風險薈訊
RMBI Newsletter
Financial Crime Risk :
Anti-Money Laundering Practices in Banking
金融犯罪風險:銀行反洗黑錢行動
What is Anti-Money Laundering?
To understand anti-money laundering, we have to understand what money laundering is. Money laundering is the process of converting illegal funds into seemingly legitimate assets with the purpose of concealing the ownership or original source of these funds. This makes it difficult for the authorities to trace the origins of the funds. To counter this, the banking sector has established a set of internal regulations and system known as anti-money laundering. These are legal controls taken by financial institutions to investigate suspicious transactions to help prevent money laundering activities within the banking sector.
甚麼是反洗黑錢行動？
要明白反洗黑錢行動，我們首先要明白甚麼是洗黑錢。洗黑錢是 指將恐怖活動犯罪、毒品犯罪、黑社會性質的有組織犯罪、走私 犯罪或者其他犯罪等違法所得及其產生的收益，透過各種手段掩 飾及隱瞞其來源和性質，使其在形式上合法化，其目的是使金錢 的來源難以追溯。為了打擊這種犯罪行為，銀行業訂立了一系列 的操作守則及程序，即反洗黑錢行動。金融機構會密切監察可疑 的資金流動，以打擊洗黑錢活動。Anti-Money Laundering has always been essential in our business world, especially the banking industry. To protect the interests of a country, regulations have been established for banks to follow to minimize opportunities for money laundering. To have a deeper understanding of this topic, I have much pleasure in introducing to you Mr. Harrison Ho, Managing Director of the Wealth Management Department at Nomura International (Hong Kong) Limited as our guest to provide us with more information on this topic.
Financial Crime Risk:
Anti-Money Laundering
金融犯罪風險
: 銀行反洗黑錢行動
The Rise of Text Mining in Financial
Markets
金融市場文字探勘的興起
反洗黑錢行動是商業世界不可避免的一個環節，特別是銀行業。為保護本國的利益，各國都透過相關法規約束銀行，藉 此減低洗黑錢的機會。我們很榮幸能邀請到野村國際(香港)有限公司的董事總經理何樂生先生跟我們分享更多有關反 洗黑錢行動的資訊，增加我們對這個議題的了解。
“ Anti-Money Laundering has always been
essential in our business world, especially
the banking industry.”
Responsible Parties 相關組織
In response to the money laundering activities within the asset management industry, several branches of the Hong Kong government have put in place certain rules and regulations which licensed financial institutions have to follow.
以下組織均在反洗黑錢行動中起了舉足輕重的作用：
FATF is an intergovernmental organization that sets the standard for money laundering measures globally and fights against terrorist financing. It is now one of the most influential organizations in anti-money laundering; it works hard to disseminate knowledge on anti-money laundering.
這是目前世界上最具影響力的國際反洗 黑錢和反恐融資領域的國際組織之一， 它設立了各種反洗黑錢的基準及相應的 措施，亦致力於宣傳反洗黑錢的知識。
Financial Action Task Force on
Money Laundering (FATF)
1國際金融反洗黑錢特別工作小組
Hong Kong Monetary Authority
(HKMA)
2香港金融管理局（金管局）
Joint Financial Intelligence Unit
(JFIU)
4聯合財富情報組
Securities and Futures Commission
(SFC)
3證券及期貨事務監察委員會（證
監會）
The HKMA supervises licensed institutions’ compliance with the legal and supervisory requirements as set out in the Anti-Money Laundering and Counter-Terrorist Financing (Financial Institutions) Ordinance (“AMLO”). It ensures that effective controls are established to detect and prevent money laundering activities.
The SFC provides guidelines on AMLO to assist licensed financial institutions in designing and implementing their own anti-money laundering and counter-terrorist financing policies, procedures and controls to ensure they comply with the AMLO and other relevant legal and regulatory requirements.
The JFIU is a department jointly operated by Hong Kong Police and the Customs and Excise Department, its main function is to receive and store the suspicious transactions reported by financial institutions and assist in the investigation of money laundering activities.
