盈餘是公司過去經營績效良窳之最終表現，而盈餘數值高低與公司股價報酬有密不可分的關係。然則，盈餘是企業營運的一連串會計處理結果，不同的會計原則及假設會影響會計處理的結果，使得當期及未來盈餘數值均會受到影響，因此在評估或預測企業的盈餘時，應對盈餘本身之品質加以探討，亦即，如何確認財務報表中那些是攸關盈餘品質優劣的資訊，透視盈餘本身的真正內涵以輔助投資人形成最佳投資策略，十分值得我們進一步研究。 近年來由於人工智慧之類神經網路快速地發展，加上類神經網路具備了平行分散式處理、關聯式記憶、自範例中學習等類似人類非線性思考的能力，在財務系統的應用上，學者所建構的類神經網路都比統計方法獲得了更好的結果。 有鑑於此，本研究即依據上述概念，以民國七十九年至八十四年共計五個年度財務報表資訊，以第一、二類上市公司一共十三個產業為研究樣本，建構了盈餘品質類神經網路預測模式，找出盈餘品質資訊內涵與盈餘成長率之關聯性。並以模式預測結果形成投資組合並據以作為投資策略操作。 在網路模式建構階段，本研究採取了過去學者所採用的盈餘品質指標作為網路之輸入結點；以每股盈餘成長率作為網路之輸出結點；以整體市場為學習範例，進行隱藏層結點個數之操弄，以找出學習效果較佳之網路模式，並以此網路模式作為後續研究採用之依據。以整體市場為樣本所進行的網路測試過程中，本研究所找出之較佳網路模式為：9-9-1。 本研究根據前述方法所進行的研究中，獲得了以下結論： 一、以整體市場為樣本所進行的測試中發現，模式之區別能力大致介於六成至七成之間。而預測能力大約是在五成至六成之間。 二、在整體市場、紡織類股以及電子類股之測試結果方面，以電子業之模式區別能力及預測能力最好，其次為紡織業。顯示以單一產業為樣本之模式學習效果優於整體市場。 三、在網路穩定性方面，則以紡織業組之穩定性較高，但與其它兩組之差異性並不明顯。 其次本研究以事件研究法進行投資策略分析，以模式之預測結果，輔以益本比評價法形成投資組合並進行投資決策，獲得了以下結論： 一、以整體市場、紡織類、電子類為投資對象均獲了超額報酬，在觀察期間內分別獲得了38.51%、34.62%以及56.89%的超額報酬率。其中以電子類股之表現最為突出。顯示本研究對於如電子業較重視研究發、資本密集之產業盈餘品質萃取能力較佳。 二、在觀察期間內，投資組合與類股報酬率表現均呈現正向相關，在類股指數上漲月份中，投資組合之超額報酬率較小，然而在類股指數下跌月份投資組合會出現較大幅的超額報酬。推論其原因在於本研究是以盈餘品質為基礎，而此類盈餘品具成長性且一致性、穩定性較高之公司較具抗跌性及長期持有之價值。 三、本研究驗證了盈餘品質網路模式能有效擷取財務報表盈餘資訊內涵，以之形成投資策略能獲取超額報酬。 關鍵字：盈餘品質、類神經網路、盈餘品質預測模式、投資組合、投資策略、累計超額報酬

財務報表記錄可說是企業經營績效良窳的反映指標，而其中所衍生出來的財務比率，向 來均是管理者、投資者進行企業診斷或未來經營績效預測的重要資訊來源。然而，相關 的研究發現，由於產業間經濟環境與市場結構特性的不同，所呈現出來的財務報表資訊 內涵亦將有所差別。因此，若進一步運用個別產業之報表資訊預測公司未來盈餘時，將 能夠提供產業間結果進行分析與比較的基礎。 如何自報表中獲取與公司經營績效相關之會計資訊，進而建構出優良的盈餘預測模式， 是近幾年來學者感興趣的研究課題之一。鑑於人工智慧之類神經網路系統擁有多項的特點，因此，對於盈餘預測會計資訊萃取的應用上，無非是提供了我們一個新的選擇途徑。 本研究即根據此項概念，以民國70年第一季至民國82年第三季為止共十五項大小產業之 股票上市公司財務報表以及股價報酬等資料作為研究樣本，進行盈餘預測模式的建構以 及投資超額報酬的計算。 進一步地說，本研究的內容可以分成三個部份，第一部份是以整體市場樣本為例，對類 神經網路主要參數如輸入變數組合、隱藏層節點數等進行調整及測試，以從中選取出盈 餘預測效果較佳之模式設定；在第二部份則是運用此一盈餘預測模式，分別對整體市場 以及紡織、電子電機 兩項產業樣本進行網路的訓練與測試，並根據模式所獲得之區別及 預測能力評估指標，探討不同產業特性樣本所建構的模式之間，其預測結果上的差異性 ；而第三部份則是利用各類產業模式預測結果的資訊，從利潤與風險兩種角度，定義"總 體"、"高利潤"、"低風險"、 "高利潤低風險"等四種不同類型投資策略，並以事件研究 法計算各項策略所能獲取之累積超額報酬，最後，則根據各策略之獲利績效，進行產業 間的分析比較，以找出本研究各類特定產業之最適投資策略。 本研究根據前述方式所進行的實驗研究中，獲得了以下三點結論： 一、類神經網路盈餘預測模式之建構 (一)以整體市場樣本為對象所進行之網路的測試中，發現模式整體區別能力大致介於五 到七成之間；而整體預測能力則介於四到六成之間。 (二)本研究所找出盈餘預測效果較佳之網路模式設定如下：1.輸入變數組合：單因子多變量變異數分析之22項顯著性財務比率 2.網路架構(輸入層-隱藏層-輸出層)：22-22-1 3.連結權數初始值設定範圍：-0.1~0.1 二、產業盈餘預測結果之分析 (一)整體而言，產業間模式測試結果的差異並不大，其中以紡織產業的模式區別及預測 能力最好(70%以上)，電子電機產業次之，而整體市場模式的結果均不及兩項單一性產業。 (二)模式預測能力穩定性方面，各產業於五個年度間預測率的波動大致還算穩定，其中 就紡織產業而言，其年度之間模式預測能力的差別不大，但電子電機產業年度間的變化 則要比前者來得明顯。 三、產業投資策略績效之分析 (一)各類型投資策略的整體結果中，紡織與電子電機兩項產業的獲利績效相當，且均要 比整體市場來得好，其中，紡織產業之"高利潤低風險"策略所獲得的累積超額報酬(43.28%) 更居全體之冠。 (二)本研究所找出之個別產業最適投資策略分別為： 1.整體市場：總體策略、低風險策略 2.紡織產業：高利潤低風險策略、高利潤策略 3.