信用評分工具：自動化信用管理的理論與實踐

內容簡介

《信用評分工具：自動化信用管理的理論與府她歸實踐》區別於其他學者的著作，對抽象理論只是點到為止，更多的則是著眼於客觀實際情況，對行業實踐操作進行系統梳理，因此具有極強的現實指導作用。本書作者恰好在南非生活和工作，以其30多年的銀行風險管理的經驗為開發中國家實現信用評估和信用管理現代化提供了寶貴的經驗。

內容概述

英文版敘

When I first learnt my book would be translated into Mandarin I was surprised, and honoured. I had visited China twice as a tourist with family, and was fascinated by the country. The first was in 2000, when infrastructure investment was driving economic growth. We saw massive undertakings, including the Three Gorges Dam while still under construction. And yet, the bulk of the population still lived off the land. My sense was that we were regarded as‘laowai’, resented because of the last two centuries’history by a proud people who felt that‘the 21st century will be ours’.

Our second trip was in 2009, one year after the Beijing Olympics. Massive urbanisation had occurred in those few years, and the trappings of the consumer economy were evident—at least in the cities. The sense of animosity had disappeared; we felt welcomed by a people who felt that better future had arrived. In particular, we were invited on countless occasions to be in people’s photographs—their cameras being obvious fixtures of a consumer society.

As for this book, I started writing it in 2003 when credit scoring had a history of

some forty years. A tool first used to assess applications for instalment finance and credit cards in the United States, had over time spread to more products, geographical regions, and stages in the credit-risk lifecycle. In recent years behavioural scores

have been co-opted by Basel II and IFRS 9 in attempts to improve the stability of the financial system—at least for retail credit. An operations research tool intended to improve lenders’decision-making and operating efficiencies now aids capital adequacy and loss-provisioning assessments. Its biggest benefit is its ranking ability (predictive power), but finance people want reliable default-rate estimates (predictive accuracy). Unfortunately, default rates are heavily influenced by data not available within the scoring assessment—economic, market, industry, and competitive factors—which makes the scores key inputs into more complex algorithms.

Several factors aided credit scoring’s success, some of which occurred long before it emerged. First, the industrial revolution that provided goods for people to buy—with producers and retailers offering credit to foster sales. Second, a collaborative revolution as information markets (e.g. credit bureaux) enabled them to determine which prospective customers had honoured or dishonoured their obligations elsewhere—which over time extended to banks and other more traditional lenders. And third, the technological revolution that enabled the compilation and assessment of the available data. China is now well into its industrial revolution, the technological revolution is under way, and the collaborative revolution is evolving.

Credit scoring’s primary benefits are its speed, consistency, and cost effectiveness. It is not a panacea though; it is hard work, success is not guaranteed, and there is the potential for mistakes along the way. Indeed, the mortgage crisis of 2008 can at least partially be blamed on lenders’ overconfidence in models developed in a benevolent economy. The biggest challenge is data, both quality and quantity, which requires a significant infrastructure investment. If China is to become a consumer-driven economy, much investment is required to enable proper risk assessment.

Credit can be a significant driving force for improving people’s lives, if offered and used judiciously. It is best for financing business ventures, education, and devices that free up people’s time for more productive activities. When used heavily for purchase of consumer-lifestyle goods, the economic benefits can be offset by the stresses put onto individuals. I live in South Africa, where excessive credit—and problematic collections practices—led to huge social pressures and increased wage demands when workers had too much month left at the end of the money. Irresponsible lending an irresponsible borrowing are often bedfellows, only the former has the upper hand and greed is a major driving force.

As it stands in 2016, China has a population of 1.4 billion, of which about 900 million are covered by the PBC’s credit bureau. Some 400 million had loan histories, and 110 million had outstanding credit from a financial institution. Those are small proportions when compared to developed economies. Savings rates of 40 percent are extremely high, not only for historical and cultural reasons, but also because credit is unavailable.

A major difference between China and the west is that ‘store credit’ is not being driven by bricks and mortar stores, but by online retailers capitalising upon their vast data. Some derive their own credit scores, but without proper data sharing—especially ‘positive data’ —and the scores derived by each vary greatly. In an ideal world, lenders should be working from the same data to derive risk assessments specific to their own circumstances. That said, there will always be peculiarities that cannot be accommodated within the standard format. The collaborative revolution needs to be carried further. I have read with interest developments relating to a ‘social credit score’. Typically, credit scores predict a credit-related outcome (non-payment, attrition, propensity, etc.). The social score sounds more like a balanced scorecard for citizens with creditworthiness just one of many factors, so the name is a bit of a misnomer—it is not really a credit score. That said, the behaviours it measures will have a correlation with ‘willingness to pay’, so may provide value when assessing those currently unable to access credit.

And finally, I hope you enjoy reading my book, and get a lot of value out of it. It was intended as course notes, but grew into a book over three-and-a-half years. Each time I turned a corner I realised how little I knew, and how much more I needed to learn. For that matter, I am still learning. I trust that you will gain benefit from the early part of my journey!

