外文翻譯---銀行信用風(fēng)險(xiǎn)管理硬信息，軟信息與控制

1、<p><b>　　中文3620字</b></p><p>　　本科畢業(yè)論文（設(shè)計(jì)）</p><p>　　外 文 翻 譯</p><p>　　外文出處 University Robert Schuman, </p><p>　　Strasbourg III Insti

2、tutd’Etudes Politiques </p><p>　　外文作者 Brigitte Godbillon </p><p><b>　　原文：</b></p><p>　　Credit Risk Management in Banks:</p><

3、;p>　　Hard Information, Soft Information and</p><p>　　Manipulation</p><p>　　Introduction</p><p>　　Information remains a crucial input for the banking industry. Banks are confronte

4、d to information’s asymmetry problems because of borrowers’ informational opacity. This opacity varies borrower’s type, SMEs being considered as the most opaque (because of lack of public information). In order to resolv

5、e this informational asymmetry, the bank can acquire two types of information: hard information, which is external, via public information (balance sheet data, rating, scoring . ), and soft informati</p><p>

6、　　Recent research on risks management in banks puts forward the importance of information’s treatment. Hakenes (2004) considers the banker as a “specialist” of information’s treatment and risk’s monitoring. Danielsson et

7、 al. (2002) analyze bank’s choice of risks management system, investigating different levels of power delegation implying more or less transmission of information. However, this stream of research doesn’t distinguish har

8、d and soft information.</p><p>　　Therefore, this article investigates the impact of information’s type on the balance sheet structure and organization in terms of credit risk management of the bank. We propo

9、se a theoretical model of the credit decision within a principal-agent framework with a bank director (or a director of the credit risk department) and a credit officer (or a bank’s agency clerk). The director allocates

10、equity for Value at Risk coverage. He also decides on the officer’s budget, as well as her wage, which are</p><p>　　We find several interesting results. We show that taking into account soft information in r

11、isk management can allow to economize equity for VaR coverage, under certain conditions. However, we also verify the existence of the soft information’s manipulation incentive by the officer. We then propose a wage schem

12、e to impeach this manipulation. The influence of soft information on banks’ organizational structure is modelled through a specific salary package. The comparison of solutions from the two </p><p>　　The rest

13、 of this article is organized as follows. In section 2, we present elements allowing to distinguish between hard and soft information, as well as recent theoretical research investigating information’s type influence on

14、banks’ organizational structure. The credit risk decision model is presented in section 3. Sections 4 and 5 provides results in the hard information’s case, and deduce pros and cons of soft information, in particular the

15、 existence of a manipulation incentive in presence o</p><p>　　Regarding soft information, its most particular characteristic is to be tightly linked to the environment and context where it was produced. In t

16、he banking framework, this environment is the bank-borrower relationship, which, through multiple interactions in time, gives access to private and confidential information, superior to publicly available information (Be

17、rger, 1999; Boot, 2000; Berger and Udell, 2002; Elsas, 2005). Thus, soft information has the advantage to increase the predictive capac</p><p>　　The adaptation of the organizational structure to the type of

18、information used in banking has also received theoretical and empirical evidence. Stein (2002) investigates the influence of banks’ organizational structure on their optimal funds allocation’s decision. In a large bank,

19、processes of collecting and treating information, needed for credit decision, are separated. Thus, the information must be easily transmissible to superior hierarchical levels. Also, it should also be easily interpreta&l

20、t;/p><p>　　versus non-hierarchical and decentralized). He then shows the existence of an adequacy between organizational structure and information type, allowing optimal funds’ allocation, through better incent

21、ives. Soft information is associated with decentralized organizations, because they provide the agent more power and authority. In such a framework, she has better incentives to make efficient use of her information in t

22、he funds’ allocation process. Hard information is associated with centralized org</p><p>　　Several empirical studies confirm this theoretical evidence (Berger and Udell, 2002; Berger et al., 2001; Berger et

23、al., 2002a; Berger et al., 2005). In a bank-borrower relationship framework, information’s production and treatment is delegated to a credit officer, who therefore receives strong authority and power, because of possible

24、 soft information’s manipulation. In this context, the officer has a crucial position within the bank. Small, less hierarchical and decentralized organizations are </p><p>　　Another strand of literature inve

