本研究目的在於探討隱含波動度不對稱效果，並分析公司規模與財務槓桿比率對於波動度不對稱的影響。波動度不對稱效果是指負向報酬衝擊對波動度增加的影響較正向報酬衝擊大。由於過去的文獻多針對現貨與期貨價格行為上的研究，並以條件變異數衡量波動度，本研究則試著以選擇權之隱含波動度作為波動度不對稱效果的衡量基礎，希冀以隱含波動度代表未來波動度的不偏估計量，反應出投資者對於未來的預期。本研究選取29支英國的個股選擇權，並利用EGARCH(1,1)模型來探討在股票價格變動下，個股選擇權所反應出來的波動度不對稱效果，研究期間主要從2000年1月25日至2003年12月31日止。而在驗證波動度不對稱效果確實存在下，我們更進一步以Pooled OLS模型、靜態Panel Data 模型與動態Panel Data模型來探討公司規模與財務槓桿比率對隱含波動度不對稱效果之關係。
本文之主要結論如下:
１. 在29家英國樣本公司中，確實均存在隱含波動度不對稱之效果，即負向股票價格變動對隱含波動度的影響較正向股票價格變動為大。
２. 在分析公司資產規模與公司財務槓桿影響波動度不對稱效果，若只以Pooled OLS模型分析，可能產生錯誤的推論，雖然公司規模為顯著性正相關，但財務槓桿則為不顯著之負相關，其實證結果與KS不一致，並且不能支持Black(1976)之槓桿效果。
３. 為了避免使用 Pooled OLS模型產生錯誤的推論，本研究另以靜態Panel Data 模型來分析波動度不對稱程度與公司資產規模和財務槓桿之關係，對於公司資產規模因素而言，不管在公司效果(固定模式)與時間效果(隨機模式)均呈現顯著之正相關，而在同時考量公司效果及時間效果(隨機模式)下，則呈現不顯著之正相關，此結果與KS的推論一致。而對於財務槓桿而言，則只有在公司效果(固定模式)呈現顯著之正相關，在同時考量公司效果及時間效果則呈現不顯著之正相關，而若單只考慮時間效果，則係數為-0.000(不顯著)，則與KS推論不符合，且不支持Black之槓桿效果假說。
４. 為了較正確反應投資市場是有記憶性與調整性，我們另以動態Panel Data 模型來作實證，而這亦是一般較符合市場之模型，實證結果顯示不管在one-step 或 two-step 下，公司財務規模與財務槓桿確實與波動度不對稱性呈現顯著正相關，其結果與KS一致，且支持Black所提出之槓桿效果，而動態的延遲項則呈現不顯著(推測受限於樣本數過少)之負相關(係數為負，且值小於1)，若以部分調整模型之經濟意義來解釋，即調整係數值均大於1，顯示出實際市場反應出來的波動度不對稱之結果，大於投資人對於波對度不對稱情形預期的調整，這可能是選擇權市場投資人之過度反應的行為所造成，故可能使得前期項對於後期項的影響為負，但會逐漸消失。 / The purpose of the research is to discuss the asymmetric effect of volatility, and analyze firm scale and debt ratio affect the asymmetric effect of volatility. Asymmetric effect of volatility is the influence of negative return is more than positive return. Most research focus on the futures and spot goods，and takes conditional variance as volatility. We want to use IV as unbiased estimator of volatility in the future, and reflect the investor’s expectation. We chose 29 call options in English, and use EGARCH (1,1) model to explore the asymmetric effect of volatility over 2000/1/25-2003/12/31 period. After confirming the asymmetric effect of volatility, we use Pooled OLS Model, Static Panel Data Model, and Dynamic Panel Data Model to discuss the relationship between firm scale, debt ratio, and asymmetric effect.
The funding of the paper are:
(a) There is certainly the existence of asymmetric effect in 29 sample firms.
(b) Pooled OLS Model may result wrong conclusions. There is a significantly positive relationship between firm scale and asymmetric effect. And there is a less significantly negative relationship between debt ratio and asymmetric effect. The result doesn’t consist with KS, and doesn’t support Black’s leverage effect.
(c) To avoid the error from Pooled OLS Model, we use Static Panel Data Model to analyze the relationship with firm scale, debt ratio, and asymmetric effect. Asymmetric effect is significantly positively related to firm scale with single corporate effect (fixed effect) and single time effect (random effect). And asymmetric effect is less significantly positively impacted by firm scale if we chose corporate effect and time effect simultaneously. Asymmetric effect is significantly positively related to debt ratio with single corporate effect (fixed effect), and is less significantly positively related to debt ratio with corporate effect and time effect simultaneously. The coefficient is -0.000 (less significantly) of debt ratio with single time effect. The result doesn’t consist with KS, and doesn’t support Black’s leverage effect.
(d) For showing capital market’s memory and adjustability, we use Dynamic Panel Data Model to analyze the problem. Asymmetric effect is significantly positively impacted by firm scale and debt ration in one-step model and two-step model. The result consists with KS, and supports Black’s leverage effect. The coefficient of lagged term is between 0 and -1 (less significantly) may be come from the real asymmetric effect is more than investor’s expectancy, and investors my have overreaction in capital market.
| Identifer | oai:union.ndltd.org:CHENGCHI/G0923570011 |
| Creators | 林景民 |
| Publisher | 國立政治大學 |
| Source Sets | National Chengchi University Libraries |
| Language | 中文 |
| Detected Language | English |
| Type | text |
| Rights | Copyright © nccu library on behalf of the copyright holders |
Page generated in 0.0026 seconds