它負責監管認可機構的運作，確保其運 作符合《打擊洗黑錢及恐怖分子資金籌 集 (金融機構) 條例》，《打擊洗黑錢條例》 的守則，以及其管理的有效性以杜絕洗 黑錢。 它提供、制定和執行反洗黑錢法規的指引 ，以幫助持牌金融機構遵從《打擊洗黑錢 條例》及其他相關的法律和監管要求。 它隸屬於香港警務處刑事及保安處刑事 部毒品調查科，是由警務處和海關人員 組成的一個聯合單位，其作用在於接收 與分析金融機構提供的可疑交易報告， 從而調查洗黑錢活動。
1 http://www.fatf-gafi.org/about/ Financial Action Task Force (FATF)
2 http://www.hkma.gov.hk/eng/key-functions/banking-stability/aml-cft.shtml Hong Kong Monetary Authority, Anti-Money Laundering and
Counter-Fi-nancing of Terrorism (AML/CFT)
3 http://www.sfc.hk/web/EN/rule-book/anti-money-laundering-and-counter-terrorist-financing/ Securities & Futures Commission of Hong Kong,
Anti-money laundering & counter-terrorist financing
Future Trends
It is believed a new global standard on AML practice will be implemented. With better technology, money launderers are able to hide their illegal activities and their ways of money laundering become more difficult to be detected by World-Check, which is used to check if an entity or related stakeholders originate from a high risk country or are on the Politically Exposed Person (PEP) list that have been identified as risky. To combat this, tighter regulations are needed.
Activation of Anti-Money Laundering
In 2012, The Hongkong and Shanghai Banking Corporation (HSBC) was fined $1.92 billion by U.S. regulators, including the Office of the Comptroller of the Currency (OCC), the Federal Reserve (Fed), and the Treasury Department, as penalty for allowing criminals to launder $881 million through HSBC to Mexico which was classified as a high-risk country for money laundering and drug trafficking.5The massive fine has raised awareness among multinational banks of the importance of implementing anti-money laundering measures and the consequences of lack of AML controls. Since then, banks have increased their spending on anti-money laundering systems.
反洗黑錢行動的誘因
墨西哥是洗黑錢及毒品走私的高危國家。2012年，香港上海滙 豐銀行有限公司（滙豐）因縱容罪犯非法轉移 881億美元到墨 西哥賬戶，被貨幣監理局、美國聯邦儲備委員會（美聯儲）及 美國財政部等監管機構罰款1.92 億美元5。 巨額的罰款促使銀行加強監察洗黑錢行動，亦令它們明白到反 洗黑錢行動不力的嚴重後果，因此增加了在反洗黑錢行動方面 的投入。Know Your Customer (KYC), a standard form that banks and financial institutions use to collect detailed client information and assign risk categories, will be more frequently administered.
It has always been difficult for banks to track suspicious transactions with thousands of transactions conducted on any single day. Big data can be used to track suspicious customers and transaction records. This can greatly reduce the manpower needed to implement AML and can identify suspicious transactions more accurately.
Mobile technology can also help monitor customers’ actions. By identifying and classifying customers’ actions with the use of data mining technology, banks can single out suspicious customers before they engage in a transaction. This greatly reduces the chances that banks will be penalized due to money laundering. With the above trends, banks will be able to detect money launder-ing activities early, and thus reduce reputational risk and legal risk
未來趨勢
科技的發展會衍生出更難以發現的罪案，因此有理由相信反洗 黑錢行動守則及準則的嚴謹性及複雜性亦會隨之增加。科技的 進步令現時所採用的措施 (World-Check)，愈來愈難以檢查出高 危客戶。 World-Check，是利用詞庫去監察客戶公司是否創建 於高危地區或有關持份者是否屬於政治敏感人物的技術。 Know Your Customer (KYC) 是一套銀行和金融機構用來詳細列 明顧客資料及其風險等級報告的標準，銀行並將利用其進行更 頻繁的監察。 另外，隨著交易數量日益增多，要追查數量龐大的可疑交易， 必定要耗費大量人力物力。大數據的運用有助分析可疑的顧客 及其交易記錄，除了節省銀行資源，也能提升準確性。 流動通訊技術的普及有助於有效監察顧客的行為特性，透過數 據分析軟件的識別及分類後，銀行可以在可疑交易進行前先將 其鎖定，從而大大減低因疏忽而引致損失的機會。以上的趨勢 有助銀行預防洗黑錢的發生，穩定經濟。5 http://www.reuters.com/article/us-hsbc-probe-idUSBRE8BA05M20121211 Viswanatha, A., & Wolf, B. (2012, December 11). HSBC to pay $1.9 billion U.S.
Why this is Related to the Wealth Management
Industry ？
First of all, let us define wealth management. It is a high-end professional service that combines financial and investment advisory, accounting and tax services, retirement planning and legal or estate planning for one single set of fee6. Wealth managers help their clients’ coordinate
and manage their money and plan for their own or their family's current and future financial needs.