電子電機產業：高利潤低風險策略、低風險策略 / Financial Statements are very important information indicating performance of corporations. Managers and investors use financial ratios as vital indexes to evaluate and predict operating results of corporations, and make their decisions. ategy, and compute CAR for each investment strategies. At last, I analyze the investing results of the four strategies for individual industry. ANN ( Artificial Nerual Network) shoot a new direction on researching application of abstracting accounting information which can efficiently predict earnings. According to results of relative researches, financial statements from different industries present and implicate different accounting information. If we further apply ANN on financial statement information to predict earnings of corporations, we can use the results as bases of analyses and comparisons among industries. Because ANN model has many advantages, in this research, I use financial statements and return on stocks from corporations as researching samples to construct prediction models and compute CAR(Cumulative Abcdrmal Return) on investments. These samples are chosen from 15 different industries and covered from the first quarter of 1981 to the third quarter of 1993. This research consists of three parts: 22 financial ratios selected by MANOVA First, I use the general market samples to adjust and predict the vital parameters of ANN models, such as the selection of input variable, the number of hidden node, and finally pick better setups for the prediction model. Second, I use this model to train and test samples from the general market, the textile, and the electron industry, and research the variation of predicting results by different models made up different industries by means of evaluation indexes . Third, I use the results predicted by the three different industry models, inspect of risk and return, to define four types of investment strategies -- "the general", "the high return", "the low risk", and "the high return - low risk" strategy, and compute CAR for each investment strategies. At last, I analyze the investing results of the four strategies for individual industry. After researching, I find:s of the textile and electron industry are better than the general markets'. 1.The better setups of ANN predition models are :industries are: (1)the selection of input variable:the 22 financial ratios selected by MANOVA (2)the ANN model topology(input node - hidden node - output node):22-22-1 rategy (3)the range of initial connection weights:-0.1~0.1 return - low risk strategy 2.The analyses of results predicted by the three different industry models are: (1)the predicting abilities of the textile and electron industry are better than the general markets'. 3.The proper investment strategies of individual industries are: (1)the general market:the general and the low risk strategy (2)the textile industry:the high return and the high return - low risk strategy (3)the electron industry:the low risk and the high return - low risk strategy

