合规、内控、风险管理 Compliance\Internal Control\Operation Risk Management

目前洗钱的主要手段

1.古董买卖———
　 你买了40只明代的花瓶，每个5000美元。你把它们都拿出来拍卖，把这些藏品分别存放在12家拍卖行内，最好分布在12个不同的城市。当一只花瓶被拿出来拍卖时，你或是把它卖给出价最高的人，或是派你最亲近的叔叔去拍卖行把它买回来。
人们买你的每件东西一般都是用支票付款，所以，当你的叔叔为买其中的一个花瓶付给拍卖行5500美元现金时（即5000美元的拍卖价加上向买主收取的10％的手续费），他们会给你一张4500美元的支票，即拍卖价减去向卖主收取的10％的手续费。
这些费用作为洗钱的开支完全可以被一笔勾销。幸运的是，你还可以把这个花瓶再次拍卖。当然，做这档生意的前提是：你多少得有点没落贵族的做派。

2.寿险交易———
保险“洗钱”主要集中在寿险领域，尤其是在团体寿险中。通过长险短做、趸交即领、团险个做等不正常的投保、退保方式，达到将集体的、国家的公款转入单位“小金库”、化为个人私款或逃避纳税的目的。

投保时别忘了“为职工谋福利”、“合理避税”这些名目———将一部分通过正常财务途径无法转化为个人收入，或直接发放须缴纳高额个人所得税的资金，通过为其职工投保寿险，然后退保，取得退保金这种途径，化公为私或逃避纳税。

而有的保险公司也明知其投保目的不纯，却积极地促成这类“保单”，有些业务员甚至以此为诱饵，为业务对象出谋划策，主动提供“洗钱”方便。他们多拉了业务，多拿了费用；而“以规模论英雄”的保险公司，则在短期内完成了任务指标，还能从“退保”中扣下一笔手续费。 因此，保险“洗钱”更像个令人不齿的多边交易。

3.海外投资———

经典做法是：进口时，高报进口设备和原材料的价格，以高比例佣金、折扣等形式支付给国外供货商，然后从其手中拿回扣、分赃款，并将非法所得留存在国外；出口时，则大肆压低出口商品的价格，或采用发票金额远远低于实际交易额的方式，将货款差额由国外进口商存入出口商在国外的账户。如果胆子足够大，可以考虑在境外银行直接建有个人账户。

如此一来，这样的情景便不令人吃惊了：某些国有企业的海外分支机构已经演变为专业的洗钱中心。另外，不妨通过代理人或移民海外的子女、亲属，用办皮包公司的办法把黑钱洗白。如果主管部门派人去查，将会得到一些令人神伤的亏损理由。接下来的后果更为传奇：一些所谓亏损公司的负责人及其亲属在几年内成了富翁，当上了幸福的美国“投资移民”；与此同时，这些中资公司在海外账户里非法持有的外汇，远远地超于国家外汇管理部门所掌握的数额。需要点明的是，此过程的主角，正是那些政府官员。

4.地下钱庄———
2002年汕头许鹏展地下钱庄案至今令人记忆犹新。许鹏展的主要“资产”，是许鹏展以汕头市金园区新兴鸿展农副产品商行等20多个空壳公司名义在几个银行开设的20多个账户。在一个设施再普通不过的写字间，许的日常业务为 造假财务报表、虚报营业额和利润，却在没有任何营业活动和收入的情况下，不断缴纳各种税收和保险。由于钱庄资金往来量巨大，许鹏展需要频繁地更换账户。每换几个新账户，他就逐步从老账户里把资金转到新账户上来。

目前，广东、福建、浙江一带的地下钱庄一般与香港的找换店连为一体，采用在境内用人民币交割，境外用外汇交割的形式，不发生资金外逃的物理过程。地下钱庄之间也经常互相拆借，组成一个相互勾联的体系。

用地下钱庄洗钱的“好处”有两个，一是成本非常低，以港币为例，假如银行的兑换牌价是1.08,地下钱庄的牌价基本上只有1.10；第二，黑钱到了境外可以经常以亲友馈赠的方式流回来。