Raymond Anderson

August 2017

譯者序

這是迄今為止非常詳盡地介紹信用評分的一本著作。它是商業銀行和貸款機構半個多世紀以來在貸款技術和風險管理方面全面的經驗總結，稱其為信用評分領域的“百科全書”毫不為過。

自20 世紀60 年代理論基礎和70 年代計算能力有了重大突破以來，自動化信貸審批和風險管理逐漸成為主流。更豐富的數據、更科學的方法、更完善的流程所帶來的風險管理技術革新，已滿足了億萬個人和企業的融資需求，促進了歐美經濟幾十年的高速增長。國內的自動化信用管理和大數據風險評估才剛剛起步，尚在混沌和懵懂之中摸索。以其龐大的人口基數和經濟總量，中國理應成為個人消費信貸和中小微企業融資蓬勃發展的藍海。因此，可以預見，消費升級將帶來中國經濟的再次騰飛，而消費信貸則是其中最重要的支撐點和爆發點。行業代表已提前行動，大批消費金融、小額貸款和金融科技公司崛起，傳統商業銀行開始轉型，許多財力雄厚的集團企業也紛紛成立消費金融和小微金融事業部，依靠其豐富的數據積累開始做線上、線下的貸款業務。然而，如火如荼“跑馬圈地”的背後卻是單一的徵信體系、缺失的制度保障和薄弱的風控技術，我們還有很長的路要走。

本書副標題是“自動化信用管理的理論與實踐”。事實上，“自動化”在數據時代和移動互聯的背景下已升級為“現代化”，使決策更科學、更智慧型。“信用管理”不僅是風險控制，還有與此相關的市場行銷、欺詐識別和賬戶管理。“理論”與“實踐”也讓本書區別於其他學術著作和大眾讀物。本書不僅蒐集、引用大量理論，而且著眼於實際情況，對行業實踐進行系統梳理，為實際套用服務。本書作者恰好在南非生活和工作，以其30 多年的銀行風險管理的經驗，為開發中國家實現信用評估和信用管理現代化提供了寶貴的經驗。

深入理解信用評分，它有如下幾層意義：（1）交易層面，準確判定好壞，預測違約風險；（2）個體層面，考慮客戶關係周期，實現利潤、利益最大化；（3）公司層面，適應經濟周期，分散資產風險，維護金融系統穩定；（4）社會層面，減少不公正歧視，保護弱勢群體，提高信貸可得性，增加社會福祉。第一點和第四點看似矛盾，但人類文明的發展權由大家共同享有，個人借錢改善生活，企業融資擴大經營，和諧社會應給予缺少徵信記錄、具有較少和負面徵信記錄的個體公平的對待。“Access to Credit”既是一種願望，也是一種權利。信用評分恰如其分地在微觀層面增加基尼係數（區分能力）和巨觀層面降低基尼係數（貧富差距）上實現了統一。

本書系統完整：第一部分介紹信用評分的背景和意義，第二部分介紹風險管理理論和套用，第三部分介紹數學和統計方法，第四部分介紹數據和信息來源，第五部分介紹建模和開發過程，第六部分介紹實施和使用，第七部分介紹信用風險管理周期，第八部分介紹監管體系和制度。本書對在校專業學生、建模開發人員、行業從業人員、專家研究人員、公司領導者、政府監管者、政策制定者，特別是對進入和正在進入這個行業的公司，都有很強的指導意義。

信用時代已經拉開帷幕，我們身處其中，每個人的生活都受到它的影響，每個人也都在為創造一個誠實守信的社會貢獻力量。願每個人都珍惜信用、積累信用、享受信用！

歡迎來到信用評分的世界！

西南財經大學金融學院信用管理系

李志勇

作者簡介

雷蒙·安德森（Raymond Anderson），加拿大人，常住南非，獲得工商管理學士和碩士學位，有30多年銀行從業經驗，從1996年開始參與信用評分系統開發，曾在渣打銀行等多個跨國金融機構擔任風險模型部高級經理，參與過十多個國家的信貸工廠建設，在小額貸款、農村貸款、小微企業貸款等次級市場、邊緣市場和新興市場的信貸決策支持系統開發上有豐富的經驗，擅長用各類程式語言建模和分析。

譯者介紹

李志勇，西南財經大學金融學院信用管理系副教授，英國愛丁堡大學商學院信用研究中心博士，Credit.Li創始人，致力於發展信用風險管理方法和信用評分建模技術、培養風控人才和團隊，專注研究消費金融和個人信用問題。

信用評分工具：自動化信用管理的理論與實踐

基本介紹

內容簡介

目錄

內容概述

英文版敘

譯者序

作者簡介

譯者介紹

內容概述

英文版敘

譯者序

作者簡介

譯者介紹

相關詞條

熱門詞條