This type of service is mainly reserved for high networth individuals, they are people who would seek professional advice on managing their assets.
Wealth management is currently one of the major sources of revenue for many banks.
As we all know, due to the 2008 financial crisis, many banks, especially those in Europe, suffered huge losses from investments which severely weakened their capital base; their reserves fell to levels too low to sustain daily operations. Many investors were also hit hard by the crisis as they could not get their money deposited in the banks back7.
In order to regain their customers’ confidence, banks were fully aware that they needed to increase their capital reserves to support their financial operations. This pushed them to find new sources of revenue. Hence, wealth management, a service that can bring in many wealthy clients and considerable service fees, quickly became a new favorite of the banking industry.
As a result, banks are now handling more transactions for these wealthy customers and volumes are higher, so there is a much larger possibility and higher risk of money laundering activities in the banking industry.
為什麼與財富管理行業相關？
首先，讓我們定義什麼是財富管理。它結合了金融和投資顧問、 會計和稅務服務、退休計劃和法律或遺產規劃，是一種高端的專 業服務 6。財富管理機構會幫助客戶分配和管理資金，並因應客 戶自己或家庭的當前和未來的財務需要制定計劃。 這種服務的主要服務對象是高資產淨值人士。此類人士希望借 助專業建議去管理其大量的財富。 財富管理已成為許多銀行的主要盈利來源之一。 衆所周知，2008年的金融危機令許多銀行（特別是歐洲的銀行） 遭受了巨大的投資損失，從而缺乏資本來維持日常業務。很多 投資者不能收回以前存放於銀行的金錢，損失慘重 7。因此， 銀行深刻意識到，要重獲客戶的信心，他們需要提高資本儲備 水平以支持其業務。這促使他們開拓新的收入來源。財富管理 由於可以吸引許多富有的客戶和資金，因此迅速成為銀行業的 重點業務。 為此，銀行現在要替這些富有的客戶處理更多更大額的交易， 洗黑錢活動的可能性和風險亦因而相應增加。 6 http://www.investopedia.com/terms/w/wealthmanagement.asp 7 https://en.wikipedia.org/wiki/Wealth_management 8 http://www.nomuraholdings.com/services/retail/wealth_management.htmlNomura as an Example
In this article, we introduce Nomura Group and show readers how Nomura’s wealth management department has adapted to this tough regulatory environment.
Nomura Group is one of the biggest investment banks in the world. Founded in Japan in 1925 as an investment bank, the group now has over 150 retail branches in Japan alone, making it the biggest retail bank in Japan. Also, it has more than 90 years of banking and related experiences with over 28,000 employees worldwide, with assets under management of over 100 trillion yen8. The reason we
choose Nomura is because we want to use a large and well-established bank as an example to assess the effects of AML practice on the banking industry in general.
野村集團
本文除了介紹野村集團外，也會闡釋野村的財富管理部門如何 適應這種艱難的環境。 野村集團是一家於1925年成立的日本投資銀行，亦是全球最大 的投資銀行之一。集團現時在日本擁有超過150個零售分行， 是日本最大的零售銀行。集團從事銀行業逾90年，在全球擁有 超過28,000名員工，所管理的資產規模超過100萬億日元 8。1. AML practice in wealth management ：
opportunities and threats
Opportunities: The industry has become more regulated and
transpar-ent. This implies there will be more guidelines and regulations that banks have to follow.
A pro-active internal AML procedure of banks can ensure the operation can be conducted in a sound, effective and efficient manner to deal with any possible risk exposed to the firm. It has become clear which types of business are legitimate and which are not. Banks can then reduce their potential risk in undertaking suspicious deals or transactions. The potential damage to the society and reputation of the banks will be minimized when the business or industry is more alert and self-regulated.
