Bayesian Mixture Laplace Autoregressive Modeling to Estimate Value-at-Risk in E-Commerce Stocks

Abstract

The e-commerce business is growing in Indonesia. However, there are not many e-commerce companies that trade their shares on the Exchange. Since the emergence of startups in Indonesia, including the emergence of four technology companies with unicorn status, only three have registered. These companies, namely PT. M-Cash Integration Tbk (MCAS), PT. Kioson Commercial Indonesia Tbk (KIOS) and PT. NFC Indonesia Tbk (NFCX), are listed on the development board. To enliven the stock market of e-commerce companies, it is necessary to analyze the risk value of daily closing price data as an investment consideration. This study aims to provide knowledge to investors who want to invest in e-commerce companies about possible risks. This research was conducted on two e-commerce companies that were first listed on the IDX, namely MCAS and KIOS. The Value-at-Risk (VaR) method with the Bayesian Mixture Laplace Autoregressive (MLAR) approach is used to determine the risk value of the company's shares. Accurate VaR based on backtesting on each company's shares has been obtained at 5%-quantile, 0.0384 for MCAS, and 0.0622 for KIOS in 1-day horizon investment.