据英国《金融时报》报道，近日一家名叫Kensho的金融数据服务商得到高盛1500万美元的投资。除了高盛，谷歌和CNBC都是该公司的投资者。Kensho目前正在研发一种针对专业投资者的大规模数据处理分析平台。该平台将能取代现有各大投行分析师们的工作。该工作平台将可以快速、大量的进行各种数据处理分析工作并能实时回答投资者所提出的复杂的金融问题。 如果Kensho的产品最后能研发成功，金融机构的分析师和研究人员将面临灾难。面对更快、更好的机器人黄金分析师，他们毫无胜算。 Kensho背后的男人叫做Daniel Nadler，是哈佛大学经济学博士。Kensho位于马萨诸塞州剑桥市，由Nadler与程序员Peter Kruskall联合创立。 自2012年1月24日起，来自对冲基金的电话就不断响起，纷纷要找Daniel Nadler。当天，Nadler在彭博新闻社上与人合著一篇文章，介绍了如何通过美元的走强来预测标普500指数每周的走低情况。在当时，交易员们正利用美元与标普指数之间的这种关系大赚特赚。 Nadler说： 他们的电话就像是说，你这个叛徒！如果你发现了这种关系，那你就利用这种关系来交易！不要公开它啊！你这样导致大家都无法进行套利交易。 Kensho对于金融分析行业的影响就好像当年谷歌给搜索领域带来的冲击一样。 Kensho想要将金融市场的一些专业知识交到大众手中，而此前只有一些顶尖对冲基金和投行使用这些专业知识来进行套利。Kensho公司的软件取名沃伦（沃伦巴菲特的沃伦）。你可以像在谷歌上搜索一样，在简单的文本框里输入复杂的金融问题。例如： 当三级飓风袭击佛罗里达州时，哪支水泥股的涨幅会最大?(最大的赢家是谁?德州工业[Texas Industries])。 又或者，当朝鲜试射导弹时，哪支国防股会涨得最多?(雷神公司[Raytheon]、美国通用动力公司[General Dynamics]、和洛克希德马丁公司[Lockheed Martin])。 当苹果公司发布新iPad时，哪家苹果公司的供应商股价上涨幅度会最大?(为iPad内置摄像头生产传感器的豪威科技股份有限公司 [OmniVision])。 要回答此类问题，即使对冲基金的分析师能找到所有的数据，这也要花上他们数天的时间。但沃伦软件可以通过扫描药物审批、经济报告、货币政策变更、政治事件以及这些事件对地球上几乎所有金融资产的影响等9万余份资料，立刻为6500万个问题找到答案。 如果广泛加以使用，沃伦软件可以撼动长期以来被彭博和汤姆森路透垄断的260亿美元的金融数据市场，并且让华尔街的大银行和Bridgewater等对冲基金的研究工具变得更为大众化。 投资公司Devonshire Investors总经理David Jegen说： 每天，基金经理会有一个投资理念，然后必须找数量分析专家来建立模型。这是多数金融服务公司内部的一个瓶颈，导致经过测试的理念少之又少。 Devonshire Investors是共同基金巨头Fidelity Investments的私募股权部门，也是Kensho公司的投资者。Jegen称，提高投资理念测试的效率、质量和数量，这对于企业来说是件好事。 Nadler曾经在美联储工作。当时，他惊奇地发现这家全球最具权势的金融监管机构仍然依靠Excel来对经济进行分析。Nadler说： 我当时曾经想象美联储会有一个有着实时信息的作战室。 但相反，他发现美联储使用的是普通的不能再普通的Excel。他将这些告诉了Peter Kruskall。Kruskall是一位少年老成的程序员，在麻省理工学院获得了计算机科学的学士和硕士学位，并且在谷歌公司担任实习生，负责开发Gmail的颜色标签。 Kruskall表示，我开始思考自己在谷歌公司所了解到的东西，使用它们来分析市场。 Kensho使用了从谷歌那里得到的灵感，例如映射化简(MapReduce，将庞大的处理任务分配给云服务器的一种方法)和BigTable(将庞大的数据压缩到可以加以处理的规模的一种方法)。2013年4月份，Kruskall从谷歌公司辞职，来到波士顿加入了Nadler的行列。 在向沃伦软件提问后，投资者会得到非常直观的答案，你还可以通过改变时间范围或被研究公司对象等变量来对答案加以优化。 投资者甚至可以在自动驾驶仪上通过沃伦软件进行搜索。 例如，如果当天美国耐用工业品订单高于预期，标准偏差值为0.25，那除了向你播报数据，Kensho还会自动分析，发现这种情况自2009年来已经发生26次，而且在这26次里，Clorox公司有23次的日回报为正。 Kensho可以帮助更多投资者在面对突发状况时，更快速地进行反应，例如美国中西部的龙卷风或者是纽约近日的抗议活动。Stanley Young表示： 在数据分析和洞悉方面，Kensho的沃伦软件确实设定了新的标准。 Young曾担任过彭博社企业产品与解决方案部门和纽约证券交易所技术部的首席执行官。目前他是Kensho公司顾问团成员之一。 要想打入华尔街，Kensho自然需要面临挑战。 Kensho公司的使命是将量化分析大众化，这将同投资界这一注重秘密和排外的群体产生冲突。FirstMark Capital的Matt Turck表示， 我认为市场将向这个方向发展，但还需要经过很长的过程。 Nadler表示，他正在敲定一个一年期的独家协议。用他的话来说，协议的另一方是华尔街一家著名的银行巨头。他同时也计划将沃伦软件租赁给基金经理和买方公司，并且参照彭博社和路透社的方式收取高昂的软件使用费。
Can Kensho Bring Google Style Search To Stock Picking? This story appears in the May 26, 2014 issue of Forbes The screaming calls from hedge funds started coming for Daniel Nadler on Jan. 24, 2012. That was the day Nadler, then a 28-year-old Harvard Ph.D. in economics, co-wrote a story on Bloomberg pointing out how the strength of the dollar can predict weekly lows in the S&P 500 index. At the time traders were making a bundle on this dollar-S&P correlation. “They were like ‘You betrayed the brotherhood! If you discover it, you trade off it–you don’t make it public and kill the arbitrage,’” says Nadler, a fast-talking Canadian with dark ringlets flopping over his forehead. In the coming months Nadler is likely to receive a few more of those calls. His new Cambridge, Mass. company, Kensho, cofounded with programmer Peter Kruskall, is out to do to financial analysis what Google did to search. That is, they want to put in the hands of the masses a level of expertise about financial markets that only