中国内地每年通过地下钱庄“洗”出去的黑钱至少高达2000亿元人民币，相当于国内生产总值的2％。

5．各类赌场———

厦门市原副市长蓝甫在远华案中落入法网。问起蓝甫的不明财产和他在境外违纪参赌的情况，他大言不惭地称，几年之内，他通过赌博赚到的钱多达65万美元外加33万元港币。蓝甫的谎言被一些已经到案、曾经陪同他参赌的人戳穿，因为蓝参赌每次都几乎铩羽而归。

如果是高手，他拿着1000万的筹码进场，输掉100万后离场，要求赌场把剩下的900万打进他的账户，他已经为将来可能的追查设置了障碍。因此，人们只注意到这些人在不停地输钱，但相对于标准洗钱模式中的大量损耗，赌场洗钱的风险常常会被认为是可以接受的。

实际上，赌场是最传统的洗钱场所。早期黑手党的毒品买卖大都是现金交易，钱上通常沾有白粉，一旦被警察抓住难以脱罪。黑手党成员便携带现金去赌场换成筹码，一旦输掉差不多30％的钱，就把剩下的筹码换回现金，顺理成章地把赃钱变成“干净”的收入。在世界经济不景气的今天，来自亚洲的赌客是惟一呈增长趋势的客户群。与现实生活中的赌场相比，网上赌场已经成为洗钱的安全天堂。赌博网站总部大多设在有“逃税天堂”之称的加勒比地区。许多网站根本没有受到政府部门的监管，也不遵守国际赌场的游戏规则，它们甚至不会查问客户的身份资料。许多犯罪集团把钱款打进在这些赌博网站开设的账户后，一般先象征性地赌上一两次，然后就马上通知网站说“我不想再玩了”，要求网站把自己户头里的钱以网站的名义开出一张支票退回来。于是，一笔笔数额巨大的“黑钱”便轻而易举地“洗白”了！初步估算，每年通过数百个赌博网站清洗的“黑钱”数额大约在6000亿至15000亿美元之间。

6，证券洗钱———
2002年12月16日，裕丰国际和富昌国际及金禾国际在香港被勒令停牌，停牌原因涉及洗钱。这三家公司，多为朋友之间相互帮衬持股，圈子外的小股东非常少，这种公司在香港市场人称“干”，而被称做“干”的一类公司往往被某些国际组织视为洗黑钱的“最佳拍档”。

马休雇佣了一个来自德崇证券(Drexel Burnham Lambert)的团队，并创建了AIG金融产品公司(AIGFP)。德崇证券是一家与垃圾债券之王迈克尔•米尔肯(Michael Milken)相关的、破了产的投资银行。“这是一个自成体系的世界，”当时参与的一位人士说。这家企业运用了美国国际集团3A级信用评级来进行衍生品交易，起初参与相对较简单的产品，如货币和利率风险对冲保值。

数年后，轮到了约瑟夫•卡萨诺(Joseph Cassano)掌权，他是一位布鲁克林出生、精明强干的交易员，是德崇证券资深高管。卡萨诺在伦敦管理业务，他住在离哈罗德一箭之遥的豪华联排别墅里。无法联系到卡萨诺就本文置评。一位邻居形容卡萨诺是个“和蔼可亲的人”，说他仍住在那里。

根据去年5月份美国国际集团投资人说明会，其金融业务部门的“分水岭式”事件发生在1998年，当时来自JP摩根(JPMorgan)的银行家们来到AIGFP，并请它参于其所谓的“小酒馆交易”( bistro trade)。这些人是经营担保债务凭证(CDO)的先行者，这种产品即一批抵押贷款支持证券的集合，在此后几年对全球金融体系造成了大破坏。

在某些方面，发行信用违约互换（CDS，对债券违约的保险）对于一家以对困难风险定价而自豪的公司而言，这是一个自然的业务延伸。美国国际集团逐渐专事为那些称为“最高等级”层次（见下）债券的持有人提供保险，这种产品被认为是最安全的担保债务凭证。在美国国际集团巨大的资产负债表和原封不动的信用评级支持下，这家伦敦的业务单位迅速发展起来。美国国际集团内部人士估计，1987年和2005年之间，该业务单位净赚逾50亿美元利润，给美国国际集团和AIGFP团队带来了丰厚的红利，他们分享一小部分利润，并且用他们自己的一些钱冒风险。