25、stigates wages’ structure and budgets’ allocation role in driving proper incentives of credit officers in charge of information’s treatment. Bernardo et al. (2001) consider a risk-neutral firm with an investment project

26、for which optimal fund allocation depends on uncertain project’s quality. A manager can be hired in order to produce information about this quality. She enjoys private benefits from controlling the allocated budget and r

27、eports project’s non verifi</p><p>　　Another model considers several agents within the same firm (Ozbas,2005). This author studies information asymmetries’ problems in a model of internal competition for res

28、ources between managers in a firm. These managers have access to information qualified as specific knowledge. The model focus on the concept of strategic communication between a principal - firm’s director - and the agen

29、ts - firm’s managers. These have a utility define on budget, thus inciting them to exacerbate return’s predict</p><p>　　A study of information’s role in financial markets is provided by Ozerturk (2004). This

30、 author investigates the influence of wage scheme on funds manager incentives to acquire precise information. The wage scheme includes fees and is also indexed on managed fund’s returns. The fund manager - the agent - ex

31、ergues costly effort in order to observe assets’ return for the benefit of a investor - the principal. Her effort influence information’s precision, which is modelled through a signal. Precisio</p><p>　　In w

32、hat follows, we aim at linking these two separate strands of literature, one dealing with the relationship between the type of information and organizational structure, and the other dealing with optimal wage in a princi

33、pal-agent framework with hidden information (?), focusing on the process of bank risk management organization.</p><p>　　Information’s quality, produced by the bank, determines its risk taking characteristics

34、. Existing literature treats this problem with a distinction of hard and soft information (Petersen, 2004). Acquiring this information can be done through two methods: transaction lending or relationship lending. The for

35、mer can use statistical methods of hard information’s treatment. This type of information presents several advantages, as low cost, economies of scale and the possibility to measure Value at Ri</p><p>　　In t

36、his article, we have focused on information’s type role in credit risk management in banks. In a principal-agent model with moral hazard with hidden information where a banker requires information on assets’ return in or

37、der to manage credit risk through equity allocation for VaR coverage, we show that using additional soft information allows to economize equity, thanks to soft information’s higher precision. However, this type of inform

38、ation being not verifiable, it requires to implement a p</p><p>　　These results provide theoretical evidence on soft information’s advantage in credit risk management, as we show that VaR can be reduced, whi

39、ch allows to economize equity, even if the bank must implement a specific organizational structure and an adequate wage scheme.</p><p>　　Source：University Robert Schuman, Strasbourg III Institut d’Etudes Pol

40、itiques</p><p><b>　　譯文：</b></p><p>　　銀行信用風(fēng)險(xiǎn)管理：硬信息，軟信息與控制</p><p><b>　　介紹:</b></p><p>　　信息仍然是銀行業(yè)的關(guān)鍵投入點(diǎn)。面對(duì)銀行，仍然存在借款人的信息不透明與信息的不對(duì)稱問(wèn)題。這種不透明在不同借款人的類型，

41、中小企業(yè)被認(rèn)為是最不透明的（由于缺乏公共信息）。為了解決這種信息不對(duì)稱，銀行可以掌握兩種類型的信息：1.硬件的信息，通過(guò)公共信息（資產(chǎn)負(fù)債表數(shù)據(jù)，評(píng)級(jí)，得分等）；2.軟信息，這些信息通過(guò)銀行內(nèi)部借款關(guān)系（判斷，意見(jiàn)，票據(jù)，報(bào)表等）。這也意味著兩個(gè)貸款方面：交易貸款與銀行的關(guān)系。一些專家最近提出要將硬軟信息進(jìn)行區(qū)分（彼得森，2004年）。1.考慮到以風(fēng)險(xiǎn)分析軟件的信息可以提高預(yù)估的借款人的品質(zhì)（萊曼，2003;古方等人，2005年），具有

42、可查性，但也有非可操縱的劣勢(shì)。這種類型的信息可影響銀行的信貸風(fēng)險(xiǎn)管理，但也可能使銀行的組織結(jié)構(gòu)，適應(yīng)軟信息，以避免處理成本造成的后果和影響。</p><p>　　在銀行的風(fēng)險(xiǎn)管理的最新研究提出了信息處理的重要性。哈格森（2004）認(rèn)為是專家對(duì)信息的處理和風(fēng)險(xiǎn)的監(jiān)測(cè)的分析家對(duì)分析銀行的風(fēng)險(xiǎn)管理制度的選擇，調(diào)查不同層次的權(quán)力下放意味著多少的信息傳輸。但是，這種研究流派不區(qū)分硬信息和軟信息。</p>&l