Threats: It may shut the wealth management industry out of
certain business opportunities, because nowadays the industry has become more regulated and there now appears to be a clearer understanding of what can be done and what cannot. Some potential clients may not be willing to do business with banks anymore due to
1. 財富管理中的反洗黑錢行動：利與弊
利：財富管理行業變得更加規範和透明，這意味著相關監管機 構將對行業施加更多的指引和法規。銀行內部積極的反洗黑錢 程序可以確保反洗黑錢行動以更合理和高效的方式進行操作， 以應對任何可能存在的風險。對於銀行來說，現在他們很明確 地知道哪些業務可以做，哪些不能做，因此銀行可以減少交易 的風險或與可疑客戶交易的潛在損失。當企業或行業更加警惕 和自我監管時，對社會的潛在損害和對銀行的聲譽的影響將被 最小化。 弊：監管越來越嚴格，加上有更清晰的指引說明什麽可以做和什 麼不能做，有可能會扼殺了財富管理行業中的一些業務。此外， 由於監管收緊，一些潛在的客戶可能不願意與銀行做生意，因為 開戶過程複雜，例如非常全面的審查程序和客戶背景調查。銀行 需要更謹慎地檢查客戶的背景，以及貨幣的來源或去向。銀行會 加強留意有政治背景的人，例如政界人士。 9 http://www.nomuraholdings.com/news/nr/asia/20070416/20070416.html 野村是一間運作完善的大型銀行，選擇其作為例子可讓大家更 容易去瞭解反洗黑錢行動對整個銀行業的影響。 野村財富管理隸屬銀行的零售部門。與其他私人銀行或大型銀 行的財富管理部門不同的是，它為高資產淨值人士、富裕家庭 以及大公司等不同類型的客戶提供私人定制的方案，並提供廣 泛的產品和服務，幫助客戶把握不同資產類別的本地投資機會 。 2007年，隨著財富管理業務需求日益增長，野村證券決定在 日本以外的地區擴展其財富管理業務，以滿足不同資產類別的 投資和財務需求9。當然，這也意味著銀行更易受到金融犯罪的 傷害。Industry Insight 行內洞悉
We arranged a private meeting with the managing director of the wealth management department of Nomura, Mr. Harrison Ho, to find out how the enforcement of AML practice in banks has affected the development of the wealth management industry. Mr. Ho have been working in this industry for over thirty years, we asked him to share his personal experience and views of it.
我們與野村財富管理部門的董事總經理何先生進行了面談，以 了解銀行中反洗黑錢行動如何影響財富管理部門的發展，特別 是在野村。
Nomura wealth management is under its retail division from Japan, it differentiates itself from other private banks or wealth management divisions of large banks by delivering bespoke solutions to different kinds of clients, including high-net-worth individuals, wealthy families as well as large corporations. It provides a wide range of products and services to help its clients access local investment opportunities across different asset classes. In 2007, Nomura decided to further expand its wealth management business outside of Japan, which were showing increasing needs for wealth management services that cater to the investment and financial needs of high-net-worth individuals and families for different asset classes9.
Of course, this implies a higher probability of dubious transactions or even money laundering activities in the bank.
Interview with Mr. Harrison Ho for giving us insights into Anti-Money Laundering.
the complicated account opening process, such as a very comprehensive KYC procedure and background checks and proves with clients. Banks are paying more attention to the backgrounds of their clients as well as where their money is coming from or going to, e.g. they are more aware of people with certain political backgrounds such as PEP, which stands for Politically Exposed Persons.
2. How will the daily operation of the department
change?
In order to meet the rising standard of the requirements by the authorities, more resources are allocated to align the firm’s internal control with the compliance standard required by local authorities, such as SFC and HKMA. More and more new hires from banks for legal, compliance, risk and control function area.
The AML practice is not only for bank's supporting and controlling functions, but also for their sales forces. Relationship managers now have more responsibilities too. Relationship managers are required to go through a more complicated operating practice requirements, which adopted to comply with AML standards. For instance, they need to conduct a more comprehensive KYC procedure for new client account opening, stricter pre-trade censoring process, more frequent clients' meeting call reports etc.
3. How does Nomura wealth management adapt to this
tightened environment?
A seamless and more compliance-focused sales and control procedure has been set up to adapt to this regulatory environment. Employees are given clearer procedures to follow and more compliance checking is required from time to time during the sales process. Authorities, like SFC, will conduct regular onsite examinations on the bank to check whether a comprehensive operating and sales procedure has been established and followed. Staff training will become more frequent to ensure employees have incorporated a compliance-focused culture in their daily operations. After implementing these changes within the department, more time and resources are needed as the heightened reporting and checking requirements have increased the daily workload of staff from different positions, such as operations, sales and supporting teams.