a handful of the top hedge funds and banks use to make billions off of short-term market inefficiencies. Kensho’s software, dubbed Warren (as in Mr. Buffett), offers up a simple Google-like text box into which you can pose very complex questions–all in plain English. Stuff like: Which cement stocks go up the most when a Category 3 hurricane hits Florida? (The biggest winner? Texas Industries.) Ditto for defense stocks when North Korea fires a test missile? (Raytheon, General Dynamics, Lockheed Martin.) Which Apple supplier’s share price goes up the most when the company releases a new iPad? (OmniVision, which makes the sensors in the iPad camera.) Teams of hedge fund analysts can spend days answering these kinds of questions, assuming they can find all the data. Out of the box Warren can find answers to more than 65 million question combinations in an instant by scanning more than 90,000 actions such as drug approvals, economic reports, monetary policy changes, and political events and their impact on nearly every financial asset on the planet. Its software runs on Amazon’s Web servers in Nasdaq’s supersecure FinQloud computing network. “This essentially is giving you a quant army,” says Nadler. If deployed widely Warren could shake up the $26 billion financial data market long dominated by Bloomberg and Thomson Reuters, and democratize the tools now closely held by Wall Street megabanks and the elite quant hedge funds like Bridgewater Associates, D.E. Shaw and Renaissance Technologies. “Every day a portfolio manager has an investment idea and has to go to a quant to build the model. That’s a bottleneck within most financial services firms, and as a result far fewer ideas are tested,” says David Jegen, managing director at Devonshire Investors, the private equity arm of mutual fund giant Fidelity Investments. Devonshire is an investor in Kensho. “Improving the efficiency, quality and volume of ideas that can be tested can be a good thing for incumbents.” Other investors think so, too. Founded last May, Kensho has raised a $10 million seed round from the likes of Accel Partners, Breyer Capital, General Catalyst, Google Ventures and NEA. Nadler grew up in Toronto, the son of immigrants from Poland and Romania. His dad, an engineer who designs ways to use sound to find minute cracks in bridges and submarines, would teach him “real math” after school. As a kid Nadler sketched diagrams for perpetual-motion machines and spent hours reading ancient Greek before attending Harvard to study math and classics. He stayed at Harvard to get a Ph.D. (he graduates this year), where, while writing his thesis on the pricing mechanisms of credit derivatives (Nadler is quick to point out that the first olive oil futures were developed by Thales in 600 B.C.), he studied poetry with Pulitzer Prize winner Jorie Graham, launched a sleep app called Sigmund and served as a visiting scholar at the Federal Reserve. The idea behind Kensho was born at the Fed when Nadler was stunned to discover the world’s most powerful financial regulator was still relying on Excel to analyze the economy. “I was expecting a war room with real-time information,” says Nadler. Instead he found the same stat kits he had in his dorm room. He told all this to Kruskall, a precocious coder who earned both a B.S. and