43、t;p>　　因此，本文探討了信息的資產(chǎn)負(fù)債表結(jié)構(gòu)和組織形態(tài)的銀行信貸風(fēng)險(xiǎn)管理方面的影響。我們建議與銀行的董事（或信貸風(fēng)險(xiǎn)部主任）和信貸人員（或銀行的代理員）委托代理框架的信貸決策理論模型，主導(dǎo)在風(fēng)險(xiǎn)保障分配公平方面的價(jià)值。他還決定對(duì)有關(guān)人員的預(yù)算，包括工資，這都是一個(gè)信號(hào)函數(shù)的唯一方面或硬信息和軟信息相結(jié)合的基礎(chǔ)。信號(hào)之間差異的兩種類型的本質(zhì)在于，更準(zhǔn)確的核查其可操作性，以及它們的精確度。一個(gè)軟硬結(jié)合的信息更精確，比硬信息更利于參

44、考，但無(wú)法證實(shí)作為軟構(gòu)件是否可操縱。因此，一個(gè)軟信息與隱藏的信息道德風(fēng)險(xiǎn)源是一個(gè)組織的修改關(guān)鍵，該銀行的潛在驅(qū)動(dòng)以限制道德風(fēng)險(xiǎn)問(wèn)題。</p><p>　　我們發(fā)現(xiàn)一些有趣的結(jié)果。在考慮到風(fēng)險(xiǎn)管理軟信息的情況下可以允許在某些條件下節(jié)約風(fēng)險(xiǎn)保險(xiǎn)的權(quán)益。然而，我們也驗(yàn)證了軟信息的人員操縱激勵(lì)的存在，因此我們提出了一個(gè)工資計(jì)劃來(lái)彈劾這種操縱。關(guān)于銀行的組織結(jié)構(gòu)是通過(guò)一個(gè)具體的薪酬待遇來(lái)仿照軟信息的影響。從兩個(gè)框架的解決方

45、案比較（硬信息與硬軟信息的組合），與數(shù)值模擬的實(shí)現(xiàn)，證實(shí)了軟信息的組件可以提供一個(gè)優(yōu)勢(shì)。</p><p>　　本文的其余部分組織如下：在第二節(jié)，我們對(duì)現(xiàn)在銀行的組織結(jié)構(gòu)元素允許區(qū)分軟，硬信息，以及最近的理論研究，及調(diào)查資料類型的影響。在第三節(jié)提出信貸風(fēng)險(xiǎn)決策模型。第4和第5節(jié)規(guī)定，從硬信息案件的結(jié)果，推導(dǎo)出一個(gè)基于軟信息的信號(hào)存在的優(yōu)點(diǎn)和缺點(diǎn)，特別是軟信息操縱激勵(lì)的存在。第六節(jié)為有關(guān)人員獎(jiǎng)勵(lì)工資計(jì)劃操作并提出解決

46、問(wèn)題。最后，從硬盤(pán)結(jié)果和軟硬結(jié)合的信息情況下進(jìn)行比較，用數(shù)值解答，并在第7節(jié)提出，在第8節(jié)結(jié)束這篇文章。</p><p>　　關(guān)于軟信息，其最特別的一點(diǎn)是緊密地聯(lián)系在一起的環(huán)境和背景。在銀行的框架下，這個(gè)環(huán)境是銀行借款人的關(guān)系通過(guò)長(zhǎng)時(shí)間的相互作用，使之進(jìn)入私人和機(jī)密的信息，公開(kāi)資料的優(yōu)越性（伯杰，1999年;布特，2000年；伯杰和德?tīng)枺?002年；阿爾薩斯，2005年）。因此，軟信息的優(yōu)點(diǎn)是能增加硬信息的預(yù)測(cè)能

