Amid sweeping disruptions instigated by Covid 19 globally, the banking diaspora has its task cut out. Unlike the financial crises of yesteryears, the pandemic has imposed a unique predicament by creating an indefinite physical barrier between the banks and their customers, owing to social distancing norms. As the situation continues to take a toll on the global economies, business houses, small- and medium-sized enterprises face restricted access to liquidity, delinquency risks and aberrations in the supply chain leading to a confounding existential crisis.

The tone of corporate or commercial banking, is rather muted and dilemmatic. Having seen phenomenal growth in the last decade, this segment of banking — which counts businesses of all size and scale as customers — is battling paranoia related to stressed assets, default risks, margin squeeze and rising cost-income ratio.

A global survey conducted during the pandemic spanning more than 250 treasurers and CFOs had serious conclusions: A whopping 64% had liquidity concerns, while 17% rued external funding challenges.

The pandemic has thrust significant responsibilities on the CFO suite to ensure an undisrupted and error-free business environment addressed through a rigorous risk management framework.

As treasury gains strategic mileage with tectonic shifts in banking architecture and digital embodiment of access and privileges, it becomes imperative for the treasury teams to retain control and ensure round the clock visibility across cash flows, fund requirements, risk scenarios, business disruptions. Organizations need to become increasingly agile and resilient to contain the impact of external shocks amidst a complex intertwining of supply chains and payment systems.

Intelligent automation helmed by artificial intelligence (AI), machine learning (ML), natural language processing (NLP) and data analytics is drawing resonance from the board and senior management of corporate banks to mitigate the situation. The bankers are leaving no stones unturned to explore, test and implement viable use cases in areas such as credit risk management, know your customer (KYC), customer onboarding, payments reconciliations, document management and forecasting.

Defaults on commercial loans

Covid-19 has spurred instances of defaults on commercial loans as small businesses fail to comply with financial obligations. Drawing upon the cognitive armoury of AI and ML, banks are putting in place a robust system to assess the creditworthiness of prospects and build fraud management capabilities into their middle office.

Early warning signals are constantly modelled to enable better understanding of risk profiles and corporate volatility. Predictive algorithms help zero in on defaults and anomalies pertaining to the behaviour of payments and the associated clients, paving way for defaulter identification. Bank of America has deployed AI and NLP to churn its credit portfolio to filter out suspects, increasing the defaulter identification probability to 5.9%.

Lingering inefficiencies related to legacy systems, exception handling, multiple interfaces, currency positions and information sharing pose challenges for cross-border payments. AI and ML frameworks can enable straight-through reconciliations ensuring that discrepancies in payments information are resolved. Citi has employed solutions to ensure that heterogeneous data in payments receipts and invoices are identified and matched appropriately through ML-based pattern detection, resulting in accelerated processing.

Citi has employed solutions to ensure that heterogeneous data in payments receipts and invoices are identified and matched appropriately through ML-based pattern detection, resulting in accelerated processing.

Onboarding in corporate banking

New clients continue to experience painfully high onboarding periods in corporate banking. Leading surveys attribute five to seven days for average customer due diligence and KYC schedules that get further extended in the absence of prompt signature-verification systems.

Many banks have deployed AI technologies such as optical character recognition (OCR) to ensure the digital capture of data, categorization of unstructured elements, extraction of relevant information, derivation of inferences and going as far as validating legal agreements and performing KYC and anti-money laundering (AML) checks with varying complexity, to round off the full digital onboarding process. Standard Chartered has created an end-to-end workflow solution that uses robotics to validate documents and leaves behind an audit trail for retrospective authentications in its trade and forex (FX) platforms.

Treasury forecasting

Treasurers have struggled with accuracy in forecasting for a long time. In order to generate actionable insights in a timely manner, most of them rely on time-consuming and complex modelling as well as simulation exercises to investigate historical forecasts to be able to make reliable long-range predictions. Banks are increasingly using advanced data and analytics mechanisms to cultivate precision in forecasting. In forecasting cash positions, treasurers routinely derive insights from a collective data pool based on significant events and triggers.

Innovations in pattern recognition ably supported by AI and ML detect anomalies in cash flows as well as seasonal variations, which make for enhanced accuracy. Risk mitigation gets a shot in the arm through advanced analytics-driven models as treasurers are better prepared to deal with adverse market scenarios and fluctuations as well as hedge their multi-instrument, multi-region positions.

Kamal Misra is a Senior Director in Financial Services with Capgemini Invent. He has strong strategy and advisory experience collaborating with leading financial services firms globally.