an M.S. in computer science from MIT and developed Gmail’s color tabs as a Google intern. Kruskall knew a better way to do this. “I began taking the things I was exposed to at Google and used them to look at the markets.” At the time Kruskall was at Google developing ways to bring the Internet to far-flung locations. For the next eight months he hacked together Warren in all-nighters and on bus rides to and from the search giant’s Mountain View campus. Kensho uses Google-inspired technologies such as MapReduce, a method for splitting up huge processing tasks among cloud servers, and BigTable, a way of compressing giant data sets to a workable size. In April 2013 he quit Google and joined Nadler in Boston. A query in Warren comes back as an elegant visualization that you can continue to refine by changing variables such as time frame or companies in the study. Warren can search on autopilot, too. During the day of a visit to Kensho’s office U.S. durable goods orders surprised on the upside by a standard deviation of 0.25. Kensho automatically identified that this exact event had happened 26 times since 2009 and that Clorox delivered positive daily returns on 23 of those 26 days. “We could all get hit by a bus,” says Nadler, “and it would continue to find price disparities.” Used in the real world, Kensho could help more investors react faster to unexpected events such as a tornado in the Midwest or a protest in the Middle East. “They are really setting a new bar when it comes to data analysis and insight,” says Stanley Young, former CEO of both Bloomberg’s Enterprise Products; Solutions and NYSE Technologies. He’s on Kensho’s advisory board. “Their mojo is all about leveling the playing field,” says General Catalyst’s Larry Bohn. “They’re trying to bring super high performance computing to the public… It’s very mission based.” The mission won’t be easy. Kensho will have its challenges breaking into Wall Street, which has seen plenty of kids with black boxes come and go. For every ten hotshot college grads at Google and Facebook earning a few hundred thousand dollars there’s a hedge fund Ph.D. making a few million. Plus Kensho’s mission to democratize quantitative analysis is going to collide with a community that prizes secrecy and exclusivity. “I think the market is going toward this type of thing, but I think it’s going to be a long process,” says Matt Turck of FirstMark Capital. “Institutions have spent a lot of money on in-house solutions and teams whose sole purpose is to build those solutions.” Nadler says he’s finalizing an exclusive, one-year agreement with, as he describes it, “one of the premier bulge-bracket banks on Wall Street.” He also plans to lease Warren to asset managers and buy-side firms on a pricey subscription basis as Bloomberg and Reuters do. “They are landing pilot programs with a number of very sophisticated asset managers that are known for being very conservative,” says billionaire investor Jim Breyer, founder and CEO of Breyer Capital and partner at Accel. “That is something we haven’t seen very often.” Plenty of people have asked Nadler why he doesn’t just use Warren’s computing power to launch his own hedge fund. “That has been done before. I don’t like being the eighth person to do something.” Nadler had better start screening his calls.