47、力，但仍然不對(duì)此進(jìn)行核查。軟信息的能力，增加硬信息的預(yù)測(cè)能力的實(shí)證研究，是在調(diào)查定性預(yù)測(cè)因素對(duì)違約風(fēng)險(xiǎn)的影響。這些研究表明，尤其是銀行的違約風(fēng)險(xiǎn)是預(yù)測(cè)模型集成內(nèi)部評(píng)級(jí)隨著硬資產(chǎn)負(fù)債表上升的經(jīng)濟(jì)因素。對(duì)這些內(nèi)部評(píng)級(jí)強(qiáng)有力的作用是基于定性因素，因此軟信息作為借款人或企業(yè)的質(zhì)量管理有四個(gè)觀點(diǎn)。對(duì)進(jìn)入違約風(fēng)險(xiǎn)因素納入預(yù)測(cè)模型的定性來(lái)增加它們重新分類的性能，由此來(lái)預(yù)測(cè)違約的準(zhǔn)確性（萊曼，2003;古方等，2005。）。此外，定性因素似乎并不那么分

48、散，顯得更穩(wěn)定。該組織結(jié)構(gòu)適應(yīng)了在銀行使用的信息類型，也得到理論和經(jīng)驗(yàn)。斯坦因（2002）考察了銀行的組織結(jié)構(gòu)對(duì)優(yōu)化資金配置的決策影響。在大型銀行，流程信息的收集和處理為信貸決策需要是分開(kāi)的。因此，信息必須容易傳播到高級(jí)的層次水平。此外，還應(yīng)該更容易解釋以統(tǒng)一的方式生產(chǎn)</p><p>　　幾個(gè)實(shí)證研究用證據(jù)證實(shí)了這一理論（伯杰和德?tīng)枺?002; 伯杰，2001）。在銀行借款關(guān)系的框架下，信息的生產(chǎn)和處理委托給信

49、貸人員，因此誰(shuí)獲得強(qiáng)大的權(quán)力和力量，就有可能對(duì)軟信息進(jìn)行操縱。在這種情況下，有關(guān)人員需要在銀行內(nèi)部有一個(gè)關(guān)鍵的位置。小型，低層次和更分散的組織關(guān)系更合適銀行。小銀行被認(rèn)為具有銀行借款人長(zhǎng)期合作關(guān)系的軟信息處理中框架的優(yōu)越能力（伯杰，2004；德?lián)P等人，2004；斯科特，2004）。當(dāng)中小企業(yè)面臨外部資金的選擇時(shí)，他們通常喜歡選擇小銀行來(lái)減少信貸配給。</p><p>　　調(diào)查的另一派別認(rèn)為該利用工資結(jié)構(gòu)和預(yù)算對(duì)信

50、貸人員適當(dāng)獎(jiǎng)勵(lì)，這起到對(duì)信息的處理費(fèi)進(jìn)行分配的作用。貝爾納等（2001）認(rèn)為投資項(xiàng)目的不確定性與優(yōu)化資金配置上項(xiàng)目的質(zhì)量取決于一個(gè)風(fēng)險(xiǎn)性中等的公司。一個(gè)管理人員能產(chǎn)生這個(gè)質(zhì)量的信息。她所分配的預(yù)算控制和報(bào)告項(xiàng)目的質(zhì)量核查并非該公司的董事的私人利益。后者取決于資金分配報(bào)告中的項(xiàng)目。經(jīng)理可就這點(diǎn)作為非允許核查中努力提高項(xiàng)目水平的回報(bào)。經(jīng)理的效用是既影響了預(yù)算和工資分配的計(jì)劃，并由處長(zhǎng)提出一個(gè)固定的組成和可變?nèi)纹凇Ｘ悹柤{等（2001年）表明三

51、種工資和預(yù)算的優(yōu)化組合情況下存在：低報(bào)項(xiàng)目的質(zhì)量意味著沒(méi)有預(yù)算中，平均質(zhì)量意味著預(yù)算中，但沒(méi)有在工資制度變量和高品質(zhì)組別中意味著預(yù)算和工資計(jì)劃中的一個(gè)變量項(xiàng)目。</p><p>　　另一種模式認(rèn)為在同一公司（歐巴斯，2005年）有幾家代理人，具有在內(nèi)部管理人員之間的競(jìng)爭(zhēng)資源與在企業(yè)模型筆者研究的信息不對(duì)稱的問(wèn)題。這些管理人員獲取信息的具體知識(shí)是合格的。公司的董事——代理——公司經(jīng)理對(duì)概念之間的主要溝通模式是戰(zhàn)略重