另一个重要的业务部分是为银行发行信贷保护。通过购买针对公司债和住房抵押贷款违约的保险，银行能降低他们必须在资产负债表上维持的呆滞资本的数额。马休出席格林伯格主持的AIGFP季度董事会会议和每周一的会议，也收到从美国国际集团信贷风险部门发来的定期最新动态资料。马休说，到他和格林伯格都离开公司的2005年，该单位的业务没有明显出问题的迹象。

格林伯格被英国人马丁•沙利文(Martin Sullivan)取代，而他现在仍身陷前纽约州总检查长艾略特•斯皮策(Eliot Spitzer)提起的民事欺诈诉讼之中。但在格林伯格不在的时候，沙利文或任何人似乎都不能掌控美国国际集团的所有业务。

在美国国际集团几名现任和前高管看来，沙利文是一位出色的保险人，但是该集团却已远远不止是一家保险公司，并且他对金融方面缺少深刻的理解。“马丁是仓促提拔的，”一位前高管说。他补充说，只有格林伯格真正把保险和金融分支都把握在手中。“他被扔出码头并要他在深水里游泳。”无法联系到沙利文就此置评。

仅仅在沙利文被任命后的几个月，2005年年底，卡萨诺担心美国住房市场的状况，决定停止对有次贷敞口的担保债务凭证提供信用保护。他是正确的。美国经济在减速，住房泡沫就要破灭。然而，基于2007年年中的数据，美国国际集团有4650亿美元最高级级信用违约互换。这项业务的年保费为2亿至2.5亿美元。

当去年美国房地产市场日益变糟时，美国国际集团仍深信信用违约互换(CDS)资产组合将能挺过风暴。美国国际集团首席风险官鲍勃•刘易斯(Bob Lewis)表示，将为美国国际集团计提“萧条比例”项目来承受风暴。卡萨诺更直言不讳，他在2007年8月告诉投资人说，“毫不轻率地说”，公司想不出“我们在所有那些交易中亏损1美元”的情形。

当去年秋天银行接踵冲销时，他们持有那些复杂工具的价值暴跌。美国国际集团继续消除人们的恐惧。在12月5日的说明会上—主题匆忙从人寿保险变成了美国国际集团的抵押贷款敞口—沙利文坚持说情况是“可控制的”。

但在之后数周内，美国国际集团及普华永道(PwC)的审计师开始怀疑这家保险商为担保债务凭证估值的方法，这些方法继而决定美国国际集团提供的信用保护的价值。美国国际集团假设，保险将比担保债务凭证价值更高。随着担保债务凭证的价格突然猛烈转向，不能再把这当成既定条件。这种估值基础的变动使得在信用违约掉期资产组合上的损失从10月份的10亿美元增至11月份时的近50亿美元。

到那时，信贷市场中的非流动性已开始影响美国国际集团已发行的对于那些次贷支持证券以外证券的保护。2月份，普华永道公布了一个令人震惊的消息，宣布这家保险商计算其敞口价值的方法存在“重大缺陷”。之后不久，美国国际集团和卡萨诺分手了。

冲销继续升级。到年中，AIGFP在信用违约保护上累计了250亿美元损失。这让格林伯格和一群以亿万富翁艾利•布洛德(Eli Broad)为首的投资者壮起了胆攻击美国国际集团的管理层，并要求改革董事会。他们声称风险管理控制被削弱了，以至于AIGFP和一个出借证券和投资于次贷支持资产的部门肆无忌惮。

美国国际集团内部人士否认了这种指控，指出格林伯格创建了这个部门并任命了卡萨诺。他们补充说，格林伯格离职后，几乎没有发行过相关产品。在他们看来，AIGFP因未能预见美国住房市场的快速崩溃和支撑担保债务凭证的抵押贷款价值随之垮台而毁灭。

