IMPLEMENTATION AND COMPARISON OF A NEW APPROACH OF VALUATION: CONTINUOUS TIME ABNORMAL EARNINGS MODEL FOR MAXIMIZING RETURN ON INVESTED CAPITAL

• Main Author: (NIM : 29107371); Pembimbing: Ir. A. Imam Istiyanto, MBA, STEPHEN
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/17305
• Institution: Institut Teknologi Bandung

Stock valuations have a very important role both in the investment process and corporate decision. For investors, it is the key factor for deciding investment portfolio. For corporations, valuation determined many corporate decisions, including public offering and merger acquisition. Decisions that made based on good valuation will lead to benefits for investors and corporations in form of high rate of return and also optimization in resource for the growth of the business. <br /> <br /> <br /> Supporting the needs of better valuation tools, this research implement a new valuation model and further develop with other theory and research to produce a good valuation model and proven. The focus of this valuation: Continuous Time Abnormal Earnings Model, developed by Yuan and Luan in 2009 in the Chinese capital market. This model is based on the theory of abnormal earnings valuation, transformed by mathematical method into continuous time model. This model represented the performance of the companies compared to their annual cost of capital. <br /> <br /> <br /> The result of this transformation reducing bias that usually occurs from factors that hardly predictable, <br /> <br /> <br /> such as future cost of capital and constant growth. This become possible since the model used historical <br /> <br /> <br /> performance compared to the macroeconomic condition. The result of this implementation later compared with abnormal earnings model, multiples, dividend discount model, and discounted cash flow. Each method became the basic of portfolio development. To prove that a valuation model is better, is only by return on invested capital it can produce, it is because all valuation tools is not accurate absolutely, but it have relative precision. <br /> <br /> <br /> Through implementation and comparison, the research showed that the error from prediction with continuous time abnormal earnings model against its real price was the lowest with 21.3% log error and 25% percentage error. Where the others result in 21.4%-91.73% log error and 26.4%-223.96% percentage error. Aligned with these results, when it was simulated as simple portfolio investment in 2003-2005, this method showed the highest return, 60.88% annually with geometric mean return <br /> <br /> <br /> calculation, when abnormal earnings, price earnings ratio, price book value, DDM , and DCF each results 40.44%, 56.72%, 33.52%, 51.63% and 42.23% in annual return. The result produced by continuous time abnormal earnings model more even better than IHSG, LQ45, and samples performance at 37.67%, 36.78%, and 38.92%. These results conclude that continuous time abnormal earnings model is the most profitable tools and able to measure the intrinsic value of securities relatively better. With further research and development in wider scope and better prediction in market value of net assets, security companies will have better investment products with better performance, investors will have higher profits, and also improving effectiveness in many financial process in corporations.