現行各國的退休金計畫逐漸地由確定給付制轉變為確定提撥制。這表示投資的風險由原本退休金計畫的發起者(雇主)轉移到了參與者(員工)的身上。為了減少每個確定提撥制計畫參與者的投資風險，本文中採用退休時所得替代率為預估的目標，藉由模擬與最適化的方法找到最適投資策略與最適提撥率。 能反映出時間性的隨機模型在精算科學的領域是日漸重要，本文試著藉由隨機性的變化來估計代替以往精算上各種假設下所求得的負債。本文藉由隨機模擬的方式，得到各種資產在市場上或者是經濟上的價值來建構相關投資標的之報酬率，並利用動態隨機規劃模型去改善財務上避險以及資產負債管理。此外，為了避免模擬分析時間過長的問題，本文採用了情境抽樣的方法去改善電腦模擬分析計算時的效率。 我們主要得到以下結論： (一)確定提撥制下的負債受薪資水準波動的影響，所以此時會持有較 多的指數連結型債券以反應薪資水準及通貨膨脹的影響。整體投 資的結果與Vigna & Haberman (2001) 文中的結果及實務上生命 週期型態(lifestyle)投資方式呈現相同的現象。 (二)考慮每期下跌風險(downside risk)時，期中的投資可能會偏向 於投資風險較高的股票。在每年觀察下跌風險的情況下其投資因 為必須考慮避免每一年的下跌風險，需要比每五年觀察下跌風險 的情況做風險較大的投資，以達到其目標。 (三)在本文的調整投資組合策略下，因為調整次數不多，所以在考慮 交易成本的情況，當交易成本很小時對於整體的最適化資產配置 與最適化提撥率的影響是很小的。在本文的調整投資組合策略 下，交易成本的影響只有在交易成本非常大的情況下才能看得出 來。 (四)均勻抽樣法抽出的400組情境幾乎可以完全的代替4000組情境， 其結果可以看出與未抽樣相同的生命週期型態(lifestyle)投資 方式。而隨機抽樣法的結果雖然也可看出趨勢，但準確性相對於 均勻抽樣法仍稍嫌不足，並不適合用來代替原先的4000組情境。 / A shift from defined-benefit pension plan towards defined-contribution pension plan is currently popular around the world. This means that a serious investment risk transfers from defined-benefit sponsors to the individual members of defined-contribution plans. In order to reduce the risk of individual DC member, we investigate the methodology of finding the optimal contribution rate and asset allocation to reach a certain target of the retirement replacement rate in this paper. Stochastic processes are getting more important to the field of actuarial science. Instead of trying to approximate liabilities by a single deterministic set of actuarial assumption, we seek to take account of market or economic valuation for both assets and liabilities using stochastic simulation. We applied dynamic stochastic programming models to improve financial hedging and asset liability management. Moreover, in order to avoid the problem of time-consuming, we use scenario sampling method to improve the efficiency of computer calculation. We draw four conclusions from our investigations： (1)We will hold more assets in indexed-linked bonds because the pension liability is highly related to the wage- index and inflation rate. The optimal investment strategy is very like the so called "lifestyle" investment strategy. (2)When we consider downside risk, we should hold more risky equities. The investment strategy is more risky when we consider downside risk every year than every 5 years. (3)Under our rebalancing strategy, if the transaction cost is small, the influence on the investment strategy and contribution rate is small. We can see the influence of the transaction cost in a situation that the transaction cost is very big only. (4)There are almost no different between uniform sampling scenarios and original simulation scenarios, so uniform sampling scenarios may replace the original simulation scenarios perfectly. And random sampling method is unsuitable to replace the original simulation scenarios.