在6月份，反叛的股东们如愿以偿。沙利文被已成为美国国际集团的董事长前花旗集团(Citigroup)银行家鲍勃•维伦斯坦德(Bob Willumstad)接替。维伦斯坦德打算出售资产和筹集资金，并告诉投资人他将在9月初公布自己的方案。

维伦斯坦德没法实施他的方案了。美国国际集团遭到了来自监管机构的压力。监管机构担心数千家银行以及与公司交织在一起的养老基金倒闭的影响。根据8月份提交给监管机构的一份文件，美国国际集团的债券被调降是最后一击，这意味着对手可以要求该公司追加145亿美元融资抵押资产。此举导致了周末的疯狂谈判，以美联储支持下的拯救行动和出售美国国际集团大多数资产的决定而告终。

格林伯格是唯一对美国国际集团热情不减的人。在抱怨因美国国际集团股价暴跌而损失数十亿美元后，格林伯格并不隐瞒极想购买这个他曾经掌管的帝国尽量可能多的股份。格林伯格告诉密友，他把这看作拯救AIG和自己名望的机会。

格林伯格希望“加强和恢复他的名誉”，谢尔普说道，“在我看来，那是他最想要的，胜过其他一切。”

译者/红岭

Mr Matthews hired a team from Drexel Burnham Lambert, the failed investment bank associated with Michael Milken, the junk bond king, and created AIG Financial Products (AIGFP). “It was a world unto itself,” says one person involved at the time. The venture deployed AIG's triple-A credit ratingto make derivative deals that at first involved relatively simple products such as hedging currency and interest rate risks.

After some years, leadership passed to Joseph Cassano, a tough, Brooklyn-born trader and Drexel veteran. Mr Cassano ran the business from London, where he lived in a grand town house a stone's throw from Harrods. Mr Cassano could not be reached. A neighbour, describing him as a “delightful person”, says he still lives there.

According to an AIG investor presentation in May last year, the “watershed” event for its financial unit came in 1998 when bankers from JPMorgan came to AIGFP and asked it to participate in what they called “bistro trades”. These were precursors to CDOs, the pools of mortgage-backed securities that were to wreak havoc on the global financial system years later .

In some respects, writing credit default swaps – insurance against bond default – was a natural extension for a company that prided itself in pricing difficult risks. AIG came to specialise in insuring holders of what it believed were the safest CDOs, known as “super-senior” tranches (see below). Backed by AIG's huge balance sheet and pristine credit rating, the London unit thrived. AIG insiders estimate that between 1987 and 2005, the unit netted more than $5bn in profits – a rich dividend for AIG and the AIGFP team, who shared a slice of the profits and had some of their own money at risk.

An important part of the business was writing credit protection for banks. By buying insurance against defaults on corporate debt and residential mortgages, banks could reduce the amount of idle capital they had to hold on their balance sheet. Mr Matthews – who was present at AIGFP's quarterly board meetings and the weekly Monday meetings chaired by Mr Greenberg, also receiving regular updates from AIG's credit risk department – said that by March 2005, when both he and Mr Greenberg left, no signs of trouble in the unit's business were evident.

Mr Greenberg, who remains embroiled in a civil fraud lawsuit brought by Eliot Spitzer, the former New York attorney-general, was replaced by Martin Sullivan, a Briton. But in Mr Greenberg's absence neither Mr Sullivan nor anyone else appeared able to get a grip on all AIG's businesses.

In the view of several present and former AIG executives, Mr Sullivan was a brilliant insurance man but the group had become much more than an insurance company and he lacked a deep understanding of the financial side. “Martin's was a battlefield promotion,” says a former executive, who adds that only Mr Greenberg was really in command of both the insurance and financial arms. “He was thrown off the dock and asked to swim in deep water.” Mr Sullivan could not be reached for comment.

Just a few months after Mr Sullivan's appointment, at the end of 2005, Mr Cassano decided to stop offering credit protection on CDOs with subprime mortgage exposure, amid fears over the state of the US housing market. He was right. The US economy was slowing and a housing bubble was about to burst. Yet, based on mid-2007 figures, AIG had $465bn in super-senior credit default swaps. Its annual premium from this business was $200m-$250m.