2. 部門的日常運作有何變化？
為了滿足當局對標準日益提高的要求，公司分配更多的資源聘 請從事法律，合規，風險和控制功能領域的新員工，使公司的 內部監管與證監會和金管局等地方當局要求的合規標準相一致 。 反洗黑錢行動不僅是為了銀行的支援和監管功能，亦是為了他們 的銷售隊伍。關係經理現在也有更大的責任，他們需要遵從更為 複雜的操作實踐要求，這些要求亦遵從反洗黑錢的標準。例如， 他們需要對新的客戶進行更全面的開戶前審查程序，更嚴格的交 易前審查程序和更頻繁的客戶會議通知報告等。3. 野村財富管理如何適應監管收緊的環境？
為了適應這種嚴格的環境，野村財富管理建立了一個更加流暢和 注重合規性的銷售程序。管理層必須向員工提供更清晰的程序， 以便他們跟隨，並在銷售過程中不時進行檢查。監管機構，如證 監會會定期對銀行的銷售程序進行突擊檢查，因此銀行必須定期 進行員工培訓，以確保員工在日常運作中熟悉反洗黑錢條例的細 節。在部門內實施這些更改後，將會耗費更多的時間和資源，因 為更多的報告和檢查工作增加了不同職位（如運營，銷售和支持 團隊）的人員的日常工作量。4. What is the most complicated case you have
encountered?
The implementation of AML practice in the financial industry may lead to the disappearance of some potential businesses opportunities. Banks are required to strictly follow the new AML standard and this may eliminate all the possible businesses from high risk area. Banks have to be more aware of the business it is dealing with and the source of funds from customers. They became very alert with some businesses involved in high risk industries and some high risk countries; such as casino & entertainment businesses, jewellery dealing and trading, shipping and some other cash transactions related businesses.
Besides, banks avoid to deal with some risky region, such as some Middle East countries, which is more susceptible to money laundering activities as it is very hard to trace the source of the money.
Complying with the Anti-Money Laundering Ordinance, Nomura is responsible for checking the source as well as the destination of the money in any transactions, so as to prevent money laundering from taking place, avoid putting the company at risk. Taking a proactive and strict AML approach, Nomura was able to retain its most prestigious clients and maintain good relationships with authorities in order to maintain its strong position in the wealth management industry.
Summary
To conclude, the tightened regulatory environment has led businesses to undergo a major transformation. However, the guidelines and procedures which financial institutions have to follow are clearer, firms now have a better idea of what they can and cannot do. Banks and their wealth management departments are adapting to this environment after severe fines are imposed by regulatory authorities. In the future, we shall see how banks and financial institutions are employing digital technology to enhance the regulations of compliance and AML practice, and transforming their procedures to prevent themselves from suffering heavy monetary sanctions. We assume there will continuously be a huge demand for talents in the fields of regulation and compliance in the financial sector.
總結
總而言之，隨著監管環境越來越嚴格，企業經歷了艱難的轉型。 然而，這些監管措施為金融機構提供了一個更清晰的指引和銷 售程序，向所有公司明確說明什麽應該做和什麽不應該做。銀 行及其財富管理部門在相關監管機構實施嚴厲懲罰措施後已經 適應了這種環境。今後，我們將看到銀行和金融機構採用先進 的數碼技術促進其在法規和反洗黑錢領域的創新，以防受到金 錢制裁。我們認為，在金融市場的監管和合規領域，法規和風 險管理人才的需求將會持續強勁。 Text 撰文TSANG Chiu Yu, Derek 曾昭宇 WONG Ching Ip, Venice 黃靖燁
Students (Year 3) of Risk Management & Business Intelligence Program 風險管理及商業智能學課程三年級學生
4.你以前遇到過最複雜的情況是什麼？
在金融行業實施反洗黑錢行動，會令一些潛在的企業機會消失了 。銀行必須嚴格遵守新的“反洗黑錢”標準，消除了所有可能的 高風險地區的業務。 銀行對其所處理的業務和客戶的資金來源提高警覺。他們對涉 及高風險行業的一些企業和一些高風險國家變得非常警惕; 例如 賭場娛樂業，珠寶交易，航運等一些現金交易相關業務。 此外，銀行避免處理一些高風險地區的交易，例如一些更易受 洗錢活動影響的中東地區，因為很難追查資金來源。 根據《反洗黑錢條例》，野村負責在任何交易中檢查貨幣的來 源和目的地，以防止洗黑錢活動和其為公司帶來的風險。雖然 野村實行了更嚴格的反洗黑錢操作守則，但仍然保留其最有價 值的客戶，並與他們保持良好的關係，以保持其在財富管理行 業的地位。1 To further understand the current adoption of Text Mining processing, we have invited Mr. Timothy Li, Team Leader of the Earning Estimates Team of the
Global Data Department at Bloomberg LP to share with us how a financial data provider provides accurate and timely data, which gives us a good insight into the background of Text Mining
我們邀請了Bloomberg LP 全球數據部門收入預測團隊負責人Timothy Li 先生和我們分享金融數據供應商如何收集並為客戶提供準確的文字數據。 以此為契機提供了 寶貴的意見，從而想出了介紹文字探勘的背景及發展。 數據充斥著整個世界及金融市場，我們每天製造的資訊量約達2.5萬億位元組，而金融分析師平均每天發佈的報告亦達4000份，其 中包括53種語言，總頁數達到36000頁。隨著市場參與者試圖透過越來越多的資訊提升競爭優勢，金融市場亦發現其有必要利用大 數據的力量，並形成了自己獲取、分析和利用數據的方式。最近一個新趨勢是市場文字探勘。因此，這篇文章會集中討論現今文 字探勘科技的發展及應用。 1 文字探勘是從文字中導出高質量資訊的過程。人們通常透過統計模式學習等手段來發現模式和趨勢，從而導出高質量資訊。金融 業典型的文字探勘應用包括文字分類、文字聚類、情緒分析和文件概要。文字探勘涉及以下過程：資訊檢索、 語言分析、命名實 體識別和模式導出。
A preparatory but essential step to acquire data. Researchers carefully identify the data source as data are collected in various formats, with varying levels of accuracy and repetition. A single article may contain a variety of information, which would affect decision making differently. The information retrieved should also be validated by different means such as manual eyeball checking, dual processing and criteria matching. In addition, the cost of quality assurance should always be weighed against the return on higher quality data.