本研究中，我們計算OGA演化投資策略在五類時間數列模型上之表現，這五類模型分別是線性模型、雙線性模型、自迴歸條件異質變異數模型、門檻模型以及混沌模型。我們選擇獲勝機率、累積報酬率、夏普比例以及幸運係數做為評斷表現之準則，並分別推導出其漸近統計檢定。有別於一般計算智慧在財務工程上之應用，利用蒙地卡羅模擬法，研究中將對各表現準則提出嚴格之統計檢定結果。同時在實証研究中，我們考慮歐元兌美元及美元兌日圓的tick-by-tick匯率資料。故本研究主要的重點之一，乃是對於OGA演化投資策略，於這些模擬及實証資料上的有效性應用，作了深入且廣泛的探討。 / In this study, the performance of ordinary GA-based trading strategies are evaluated under five classes of time series model, namely, linear ARMA model, bilinear model, ARCH model, threshold model and chaotic model. The performance criteria employed are the winning probability, accumulated returns, Sharpe ratio and luck coefficient. We then provide the asymptotic statistical tests for these criteria. Unlike many existing applications of computational intelligence in financial engineering, for each performance criterion, we provide a rigorous statistical results based on Monte Carlo simulation. In the empirical study, two tick-by-tick foreign exchange rates are also considered, namely, EUR/USD and USD/JPY. As a result, this study provides us with a thorough understanding about the effectiveness of ordinary GA for evolving trading strategies under these artificial and natural time series data.

在台灣的高科技產業發展中，創投扮演了重要的角色，其所提供的資金與協助促進了新創事業的成長，造就了台灣資訊電子產業的躍居世界舞台與傲人的經濟成長。2000年人類基因圖譜的解出，似乎加快了基因技術的應用與相關產業的發展，進而成為全球經濟成長的新動力，因此我國創投的投資觸角也積極地延伸到生物技術產業上。 　　過去有關創投的相關研究多集中在投資評估準則的總體性研究，甚少針對某一產業的特性不同，來做進一步的研究，特別是在生物科技產業方面，因此，本研究將針對生物技術產業的特性與創投的股東背景、經營團隊與合作網路等組成因子，來探討其所產生的投資策略與行為。 　　本研究採用個案訪談之定性研究，選擇六家在生技產業投資比重較大的國內創投公司做訪談，再依據本研究架構進行分析整理，得到了以下之結論： 　　一、 生技產業的特性對投資策略與行為之影響 　　　1. 創投因看好生技產業的成長潛力而將提高此方面的投資比重，而生技產業的投資金額以美國為最高。 　　　2. 創投在生技產業的投資階段傾向涵蓋不同的階段。 　　　3. 創投在生技產業的投資以醫藥產業及其週邊之醫療器材為主，主要考量是醫藥產業是目前為止較高報酬的領域。 　　　4. 創投在生技產業的投資區域以美國為主，其中最重要的原因與該地區之產業群聚有關。 　　　5. 創投在生技產業的投資傾向以投資組合管理及聯合同業投資來降低投資風險。 　　　6. 創投在生技產業的投資傾向不聘任外部顧問，而傾向以經營團隊之專業評估為主，再以已投資公司與事業夥伴為諮詢對象。 　　　7. 創投在生技產業的投資回收策略為上市或購併，投資回收期間並不會因為生技產業的產品開發期長而延長。 　　　8. 創投在生技產業的投資評估，著重整體性評估（不會只看技術或智慧財產權），會因事業投資階段而有不同的評估重點，投資案愈偏早期，愈著重技術與人。投資案愈偏成熟期，所需評估的項目愈多。 　　　9. 生技產業的特性雖對創投的附加價值沒有影響，但創投對生技產業的投資案有提供附加價值，會因投資案事業發展階段之不同，而提供所需之協助，附加價值則以資訊蒐集與人脈介紹為主。 　　二、創投的組成對投資策略與行為之影響 　　　1. 創投的股東對外部顧問策略、投資案源與投資評估有影響，對投資金額、投資階段、投資領域、投資區域、風險控管、回收策略與附加價值沒有影響。 　　　2. 創投的經營團隊對投資金額、投資階段、投資領域、投資區域、風險控管、外部顧問、回收策略、投資評估與附加價值有明顯的影響，對投資案源則有些影響。 　　　3. 創投的合作網路對外部顧問、投資評估與附加價值有影響，對投資案源更是有明顯影響，而對投資金額、投資階段、投資領域、投資區域、風險控管與回收策略沒有影響。　 　　三、生技產業的特性對創投的組成之影響 　　　1. 生技產業的特性對創投董事會內的股東背景沒有影響。 　　　2. 生技產業的的特性影響到創投招募技術專業之經營團隊。 　　　3. 創投未因生技產業的特性而建構新的合作網路，而傾向運用集團中原有之合作網路，尤其是過去的已投資公司，為創投主要的諮詢者。 / Venture capital plays an important role in the development of high technological industries in Taiwan. It provides the essential fund and useful assistance to promote the growth of start-up companies. Because of it, the growth of economy in Taiwan dramatically increases and Taiwan has become the kingdom of information and communication industries around the world. In the year 2000, the complete sequence of human genome has enhanced the speed of the development in the field of biotech and its associated industries. In addition, the investment in biotech industry is expected to stimulate another trend of global economic growth. Therefore, venture capital in Taiwan also actively extends its influence in the field of biotech industries. 　　The majority of researches in venture capital seems to concentrate on the overall evaluation of the general criteria of investment, few studies focused on one particular industry, especially the biotech industry and its characteristics and aspects. Therefore, this study will aim at the characteristics of biotech industry and the constituent factors of venture capital to explore the strategies and behaviors of investment. 　　