As the US property market turned increasingly sour last year, AIG remained confident its CDS portfolio would weather the storm. Bob Lewis, AIG's chief risk officer, indicated that it would take events of “Depression proportions” for AIG to suffer. Mr Cassano was even blunter, telling investors in August 2007 that “without being flippant” the company could not envisage a scenario that would “see us losing $1 in any of those transactions”.

When last autumn brought a rash of writedowns by banks as the value of their holdings of these complex instruments plunged, AIG continued to reassure. In a December 5 presentation – the subject hastily changed from life assurance to AIG's mortgage exposure – Mr Sullivan insisted the situation was “manageable”.

But in the following weeks AIG and its auditors at PwC began to have doubts about the way the insurer was valuing CDOs, which in turn determined the value of the credit protection AIG had provided. AIG had assumed that the insurance would be valued more highly than the CDOs. With wild swings in CDO prices, it could no longer take this as given. A decision to change its valuation basis increased losses on the credit default swap portfolio from $1bn to almost $5bn for October and November.

By that time, the illiquidity in credit markets had begun to infect the protection AIG had written on securities other than those backed by subprime. In February PwC dropped a bombshell, declaring there was a “material weakness” in the way the insurer valued its exposure. Shortly afterwards, AIG and Mr Cassano parted company.

The writedowns continued to escalate. By midyear, AIGFP had racked up about $25bn of cumulative losses on its credit default protection. This emboldened Mr Greenberg and a group of investors led by the billionaire Eli Broad to attack AIG's management and demand boardroom changes. They claimed that risk management controls had weakened, allowing AIGFP and a unit that lent securities and invested in mortgage-backed assets to run wild.

AIG insiders deny the allegations, pointing out that Mr Greenberg created the unit and appointed Mr Cassano. Very little business was written after Mr Greenberg's ousting, they add. In their view, AIGFP was undone by the failure to predict the rapid collapse of the US housing market and the subsequent meltdown in the value of mortgages underlying the CDOs.

Nevertheless, in June the rebel shareholders got their wish. Mr Sullivan was replaced by Bob Willumstad, a former Citigroup banker who had been AIG's chairman. He aimed to sell assets and raise capital. Mr Willumstad told investors he would unveil his plan by early September.

He would not get to do so. AIG came under pressure from regulators worried about the effects of a failure on the thousands of banks and pension funds that had become intertwined with the company. The final blow came when AIG's debt was downgraded, meaning counterparties could demand $14.5bn in collateral, according to a regulatory filing in August. The move led to the frenzied weekend of talks that culminated in the Fed-backed rescue and the decision to sell off most of AIG's assets.

The one person whose enthusiasm for the company appears undimmed is Mr Greenberg. After complaining of having lost billions of dollars through the collapse in AIG shares, he has made no secret of his desire to buy as large a chunk as possible of his former empire. He has told close associates he sees this as his chance to save the company and his own standing.

Mr Greenberg hopes to “enhance and restore his reputation”, says Mr Shelp. “That is what he wants, in my view, more than anything else.”

英国《金融时报》弗朗西斯科•格雷拉(Francesco Guerrera)纽约和安德烈•费尔斯特德(Andrea Felsted)伦敦报道 2008-10-14

1972年，理查德•尼克松(Richard Nixon)历史性地访问了毛泽东领导下的中国。随后不久，汉克•格林伯格(Hank Greenberg)把美国国际集团(AIG)一位高级雇员罗恩•谢尔普(Ron Shelp)召到了办公室。

这家保险巨头的首席执行官，只是简单地说：“尼克松打开了中国的大门。”谢尔普回忆道。格林伯格说：“我要回去，罗恩，我想第一个回去。”最早，这家保险巨头起源于上海的一家保险经纪行，后来被共产党政权赶出中国。

格林伯格这位经历过诺曼底登陆和朝鲜战争的老兵，在身居美国国际集团最高位置的三十年多里，展现了争当第一的不屈愿望。美国国际集团在其最高峰时，资产规模近1万亿美元，实现年净利润140亿美元，在130个国家里有10.6万名雇员。

