:

Comprehensive social media screening makes it possible to identify potential reputational damage at an early stage. Technological solutions such as machine learning or predictive analytics can be used to automate and accelerate the collection and forecasting quality of social media activities. When implementing relevant indicators, however, it should be noted that the high volatility of the number of posts on social media platforms and the lack of historical data can make analysis difficult.

In the white paper, we use a prototype KPMG solution to show how efficient social media monitoring can work in practice.

As soon as suitable indicators have been established, defined threshold values can help to identify emerging external reputational risks and gradually take countermeasures. It is essential that the data collected is available to those responsible for liquidity management and that corporate communication is integrated into the newly designed processes. Reputation monitoring should be carried out at high frequency and automatically. In addition, escalation levels should be determined in advance and communicated to all stakeholders.

If warning indicators are recognised or threshold values for a crisis-like development are reached, it should not be necessary to identify possible countermeasures first. Instead, it is important to initiate stabilisation processes for the liquidity situation as quickly as possible, flanked by a content offensive on digital platforms and in "traditional" media. The aim is to defuse the situation, limit damage and, if necessary, counter online rumours with facts or correct misinformation. Involving external partners, such as marketing agencies, can be helpful here. The response in an emergency should also be tested regularly and with the involvement of all stakeholders.