A qualitative research was conducted in six important venture capital firms in Taiwan using a method of intensive personal interview. The summaries of the research findings are as follows: 　　I. The impacts of the characteristics of biotech industry on the strategies and behaviors of investment: 　　　1. Venture capital firms will increase investment percentage in biotech area because of its potential of growth, and the majority amounts of venture capital seem to be invested in the United States. 　　　2. The venture capital firms’ investments in biotech tend to cover various stages. 　　　3. The fields venture capital firms invest in biotech appear to be focused on pharmaceuticals and medical devices, because the operating return from this area is higher than any others so far. 　　　4. The location of biotech companies venture capital invested are focused on the United States, which seem to have obvious phenomenon of clustering. 　　　5. Through portfolio management and co-investment, venture capital firms can reduce risk while investing in biotech. 　　　6. Instead of relying on outside consultant when invest in biotech, venture capital firms prefer to depend on inside management teams for due diligence. In addition, the past invested firms and partners are helpful while needed. 　　　7. The exit strategies of venture capital firms are initial public offering or merger & acquisition when invest in biotech companies, and the period of investment appears to be not correlated with the long product life cycle of biotech industry. 　　　8. Instead of emphasis on technologies or intellectual properties, venture capitalists emphasize all factors which evaluating biotech companies. Their decision criteria depend on venture development stages, the earlier stages these cases are, the more important technologies and management teams are, the later stages these cases are, the more factors are considered. 　　　9. Venture capitalists add values to the biotech companies they invested, not because of the characteristics of biotech industry, but differ from development stages of cases. Most of add values are information collection and networking. 　　II. The influences of the constituent factors of VC on the strategies and behaviors of investment: 　　　1. The stockholders of venture capital affect outside consultant strategy, deal flow, due diligence, but make no influence on investment amount, venture development stage, field, location, risk control, exit strategy and value-added. 　　　2. The management teams of venture capital obviously influence investment amount, venture development stage, field, location, outside consultant strategy, risk control, exit strategy, due diligence and value-added, and make a little influence on deal flow. 　　　3. The networking of venture capital make a little influence on outside consultant, due diligence and value-added, and make obvious influence on deal flow, but do not affect investment amount, venture development stage, field, location, risk control and exit strategy. 　　III. The influences of the characteristics of biotech industry on the components of venture capital: 　　　1. The characteristics of biotech industry don’t appear to affect the background of stockholders in the board. 　　　2. Venture capital firms recruit professional management teams because of the same particular characteristics of biotech industry. 　　　3. While investing in biotech industry, venture capital firms utilize networks, especially the past invest , as their main consultants.

