為因應2008年的金融危機，國際會計準則理事會在2008年10月13日發布IAS 39和IFRS 7的修訂條文，允許企業在特定情況下，將非衍生性之交易目的金融商品重分類至其他類別，並可將基準日追溯至當年7月1日。基於相同的理由，我國在2008年10月17日亦隨即修訂財務會計準則公報第34號，與國際接軌。企業若將交易目的金融商品重分類至備供出售，將使原先進入損益表的損益數字直接進到資產負債表，使盈餘波動較少。本研究提出維持平穩化盈餘趨勢假說，測試重分類決策與歷史盈餘平穩化的關聯；並提出市場效率反應假說，測試市場對重分類行為的反應，是否僵固於報導的盈餘數字。我國上市櫃的一般產業，有超過半數持有交易目的金融資產，其中約有15%在2008年時將部份金融商品由交易目的重分類至備供出售。實證結果發現，在控制當期營運表現、金融商品持有部位、企業規模、債務契約、獎酬契約、公司治理等要素後，企業重分類之行為，與歷史盈餘平穩化程度顯著相關；歷史盈餘平穩化程度較高的企業，較傾向於重分類。而2008年的重分類盈餘影響數，與當期股票報酬呈現負向關係，與次期2009年股票報酬呈現正向關係；顯示投資人評價時會將重分類資訊納入考量，不會僵固於報導的盈餘數字。此外，本研究利用重分類與否進行套利投資組合測試，無法獲得超額報酬，顯示投資人並未因企業將金融商品重分類，而造成錯誤評價，市場在重分類事件上呈現效率反應。

In response to the financial crisis, the IASB issued an amendment to IAS 39 and IFRS 7 on October 13, 2008, which permits entities to reclassify non-derivative financial assets out of the held-for-trading category. Following the same logic, Taiwan's Financial Accounting Standard Committee immediately amended the Statement of Financial Accounting Standard (SFAS) No.34 on October 17, 2008, which allows firms to reclassify their trading financial instruments into other categories from July 1, 2008 on. Reclassifying trading securities to available-for-sale securities allows firms to shift valuation losses or gains from income statement to the balance sheet, which results in smoother earnings. In this study, we predict that firms' previous income smoothing experience may affect their choice of reclassification. We also examine whether the market reacts efficiently to the reclassification decision. Our study finds that more than half of the listed nonfinancial companies held trading purpose securities before the amendment, and 15% of these firms reclassified the securities in this category mostly as available-for-sale purpose. The empirical results show that, after controlling for operating performance, financial asset holding, size, debt covenant, compensation contract variables, and corporate governance factors, firms with more income smoothing experience before the financial crisis are more likely to reclassify the securities from trading purpose to available-for-sale, which lends a support of income smoothing hypothesis. Moreover, the earnings effect of reclassification in 2008 is negatively correlated with the concurrent year security return but positively correlated with the return in 2009, suggesting that investors are not fixated on the earnings. Finally, our results are inconsistent with the hypothesis that investors are misled by the reclassification as no significant difference exists in abnormal returns between the reclassification firms and the other firms. In sum, the market reacts efficiently to the reclassification decision.