Warren来了，分析师去哪儿？ 预测美国灾难性飓风或中东国家爆炸事件的市场影响再也不需要分析师们绞尽脑汁了，一款由交易技术专家和前谷歌工程师合作开发的类苹果Siri应用“虚拟市场助手”Warren现已横空出世。 你在看到埃及抗议的新闻时就想知道中东地区动乱会对能源价格造成何种影响吗？你是否在四级飓风登陆美洲大陆之始就想知道会对家得宝（Home Depot）、住宅开发商和水泥公司的股价带来什么损失？抑或，想先与他人知道油价位于100美元/桶上方对能源公司来说意味着什么？Warren都能在第一时间帮你一一解答。 Warren是位于美国马萨诸塞州坎布里奇市的创业公司Kensho最新发布的一款类似于苹果Siri的虚拟助手，目标用户面向各类投资者，声称可以即时针对全球事件中的各种金融相关问题作出回答。据Kensho公司表示，之所以将该应用命名为Warren，是为了传承投资大师巴菲特（Warren Buffett）的精神。目前，该应用已经接受了许多资产管理者和研究团队的测试。 Kensho公司成立于2013年，现有约20名员工，其中不乏来自麻省理工学院、哈佛大学和斯坦福大学等名校的高材生。该公司在成立之初就致力于开发能够沟通自然语言和智能电脑系统的虚拟私人助手，或将在未来五年内走在全球科技创新的尖端。有分析认为，Kensho的成功不仅会为消费品行业带来翻天覆地的变化，其影响力甚至还会波及医药、金融、能源和国防等领域。Kensho公司的CEO Daniel Nadler年仅30岁，曾是美联储的访问学者。 Warren基于纳斯达克OMX云计算平台，该平台是专门为金融服务部门设计的。使用云计算平台的好处是，让那些仍然依赖Excel工作的小公司能够与每年花费数百万美元用于完善计算机基础设施的大型对冲基金一较高下。Nadler在向客户介绍Warren时称，该应用可以深度分析全球事件和数据，并将分析结果用直观的图形界面呈现给用户，而所用时间远远少于机构研究团队，也不用写代码。在Warren的定价方面，未来将按照不同投资者的类型收取相应的费用。 Kensho公司在1月22日公开宣布，已经获得了来自包括Google Ventures、Accel Partners、Devonshire Investments和Fidelity Investments四家机构共计1000万美元的风险投资。这些资金足以让Kensho公司雇佣来自世界各地最优秀、最有经验的工程师来进行Warren的研发和更新。Devonshire Investors负责人David Jegen认为，积极资产管理需要持续创新，而我们正在见证技术进步是如何改变现有投资方法和理念的。Kensho公司一直走在技术创新的尖端，目标是为投资组合管理者提升业绩，我们很荣幸加入这个团队。从前期研发成果来看，Kensho公司的软件已经为投资研究和投资组合管理过程完善创造了不小的价值。 Warren的横空出世表明，越来越多的公司和投资者希望能够将技术进步充分应用在消费品市场如iPhone虚拟助手Siri或金融服务领域。通常情况下，只有一小撮精于研究的投资者和机构能够顺利驰骋金融市场。但Kensho公司试图通过Warren为更多的投资者提供专家级别的研究和投资分析。该公司表示，Warren能够将投资研究周期从此前的几天缩减到几分钟。目前，该应用能够回答超过100万个问题，未来有望突破1亿个。另外，Warren或于今年底能够实现语音识别功能。 Nadler对于Warren的市场定位和前景颇为乐观，他在日前表示，只要我们看看苹果的Siri、IBM的Watson和谷歌的Now，就会意识到大多数科技巨头在2020年之前都会在研发虚拟智能助手上“下血本”。到现在为止，我们还没有看到一款专门针对金融领域的虚拟助手。而Warren的出现正好填补了这一空白。 彭博前CEO、Kensho咨询委员会成员Stanley Young认为，Warren的亮点是基于对数据的深入分析来回答人们提出的各种金融相关问题。据悉，该公司的咨询委员会成员还包括前美国中央情报局亚洲负责人、前美国国防部亚洲助理部长James Shinn。Shinn认为，事件驱动型统计数据分析是最有利于理解这个世界的方法之一，而油价和货币价格等受地缘政治事件影响的资产价格波动原因及影响恰恰是让投资者最头痛的。如果没有一定的技术支持，这一领域或将成为最困难、最费脑力和研究代价最高的投资难题。 Warren的出现无疑是史无前例的，也填补了金融服务领域里的一大空白。但先走一步并不意味着一定能笑到最后。首先，鉴于金融信息对于精准性和即时性的要求，Warren数据库的可靠性及更新对Kensho公司来说都是不小的挑战。其次，Warren能否实现即时、准确回答用户的问题也在很大程度上依托于自然语言处理技术和语音识别技术在未来的发展。再次，要实现Warren的全部功能，需要智能电脑系统能够读取并记忆几百万页的自然语言信息，并在几秒钟内作出分析，这本身就是金融技术史上遭遇的最大挑战之一。（文章翻自Boston Business Journal，Sara Castellanos著）
我认为他们那个东东主要影响事件驱动型的基金，对data-crunching型的量化基金影响不大。CFA级别的分析师可能会受冲击，毕竟他们太初级了，Quant不会，尤其是拥有独门数学模型的更不会。所以说想保住饭碗还需往行业的顶端爬呀。