52、點(diǎn)。這些具有實(shí)用定義的預(yù)算，加劇煽動(dòng)他們對(duì)返回成本的預(yù)測(cè)，以增加獲得更多的資金的幾率。然而，他們的預(yù)言實(shí)現(xiàn)并非減少經(jīng)理人的聲譽(yù)，戰(zhàn)略溝通之間的委托人和代理人的質(zhì)量惡化與組織的整合以及兩種類型的代理人（好的和壞的）和兩個(gè)項(xiàng)目（高回報(bào)和低回報(bào)）類型的存在。不良代理人低回報(bào)項(xiàng)目的項(xiàng)目經(jīng)理具有良好的確定性，可以有高或低回報(bào)的概率為p和1 - P的分別。歐巴斯（2005）研究多種類型的組織（有不同程度的整合）并認(rèn)為在預(yù)算分配硬性規(guī)定下可以提高通信

53、的質(zhì)量和效率，減少管理人員的內(nèi)部競(jìng)爭(zhēng)。集權(quán)可以提高工作效率，但只有當(dāng)溝通是有效的時(shí)候。一個(gè)更高水平的集權(quán)意味著溝通的惡化，這仍然是有效分配資金的關(guān)鍵因素。</p><p>　　一個(gè)信息在金融市場(chǎng)的研究中提供了作用（歐組特，2004）。筆者探討了基金管理人獎(jiǎng)勵(lì)工資計(jì)劃的影響并獲得準(zhǔn)確的信息。該計(jì)劃包括工資費(fèi)用，也是對(duì)管理基金回報(bào)的索引。該通過(guò)基金經(jīng)理——代理的努力，來(lái)觀察資產(chǎn)為一個(gè)投資者利益的回歸 ——這是最重要的

54、。它的努力影響信息的精度，這是通過(guò)一個(gè)信號(hào)建模來(lái)實(shí)現(xiàn)的。精密和努力以及信號(hào)的實(shí)現(xiàn)是不可觀察的主體。當(dāng)市場(chǎng)的參與者有理性的預(yù)期時(shí)，一個(gè)線性工資的投資組合回報(bào)將鼓勵(lì)代理獲取更多的信息。由于資產(chǎn)價(jià)格對(duì)需求的依賴，影響經(jīng)理人獲得更精確的信息的激勵(lì)機(jī)制。</p><p>　　接下來(lái)，我們的目標(biāo)是連接這兩個(gè)派別，一個(gè)處理之間的信息和組織結(jié)構(gòu)類型之間的關(guān)系，以及最佳工資與其他交易在委托代理框架下的隱藏信息，重點(diǎn)是銀行風(fēng)險(xiǎn)管理的

55、組織過(guò)程。</p><p>　　信息的質(zhì)量是由銀行制定的決定其承擔(dān)風(fēng)險(xiǎn)大小的特點(diǎn)?，F(xiàn)有文獻(xiàn)對(duì)待硬信息和軟信息（彼得森，2004年）區(qū)別這個(gè)問(wèn)題中認(rèn)為，獲取這些信息可以通過(guò)兩種方式：貸款或交易關(guān)系的貸款。前者可以使用硬信息的處理的統(tǒng)計(jì)方法。這種類型的信息呈現(xiàn)諸多優(yōu)勢(shì)，成本低，具備規(guī)模效益和可以在可能范圍內(nèi)來(lái)衡量信用風(fēng)險(xiǎn)模型的風(fēng)險(xiǎn)值。另一方面，還獲得貸款提供的軟信息從而增加它對(duì)借款人素質(zhì)的預(yù)測(cè)估計(jì)精度，而且意味著將有

56、操作的問(wèn)題，因?yàn)檫@種類型的信息是無(wú)法證實(shí)的。</p><p>　　在這篇文章中，我們著重于信息在銀行信貸風(fēng)險(xiǎn)管理中扮演的角色。在一個(gè)隱藏的信息中，有一個(gè)銀行家需要資產(chǎn)回報(bào)的資料，用以管理信貸風(fēng)險(xiǎn)，并通過(guò)股權(quán)的分配風(fēng)險(xiǎn)與保險(xiǎn)道德風(fēng)險(xiǎn)建立委托代理模型，我們表明，通過(guò)其他信息，軟信息將節(jié)約股權(quán)，且使軟信息的精度更高。然而，這種類型的信息不被核查的，它需要實(shí)現(xiàn)一個(gè)特定的工資計(jì)劃以避免信貸人員進(jìn)行操作。</p>