Comparison study using ARIMAX and VARX in cash flow forecasting

Abstract

The economic crisis caused by the Covid-19 pandemic has made it difficult to achieve sustainable economic growth, which is one of the sustainable development goals. The great pressure on the economic sector caused most companies, including the banking sector, to experience cash flow problems. The ability to maintain a balance of cash flows during the pandemic is key for the banking sector to survive. It is very important to do a study that is able to create a model to forecast the value of cash flows in the banking sector. By having a model that is able to accurately predict the value of cash flows, the problem of cash flow difficulties can be avoided early on. In this case, the banking cash flows consist of the value of cash in to the office, the value of cash out of the office, the value of cash inflows from the e-channel, and the value of cash out of the e-channel at a bank in Indonesia. In addition, modeling is carried out by taking into account the effects of daily and holiday effects as exogenous variables. The results showed that two variables had a significant effect on the value of office cash flows. Meanwhile, the value of e-channel cash flow is only influenced by daily effects. By comparing the accuracy of the Autoregressive Integrated Moving Average with Exogenous Variable (ARIMAX) method and the Vector Autoregressive with Exogenous Variable (VARX) models, it is found that the value of cash in office, cash out of office, and cash out of e-channel provide predictive value which is more accurate with the ARIMAX model. On the other hand, the VARX model is more suitable for predicting the value of e-channel cash inflows. Thus, predictions of the value of bank cash flows can be made so that existing policies can be adjusted to support the banking sector in maintaining cash flow balance during the Covid-19 pandemic.