本研究的主要目的是希望喚起國內、外學者對演化科學在經濟學上的重視，結合電腦、生物科技、心理學與數學於經濟學中，希望對傳統經濟學上因簡化假設而無法克服的實際經濟問題，可以利用電腦模擬技術獲得解決，並獲取新知與技能。 本研究共有六章，第一章為緒論，敘述緣由與研究動機。第二章介紹傳統經濟學的缺失，再以資料掘取知識及智慧系統建構金融市場。第三章則介紹各種不同人工智慧的方法以模擬金融市場的投資策略。第四章建立無結構性變遷時間序列模型--交易策略電腦模擬分析，僅以遺傳演算法模擬金融市場的投資策略，分別由投資組合、交易成本、調適性、演化、與統計的觀點對策略作績效評分析。第五章則建立簡單的結構性變遷模型，分別由調適性與統計的觀點，採取遺傳演算法再對投資策略進行有效性評估分析。第六章則利用資料掘取知識與智慧系統結合計量經濟學的方法，建構遺傳演算法發展投資策略的步驟，以台灣股票市場的資料進行實証研究，分別就投資策略、交易成本、調適性與演化的觀點作分析。最後一章則為結論。 未來研究的方向有： 1. 其他各種不同人工智慧的方法的比較分析，如人工神經網路、遺傳規劃法等進行績效的交叉比較分析。 2. 利用分類系統(Classifier System)與模糊邏輯的方法，改善標準遺傳演算法對策略編碼的效率，並建構各種不同的複雜策略以符合真實世界的決策過程。 3. 建構其他人工時間資料的模擬比較分析，例如ARCH (Autoregressive Conditional Heteroskedasticity)模型、Threshold 模型、 確定性(Deterministic) 模型等其他時間序列模型與更複雜的結構性變遷模型。 4. 進一步研究遺傳演算法所使用的完整資訊(例如，各種不同指標的選取)。 5. 本研究係採用非即時分析系統(Offline System)，進一步研究即時分析系統 (Online Sysetem)在實務上是有必要的。 / Historically, the study of economics has been advanced by a combination of empirical observation and theoretic development. The analysis of mathematical equilibrium in theoretical economic models has been the predominant mode of progress in recent decades. Such models provide powerful insights into economic processes, but usually make restrictive assumptions and appear to be over simplifications of complex economic system. However, the advent of cheap computing power and new intelligent technologies makes it possible to delve further into some of the complexities inherent in the real economy. It is now feasible to create a rudimentary form of “artificial economic life”. First, we build the framework of artificial stock markets by using data mining and intelligent system. Second, in order to analyze competition among buyers and sellers in the artificial market, we introduce various methods of artificial intelligence to design trading rules, and investigate how machine-learning techniques might be applied to search the optimal investment strategy. Third, we create a miniature economic laboratory to build the artificial stock market by genetic algorithms to analyze investment strategies, by using real and artificial data, which consider both structural change and nonstructural change cases. Finally, we use statistical analysis to examine the performance of the portfolio strategies generated by genetic algorithms.

在全球持續電子化的過程中，台灣一路扮演著落實有效製造與實現設計的推手，無論從主機板、被動元件、面板、晶片設計、晶圓製造以及封裝甚至高精密組裝無所不包，在如此資本密集產業，如何操作才能達到供需平衡，為整體產業經濟創造出良性的競爭平台，避免惡性競爭，就成了不可輕忽的課題。 本文以晶圓產業為探討對象。全球產能平均利用率多年來總維持於中檔 (88%)， 且平均銷售單價只能緩降而無強勢反彈，近年企業無不減少資本支出來度過低潮，整合元件製造廠(IDM)相繼喊出工廠資產輕化(fab-lite)的營運策略，這時我們的命題即是：要不要繼續投資？如何調整價值鏈？ 本研究中，我們會先檢視目前市場對半導體成長預測的準確度，再經由產業價值活動代表性指標回歸分析法對相關參數做一整合之解析，對晶圓需求量與銷售價建立配適之模型，找出先行指標來達到預測，並定義上下限以供快速比對分析，最後再根據分析結果提出可能之產業趨勢議題。 / This thesis analyzes the semiconductor industry growth worldwide. The leading index via regressions has been established to achieve a reliable forecast on worldwide ASP, wafer demand and revenue. In the long run, we expect the semiconductor demand will continue to grow at CAGR 8% (compound annual growth rate) and display less extreme cycles than past decade. However, revenue’s CAGR might be diluted to around 5%, lower than demand’s growth. Moreover, it might go down to zero-growth for some times since the ASP still slightly trend down before emerging market demand really expanded. Continuous outsourcing is one possible solution for IDM to be fab-lite, since fab’s fix charge is billion- base that needs high utilization to maintain break-even operation. But what is the solution for foundry side to avoid ASP erosion all the way down? Our analysis identifies a need for executive managers to well predict the demand on capital expenditure when making decision.