它是獲取數據的一個必要步驟。由於數據會 以各種格式存在、準確度不一，並且可能相 互重複，因此研究人員需要小心識別數據來 源。一篇文章可能包含多方面資訊，並會影 響決策的制定，所檢索的資訊也應透過不同 的方法進行驗證，包括眼球手動檢查、雙重 處理和標準匹配等，保證成帶來較高質量的 數據。
Information retrieval
資訊檢索
The Rise of Text Mining in Financial Markets
金融市場文字探勘的興起
The world is awash in data. Financial markets are awash in data. We are generating around 2.5 quintillion (2.5×1018 ) bytes of information
every day, and there is an average of 4,000 brokerage reports a day comprising around 36,000 pages in 53 languages. As market partici-pants try to maximize their competitive edge from the growing mountain of information, the financial world increasingly feels there is a need to harness the power of big data and it has been shaping the way they acquire, analyze and utilize data. The recent development is the rapid expansion of text mining. Hence, this article will focus on the development of Text Mining technology as well as Text Mining technique.
1Text mining is the process of deriving high-quality information
from text. High-quality information is typically derived through discovering patterns and trends through techniques such as statistical pattern learning. Typical text mining applications in finance include text categorization, text clustering, sentiment analysis and document summarization. Text mining typically involves the following processes: information retrieval, linguistic analysis, named entity recognition and pattern derivation.
“ The world is awash in data.
Current adoption of text mining:
How ways to acquire, analyze and utilize data
have changed.
The sources from which data can be acquired are expanding. Market participants no longer restrict themselves to analysts’ reports for gold digging, but to more unconventional forms of texts ranging from broader news, viewer comments and social media. For instance, associate professor Alex Edmans of the University of Pennsylvania’s Wharton School of Business has used the textual data from FORTUNE Magazine’s “100 Best Companies to Work for in America” as a proxy for employee satisfaction and found a positive correlation between employee satisfaction and stock out-performance2. Traders and investors have since retrieved relevant
employee feedback from job sites and are looking to expand their data sources to include social media, blogs and chat rooms. The ways to utilize data has also changed, in the sense that humans are no longer the only ones analyzing or making trading or investment decisions with the given data. Financial markets have witnessed a rapid growth in algorithm trading, which is becoming increasingly efficient since automated trading algorithm can read larger amounts of data and achieve a higher market coverage at a much higher speed.
現時的文字探勘：
獲取、分析和利用數據的方式有何改變？
可以獲取數據的來源正在不斷擴大。市場參與者毋需再局限於 分析師的報告，而可以參考更多非常規形式的文字，例如各類 新聞、觀眾評論和社交媒體等。賓夕法尼亞大學沃頓商學院 (Pennsylvania’s Wharton School of Business) 亞歷克斯·埃德曼 斯教授以《財富》雜誌中「美國100家最佳僱主企業」的文字 數據作為員工滿意度的副本，發現員工滿意程度和股票表現之 間存在著正面的關聯性2。交易者和投資者已經從作業現場獲取 相關員工的反饋，並打算將數據來源擴大到社交媒體、博客和 聊天室等。 運用數據的方式也改變了，利用數據作分析或作出交易決策的 投資者，已不再僅限於人類。金融市場已經見證了算法交易 (algorithm trading) 的快速增長，由於自動交易算法可以更高速度讀取更 大量和更高市場覆蓋率的數據，因此算法交易變得越來越高效 。2 Why Employee Satisfaction Matters to Shareholders - http://knowledge.wharton.upenn.edu/article/why-employee-satisfaction-matters-shareholders/
Linguistic analysis
語言分析
Pattern derivation
模式導出
Named entity recognition
命名實體識別
The application of extensive natural language processing, such as part of speech tagging and synthetic parsing, to understand the textual data acquired.
