臺灣博碩士論文加值系統

亞洲新興證券市場主要特色之一是投資者可獲取較高的證券投資報酬
率. 對新興市場投資人而言, 投資者若運用不同的國際平均數- 變異數理
論實證模型來從事證券投資, 是否含有Frankel and Engel(1984) 所謂投
資者本風險態度考量而導致Uppal(1992) 所謂資產偏差現象? 而此是值得
探究的主題, 由於以往學者未曾做過在新興市場資產偏差研究, 故引發本
文研究之動機. 本研究的目的是以美國與亞洲新興證券市場的國際投
資組合為主體, 來探討因不同實證模型是否造成當地資產偏差現象? 另探
討風險趨避態度高低是否對國際證券投資具有影響性效果? 研究對象包括
美國與亞洲新興市場中臺灣. 香港. 韓國與新加坡等五個國家, 研究時間
範圍是1990年1月至1996年6 月止. 本研究以Glassman and
Riddick(1996)國際平均數- 變異數理論模型為實證模型, 以圖形與迴歸
分析方法來比較驗證一般模型與三個簡化實證模型間關係. 主要研究結論
如下: 在相同風險態度下, 實證結果顯示此三個實證簡化模型會異於
一般實證模型, 具有資產偏差現象. 若以預期報酬平均數, 其中以PPP 簡
化實證模型會具有誇大資產偏差現象. 無二階簡化實證模型則無定論, 名
目化簡化實證模型則無誇大資產偏差效果. 此論點與Glassman and
Riddick(1996)結論一致. 在不同風險趨避程度下, 實證結果顯示除名
目化實證簡化模型投資權數不變外, 運用不同實證簡化模型, 大致上與一
般實證模型不同, 對部份國家則預期會減少投資, 另對其它國家則預期會
增加投資, 產生不一致現象, 造資成資金移轉, 故風險趨避態度對投資組
合權數具有影響性效果. 此論點與Upall(1992) 與Frankel and
Engel(1984) 相同.

One of the characteristics of Asian Emerging Stock Markets
has ahigher rate of return. Investors using different
international mean-variance models and risk attitude
considerations may be resulted in the home asset bias. In this
paper we show how different empirical models contribute to this
home asset bias. We also show how differentlevels of risk
avesion contribute to this home asset bias. The samplecontains
five countries-the U.S., Taiwan, Hong Kong, Korea, and
Singapore. The observation sample period is from 1990/1 to
1996/6. In this paperwe use graphical comparisons and
statistical comparisons to test our proposition by comparing the
portfolio weights produced by Glassman andRiddick(1996) general
and simplified international mean-variance models.Our main
findings can be summarized as follows: Under the same level
of risk aversion, all three models produce port-folio weights
that differ from general model weights in a statistically
significant way and in a graphical comparison way. The
assumption that PPP holds consistently exaggerates the degree of
the home asset bias, andin some cases the no second order terms
assumption does as well. Under different levels of risk
aversion, the conclusions regarding theeffects of the
simplifying assumptions are robust to different levels of risk
aversion. At the same time, we find that there is a fund
removing among thesecountries. So different levels of risk
aversion had significantly impacton portfolio weights.