如今，在格林伯格迫于监管机构压力下离职的三年后，美国国际集团大厦摇摇欲坠。现年83岁的格林伯格仍是美国国际集团的最大股东。一场划时代的金融危机，加上过多地大量下注深奥难懂的证券，内部控制欠缺和监管不善，这些因素结合起来，迫使美国国际集团上个月接受了来自美联储的850亿美元贷款，从而落入政府之手。

最近，在经过偿还贷款和拯救自己免于灭顶的最后努力后，美国国际集团的临时看守管理团队宣布了庞大的处置方案。如果该投资剥离方案成功，那么美国国际集团的规模将较此前大大缩小，如果失败，那么公司将从此湮没。

这一切是怎么发生的？过去几个月里发生的事情显然是这种灾难性命运的催化剂。规模相对较小的美国国际集团业务部门因大量没有监管的衍生品而产生了巨额账面亏损。但美国国际集团从内部跨塌的根源可以追溯到其独一无二的历史和不同寻常的特殊结构上。

美国国际集团的创始人科利厄斯•范德•斯塔尔(Cornelius Vander Starr)是前加利福尼亚一位冰激凌销售商，他移民到中国，追求更美好的未来。他赚钱的观念是，向竞争对手无法染指的市场销售保单——在他这个案例中就是中国人和中国企业——这个观念成为了美国国际集团公司文化的核心部分。

格林伯格1968年接替斯塔尔的位置，加强了斯塔尔的开创性精神，在美国内外创建了一个政界、商界和监管关系网络，并让一名分析师把公司标榜为“主权公司国家”，通过这样使美国国际集团扩张。他精力充沛地构建网络并积极追求利润，这给股东们创造了惊人的成果。直到最近的危机前，在格林伯格誓言每年让收益增长15%的支持下，公司股票一直是纽约证券交易所里表现最好的股票之一。

爱德华•马休(Edward Matthews)表示，他能够看到人们看不到之事，“他可以先于其他任何人看到保险行业向什么方向发展。”爱德华•马休是格林伯格在美国国际集团以及后来在史带投资集团(C.V. Starr)的最资深副手。

但是，美国国际集团发展成一群公司的松散集合，仅靠格林伯格这位中心人物联合起来，这是有缺陷的。格林伯格拒绝对本文置评，但前任和现任美国国际集团的高管们说，公司的结构是笨拙和不透明的，其业务为任何东西提供保险，从在韩国的断骨，到华尔街的有担保债务。一名前高管人士在回忆美国国际集团有4300个法人实体散布在全世界时为之叹息，这种结构使得内部协调和外部监管在事实上都成为不可能。

另一位前员工指出，由于美国国际集团财务部门需要应付的数字规模过于巨大，常常使得季度收益被拖延到其报告期结束几周后才公布。

确实，多数观察人士同意，格林伯格和一小撮副手们对公司有清晰的了解。“整个公司集中在汉克身上，就是汉克一个人，”另一位前高管说。格林伯格给不同的部门足够空间管理他们自己的事务，只有一条：他们要达标，并把支票寄回到总部。“有一种分权精神：人人能上场并做他们自己的事情，只要你能赚钱，”谢尔普表示。他于1985年离开了美国国际集团。

但格林伯格远远不是彻底放手不管。前同僚们回忆，他是一个几乎偏执地专注于细节的人。在上任初期，当他注意到高级保险从业人喜欢在美国国际集团餐厅吃午饭时来上几杯马提尼，于是他签发了一份备忘录，限每餐只能喝一杯鸡尾酒。“格林伯格是什么都管，”马休说。“他管理组织自上而下的三个分层。”

因此，当格林伯格决定开设一个全新的业务，没有人提出质疑。他的多元化举措是基于一条确信，在许多个丰收年后，保险周期就开始拐头。“我们能够不再逆潮流而行：我们就是潮流，”马休回忆格林伯格时说。“我们需要向金融服务业拓展。” (待续)

Soon after the historic visit by Richard Nixon to Mao Zedong's China in 1972, Ron Shelp, a senior employee at American International Group, was summoned to Hank Greenberg's office.