The process of applying gazetteers or statistical techniques to identify named text features, such as people, organizations, stock ticker symbols, and so on. Disambiguation may be required to narrow down the meaning of words; for instance, "Ford" can refer to a former U.S. president, a vehicle manufacturer or some other entity.
A process that depends on the final task of text mining. It typically involves identifying association between entities and other information in the text as well as extracting a conclusive statement about the entity from the text.
它是應用廣泛的自然語言處理，例如部 分詞性標註 (speech tagging) 和語法分析 (synthetic parsing) 幫助機器理解所獲取 的文字數據。 它是應用詞典(Gaztteers)或統計技術來 識別命名文字特徵的過程，例如人、組 織及股票代號等等。這可能需要消除字 句含糊意思以縮小詞語的含義，例如， 「福特」可以是指美國前總統、汽車製 造商或其他某些實體。 它是依賴文字探勘最終任務的過程。它 通常涉及識別實體和文字中資訊之間的 關聯，以及從文字中提取關於實體的結 論性陳述。
As financial markets are extremely sensitive to breaking news, machines are now programmed to read from news and other sources, capturing keywords such as mergers and acquisitions, stock splits, annual returns and dividend announcements much faster than humans can. Machines have also been fed algorithms extensively for more sophisticated trading decision making. Therefore, responding to market events a few milliseconds faster than competitors could mean better prices and improved profitability for traders, this heightens the importance of timely data for those working in the finance sector.
Consequently, changing the ways we interact with data will change the other processes. For example, the fact that algorithms are increasingly reading the data demands higher quality and more intelligent algorithm designs. This has prompted the development of products such as Bloomberg News Events Feeds, which is a highly structured, real-time feed specifically designed to deliver machine-readable news and it has changed how data are acquired. The News Events Feeds adopt rule-based data mining and computers are programmed to scan text for patterns of language that match certain event types. These events are specific to an entity, a place, or an organization, for example, the Fed or European Central Bank (ECB), or a tradable instrument. The data provided also include a sentiment score that allows clients to trade directionally, e.g. earth-quakes are negative for an economy, higher unemployment guidance is negative, higher GDP guidance is positive and so on. Computer programs are also increasingly tasked with analyzing textual data, an area which used to be the purview of humans. This greatly enhances the efficiency of trading decision making.
另外，由於金融市場對突發新聞非常敏感，因此機器現已被教 會從新聞和其他來源讀取數據及捕獲關鍵字，如併購、股票分 拆、年度收入和股息公告等，速度更遠勝人類。機器也更被廣 泛地用於更複雜的交易決策算法， 因此，比競爭對手快幾毫秒 回應市場事件便意味著為交易者提供更好的價格和提升其盈利 能力，從而突顯即時檢索數據對金融從業者的重要性。 改變與數據互動的方式將改變其他科技進程，例如，算法機器 讀取更多的數據需要更高質量和更高智能的算法設計，這促進 了Bloomberg News Events Feeds 產品的開發。這是一種高度結 構化及提供實時報告的軟件，專門提供機器可讀取的新聞和改 變數據的獲取方式。News Events Feeds 採用規則性數據探勘 （Rule-based Data mining ），電腦則被設置成掃描符合某些事 件類型語言模式的文字。事件明確針對實體、地點或組織，如 美聯儲或歐洲中央銀行(ECB)或交易工具，它提供的數據還包括 客戶交易的情緒分數，例如地震對一個經濟體是負面的，更高 的失業率是負面的，GDP增長是正面的等等，這大大提高了交易 決策的效率。
Future developments
Market giants Google, Facebook, and Apple are spending huge amounts of money every year acquiring smaller firms for market entry, strategic intellectual property and talented employees. However, they all face the same challenge in identifying companies with the right products and technologies in the vast startup universe. Text mining in the future can provide entrepreneurs, venture capitalists and analysts with a new way to navigate the constantly changing landscape of mergers and acquisitions.
