Papers by Tadashi Kikugawa
Securities Analysts Journal, 2017
Actively-managed bond funds in Japan have one key trait in common—a tendency to overweight credit... more Actively-managed bond funds in Japan have one key trait in common—a tendency to overweight credit exposure. However, such funds do not utilize another source of returns, namely, carry-and-roll-down (CaRD) exposure in the JGB curve. This is puzzling since CaRD offers attractive risk-adjusted return and low correlation to credit and equities. This paper explores this issue further, highlighting the importance of understanding and improving factor exposure in fixed income funds.
Cobalt-blue colorant was first used in the 18th Dynasty in the New Kingdom of Egypt. The source o... more Cobalt-blue colorant was first used in the 18th Dynasty in the New Kingdom of Egypt. The source of this cobalt was cobaltiferous alum from the Western Oases of Egypt. The use of this alum, especially in glass, was suddenly limited at the end of the 18th Dynasty. There is little evidence of the production of cobaltblue glass in the Ramesside Period (the 19th-20th Dynasties) in the New Kingdom of Egypt. In this study, we brought a portable X-ray fluorescence spectrometer to two archaeological sites located in the Memphite region and used it for onsite analyses of Ramesside cobalt blue-colored glasses and faiences. This method revealed that the compositional characteristics of the cobalt-blue colorant in these Ramesside glasses and faiences is different from the colorant derived from cobaltiferous alum used in the 18th Dynasty, based on the comparison of transition metal composition and alumina content with those of the cobalt blue-colored artifacts from the 18th Dynasty. This result suggests that a new cobalt source other than cobaltiferous alum from the Western Oases was utilized in Egypt during the Ramesside Period.
Journal of Archaeological Science, 2012
The Journal of Fixed Income, Sep 1994
Conference Presentations by Tadashi Kikugawa
JSAI Special Interest Group on Business Informatics (SIG-BI #8), 2018
Standard finance theory states that returns on assets are predictable. However finding empirical ... more Standard finance theory states that returns on assets are predictable. However finding empirical evidence of predictability is statistically difficult, and data-mining for detecting more significant evidence leads to overfitting, by which it looks like significant in appearance but is senseless in fact. Particularly in recent years, an enormous amount of information or big data becomes available at low cost, and machine learning attracts an increasing interest in engineering aspects. Big data and machine learning can contribute to improvement in prediction accuracy, while they increase possibility of overfitting. This study considers data-mining with both variable selection and model selection for return predictability in the time-series. Our results demonstrate that overfitting makes large degree of influence on backtests, giving rise to specious identifications.
Uploads
Papers by Tadashi Kikugawa
Conference Presentations by Tadashi Kikugawa