The chief executive of the insurance giant, which had started as a Shanghai-based underwriting agency before being kicked out by the communist regime, had a simple message to convey. “Nixon has opened up China,” Mr Shelp recalls him as saying. “I wanna go back, Ron, I wanna go back first.”

Throughout more than three decades at the top of AIG, Mr Greenberg, a D-Day and Korean war veteran, showed an uncompromising desire to be first. At its apogee, AIG had nearly $1,000bn in assets, made $14bn in annual net profits and employed 106,000 people in 130 countries.

Today, three years after Mr Greenberg – now 83 and still AIG's largest shareholder – left under pressure from regulators, his edifice totters. The combination of an epochal financial crisis, outsized bets on exotic securities, inadequate internal controls and poor regulatory supervision forced AIG last month to accept an $85bn loan from the Federal Reserve that places it in government hands.

Last week, in a desperate effort to repay the loan and save itself from extinction, AIG's caretaker management team announced huge disposals. If the divestment programmes is successful, AIG will become a shrunken shadow of its former self. If it fails, the company faces oblivion.

How could this happen? The events of the past few months, in which a relatively small AIG unit racked up huge paper losses on a swath of unregulated derivatives, were clearly the catalyst for so ruinous a fate. But the roots of AIG's implosion can be traced back to its unique history and idiosyncratic structure.

AIG's founder, a former ice-cream seller from California called Cornelius Vander Starr, had emigrated to China in search of a better future. His notion of making money by selling insurance policies to markets that rivals would not touch – in his case Chinese people and businesses – was to become a core part of AIG's corporate culture.

Mr Greenberg, who succeeded Starr in 1968, reinforced this pioneering ethos, expanding AIG by creating a web of political, business and regulatory relationships in the US and abroad that led one analyst to brand the company a “sovereign corporate nation”. His high-powered networking and aggressive pursuit of profits worked wonders for shareholders. Until the latest crisis, its shares had been one of the best-performing on the New York Stock Exchange, fuelled by Mr Greenberg's pledge to increase earnings by 15 per cent annually.

“He could see around corners,” says Edward Matthews, one of Mr Greenberg's most senior lieutenants at AIG and then at C.V. Starr, his investment vehicle. “He could see where the insurance industry was going before anybody else.”

But AIG's development as a loose collection of companies unified only by the central figure of Mr Greenberg had drawbacks. Mr Greenberg declined to comment for this article but former and current AIG executives say that its structure, with businesses insuring anything from broken bones in South Korea to collateralised debt obligations on Wall Street, was cumbersome and opaque. One former senior executive sighs when recalling that AIG has more than 4,300 legal entities scattered around the world, a set-up that made both internal co-ordination and outside scrutiny virtually impossible.

Another former employee notes that the sheer amount of numbers to be crunched by AIG's finance department often delayed the release of quarterly earnings until several weeks after the end of its reporting period.

Indeed, most observers agree that only Mr Greenberg and a handful of aides had a clear view of the group. “The whole company came together at Hank, and Hank only,” says another former executive. Mr Greenberg gave different units ample leeway to run their own affairs, with one caveat: they had to “make their numbers” and send the cheque back to headquarters. “There was a decentralised spirit: everybody could go on and do their own thing, as long as you made money,” says Mr Shelp, who left AIG in 1985.

But Mr Greenberg was far from entirely hands-off. Former colleagues recall an almost obsessive attention to detail. When, early in his tenure, he noticed that senior underwriters liked to lubricate their lunches in AIG's canteen with several martinis, he issued a memo imposing a one cocktail per meal limit. “Mr Greenberg was hands-on,” says Mr Matthews. “He managed three layers down the organisation.”

Hence, when Mr Greenberg decided to launch into a whole new business in 1986 no one raised an eyebrow. His move to diversify was based on a conviction that, after a few bumper years, the insurance cycle was turning. “We can no longer go against the tide: we are the tide,” Mr Matthews recalls him as saying. “We need to develop into financial services.”(To be continued)