未來發展
谷歌、Facebook和蘋果等市場巨頭每年都花費巨額資金收購小 型公司，以拓展新市場、吸納知識產權和人才，但事實上，他 們都面臨著同樣的挑戰：如何在廣大的創業市場物色具有合適 產品和技術的公司。未來的文字探勘可以為企業家、資本家和 分析師提供一個新的途徑去了解日新月異的併購平台。“ Market participants no longer restrict
themselves to analysts’ reports for gold digging,
but to more unconventional forms of texts ... ”
Interview with Mr. Timnothy Li who gave us insights into Text Mining.
Text 撰文
CHIU Hok He, Angus 趙學希 LI Chin Wa, Chin 李芊嬅
Students (Year 3) of Risk Management & Business Intelligence Program 風險管理及商業智能學課程三年級學生
3 UTA researchers use text-mining techniques to improve market intelligence on potential mergers and acquisitions of startups in the high-tech sector -
https://www.uta.edu/news/releases/2016/07/Moo-text-mining.php
4 Jim Clifton- American Entrepreneurship: Dead or Alive?-http://www.gallup.com/businessjournal/180431/american-entrepreneurship-dead-alive.aspx
Gene Moo Lee, a researcher at The University of Texas, has created a new method called Topic Modelling3 which uses big data analytics
and text-mining techniques to improve market intelligence and explain potential mergers and acquisitions of startup companies. This text-mining technique analyzes the language used in the startups’ business descriptions about their products, technologies and the market. Then the likelihood of a possible merger between two companies will be computed, taking into account business proximity, geographic vicinity, and social links between individuals within the two firms as well as common investor ownership to reflect the strongly networked nature of the startup world.
In fact, according to a U.S. Census Bureau report4, there are 400,000
new businesses each year in the US, but at the same time 470,000 go out of business. In particular, during 2016 the number of startups receiving funding, as well as the number of funding commitments, dropped by 11% compared with 2015. This new text-mining technology is a great opportunity for startups to seek venture capital investment. Li said “Our system is particularly appropriated when the firms under study are small and privately held so industry classification in largely unavailable, which is the case for startups.” Thus, if this Topic Model-ling technology is adopted by the market, it will provide a new channel for market giants and startups to engage in mergers and acquisitions.
Topic Modelling is just a sneak peek of the possible wide impact that text mining or analytics has on innovation systems, market efficiency and even the economy by uncovering hidden commercial, economic and societal values from data. This is where text mining techniques find its value and we believe it will become more and more important in business analytics.
德克薩斯大學（The University of Texas）研究員 Gene Moo Lee 開發了一種稱為主題建模（Topic Modelling）的新技術 3 ，該 方法使用大數據分析和文字探勘技術來加強市場情報及分析初 創企業的併購潛力。這種文字探勘技術分析了初創公司的產品 、技術和市場的業務描述，然後考慮兩間公司之間的商業關聯 性、地理位置、社會聯繫以及共同投資者的股權，以計算兩間 公司合併的可能性，這反映了創業領域強大的聯網性質。 事實上，根據美國人口普查局的報告4，每年美國有40萬個初 創公司成立，但同時有47萬個步入清盤。在2016年，接受融資 的創業公司數量比2015年下降了11％。這種新的文字探勘技術 為尋求資本投資的初創公司提供了一個很好的機會。 Li 說：「 我們的系統是特別適合小型和私人持有的公司，因為行業分類 在很大程度上不適用以上的公司 。」因此，如果市場採用主題 建模技術，它將會成為市場巨頭和創業公司合併和收購的新渠 道。 通過主題建模的例子，我們看到文字探勘和分析可以揭示數據 中隱藏的商業、經濟和社會價值，對創新系統、市場效率甚至 經濟產生廣泛影響。我們相信文字分析將在商業和金融業務分 析中變得越來越重要。
Prof. Lancelot JAMES (Program Co-director) 在林壽教授 Prof. Xianhua PENG (Program Co-director) 彭獻華教授
Dr. Jean Wang (Undergraduate Program Coordinator) 王繼英博士 Some photos in the newsletter are from
http://www.unsplash.com http://www.compliancetrendz.com http://www.thestaffingstream.com http://alumasso.hkbu.edu.hk http://www.nomura.com http://infonegocios.com.py http://www.servicetrac.com
Risk Management and Business Intelligence Program The Hong Kong University of Science and Technology Phone: 3469 2399 Fax: 3104 0026
Email: [email protected]
Website: http://www.rmbi.ust.hk
All contents and information are subjected to copyright protection. Republication, redistribution or unauthorized use of any content is expressly prohibited without the prior written consent.
