財務金融研究

Abstract

本論文以兩個財務實證研究為主。第一個主題為股票指數期貨合約的轉倉效果分析，第二個主題是資本市場中的地雷股分析。 針對股票指數期貨合約的轉倉效果分析，本文提出一個新方法來衡量股票指數期貨合約的轉倉報酬，並分析轉倉效果。由於期貨合約到期時若轉倉進入下一個合約，使用傳統的轉倉報酬衡量方法，並未考慮價格跳動、及合約不一致的問題，有必要加以修正。本文所提出的新的衡量方式，是在轉倉日計算報酬時，以將轉入的合約價，取代目前將到期的合約價，並將總報酬拆解為轉倉效果與資本利得兩部份來分析。在轉倉日的選擇上，本文也比較了最後交割日、到期日前七個交易日、及到期當月第一個交易日的轉倉效果，同時也考慮了理論與實際價格的差異，及觀察便利殖利率的變化。本文所觀測的市場為美國 S&P500、德國 DAX、及台灣 TAIEX 股票指數期貨合約，發現在最後交割日轉倉的收益最大，並由價格到期收斂的路徑，也發現這三個市場有很顯著的到期日效果。 而本文第二個主題為財務危機公司之研究。資本市場中，財務陷入危機的公司，猶如戰場上的地雷一般，不容易預測，但會造成極大傷害，所以有必要針對資本市場中的地雷股進行更進一步的分析與預測。本文以違約距離模型（distance-to-default model）與稀有性羅吉斯模型（rare event logit models）來預測公開之市場資料，並以cumulative accuracy profiles 及 receiver operating characteristic 兩個指標衡量預估結果，發現稀有性羅吉斯模型表現均優於違約距離模型。本文所觀察的樣本為美國 S&P 500 有信用評等紀錄的上市公司資料，總計由 1986 年 01 月到 2012 年 12 月，共包含 2,138 家公司，或 271,912 個公司月份，其中 444 家為財務危機公司。在預估上，本文設定各時點違約機率最高的 6% 的公司為可能有問題的公司，並以貝氏定理的後驗機率（Bayesian posterior probability）來評估，結果發現此計算方式能更早一步捕捉到 40-60% 的準確度，甚至有時可高達 70%。因此，本文認為稀有性羅吉斯模型是個可用來預估財務危機公司的好方法，且此項預估至少可提早三年就能有效觀測使用。

This dissertation is mainly focus on two empirical finance studies: the rollover effects in stock index futures contracts, and searching for the distressed firms in equity markets. For research of the rollover effects in stock index futures contracts, this dissertation proposes a new method for evaluating stock index futures contracts rolling returns with the rollover effect analysis. Due to the limited lifespan of futures contract, traditional return series construction may be distorted by the price jump and contract inconsistency problems. We amend it by replacing the rollover price of the nearest-to-maturity to the next-to-maturity price at the rollover day, and decomposing the total return to the rollover effect and the capital gain or loss. Three rolling points, the delivery day, the seventh day, and the first day of the expiration month, are considered. Differences between the actual and theoretical futures prices are also discussed. Convenience yields for the near close and next-to-maturity futures are also explored. We investigate S&P500, DAX, and TAIEX stock index futures, and find that with the risk-return and rollover effect analysis, rolling into the next contract on the delivery day is often a better choice. By the convergent paths of the convenience yields we observe that expiration day effects are very apparent in all three markets. The second topic is on the distressed firms research and prediction. Distressed firms in equity markets are like landmines in the battlefields due to their undetectability and devastating effects. This dissertation is concerned with distressed firms forecasting by the distance to default and rare event logit models via public available data. Comparing these two models by Cumulative Accuracy Profiles and Receiver Operating Characteristic curves, we conclude that the rare event logit model performs better than the distance to default model. The data contains U.S.-listed firms on the S&P 500 for the period January 1986 to December 2012, including 2,138 companies and 271,912 firm months, with 444 distressed firms. We set the dynamic thresholds as the last 6% of firms based on the historical cross-section distress rates. Upon Bayesian posterior probability examination, the rare event logit model shows about 40% to 60% affinity with S&P Domestic Long Term Issuer Credit Rating records on average, and the rate increases to 70% in some situations. We conclude that the rare event logit model can be a good warning indicator of distress in firms at least three years ahead.