According to the Association for Financial Professionals 2020 Strategic Role of Treasury Survey Report, 36% of companies believe that Technology and Automation will enable greater focus on higher value-added work in the next three years. Emerging technologies such as Robotic Process Automation (RPA), Artificial Intelligence (AI) and Machine Learning (ML) promise to streamline operations, increase efficiency, reduce costs and improve decision-making. However, Corporate Treasury department has lagged behind others in using cutting-edge technologies for automation. Often it is due to relatively smaller size of the Treasury team in a larger Finance organization, lack of understanding of how such technologies can help Treasurers and the high-cost of such technologies.

Though RPA technologies are getting cheaper and cheaper over time, ROI may still be difficult to achieve for Treasurers. Licensing costs of an RPA bot alone can range from $5,000 to $20,000 per year. In addition, the implementation costs, which include costs such as hiring RPA trained SMEs, consulting cost, infrastructure cost, 3rd party integration expenses, initial configuration and support costs etc., can easily be up to 20 times the licensing expense. This high cost coupled with steep learning curve for the team embarking on an RPA journey, can easily discourage the Treasurers from leveraging RPA. However, there are lower-cost alternatives to expensive RPA platform in automating Treasury and related processes.

Gap left by TMS, ERP and Bank Portal

Many treasury teams use Treasury Management Systems (TMS) or Enterprise Resource Planning (ERP) for automating Treasury processes. Though both ERP and TMS are very useful tools, they don’t automate day-to-day Treasury processes fully. For example, Treasury and payment processing team may be forced to use bank portal to manage ACH positive pay, stop payment, ACH debit fraud etc. since TMS/ERP typically fail to automate these aspects of payment automation. Similarly, the vast majority of TMS/ERP cannot manage bank signatories electronically, download lockbox data, download monthly bank statements, or ACH addenda – the list of ERP/TMS shortcomings go on and on when it comes to these tools being a realistic alternative to the bank portals. It is no surprise that treasury teams continue to perform several tasks directly on the bank portals. This leaves many Treasury teams disillusioned with TMS, looking for a better and more complete solution. ERPs often lag behind TMS in automating treasury processes, though some of the ERPs are making an effort to make a dent into the TMS market. Overall, ERP and TMS still leave a lot to be desired by Corporate Treasury when it comes to automation.

The lack of automation negatively affects not only the Treasury team but also other Finance teams such as Accounts Payable, Accounting and Cash Applications, who resort to manual processes for downloading and processing data from the banks. For example, a Cash Apps analyst may spend hours manually reconciling cash receipts with open AR and remittance information. Similarly, an AP Analyst may regularly logon to the bank portal to issue stop payments on checks or manage check/ACH positive pay.  Corporate Accounting teams may be downloading monthly bank statements manually to reconcile cash as part of monthly closing. An auditor might have to ask for email proofs of payment approvals since TMS or ERP fail to implement Corporate Delegation of Authority in payment approval workflow.

Automate using banking API and Browser automation

Some new developments in banking technology and automation have already begun to address some of this unmet need. The solution space for automating data processing from banks, merchant card processors and FinTechs is getting more and more interesting. Banks and other Financial Institutions have started offering APIs to conduct various banking services. In fact, Goldman Sachs has been building a digital banking infrastructure from the ground up called Marcus. Similarly, other major banks have been expanding the range of services that can be accessed via APIs.

Bank APIs can be a much more effective and cheaper alternative to TMS. TMS works on a file based communication with the bank. The files are exchanged based on secured internet protocols such as SFTP or AS2, which needs significant involvement from the IT team in the setup and maintenance. In addition, the data exchange between the bank and TMS/ERP is triggered on predetermined scheduling intervals. This limits Treasury team’s ability to view the banking data in real-time. In contrast, an API based communication is designed to be on-demand and provides results in almost real-time. Bank APIs are also much cheaper than SFTP based integration with the banks, at least for now. It is no surprise that some forward-looking Treasurers have started to question whether they really need a TMS and are planning to use bank APIs to perform bank transactions instead. Several FinTechs have already started taking advantage of the bank APIs to connect with the banks in downloading data and performing banking transactions on behalf of the clients. 

In addition to interfacing with the banks via the APIs, one can use browser based automation to mimic a user action in downloading data and performing banking transactions on the bank portal. At Crestfin, we built a solution using Python programming language that automatically downloaded bank balances and transaction on a daily basis and on demand. Browser automation allowed us to download any information that is available on the bank portal not just what the bank could offer via SFTP or APIs. For example, we were downloading hundreds of electronic lockbox images daily from the bank portal and storing them on SharePoint for ease of access. Such automation helps not only the Cash Application team in their day-to-day work but also the Treasurer in saving bank fees associated with lockbox-file transmission. By downloading and storing the data on the Corporate servers, we could save thousands of dollars in bank fees alone since this obviated the need to use bank’s expensive data archival service as well as the bank fees associated with file transmission. Browser automation can fully replace bank API or complement it based on how much automation a Corporate Treasurer wants to achieve. Using browser automation, we could build a better notification system than what the bank or the TMS/ERP offered.

RPA tools can certainly do browser-based automation too. However, it takes a huge effort to create even a simple solution on the RPA platform. At Crestfin, we hired some experienced Python developers who created much more sophisticated automation solutions at a fraction of the cost using open-source software libraries. With this solution, there is no need to pay for either the RPA bot licensing fees or the expensive RPA consultant fees. Similarly, we are building solutions to download credit card fee data from Payment Processors which can then be married with customer receipts to provide interesting insights into customer payment behavior and expensive credit card fees.

The examples provided above are just the tip of the iceberg of what is possible in building better automation and data analytics for Treasury and Corporate Finance. Unfortunately, there is no product in the market yet that can put all these data and processing in a single package to automate Corporate Treasury. However, the good news is that low-cost tools, technology and expertise are already available for Corporate Treasurers and CFOs to leverage today.