自动执行的交易信号

I'm Tashara Fernando. 我是 Tushara Fernando。 I'm head of data and AI at Mang Group. Man Group 数据与 AI 主管。 Mangroup are an alternative investment manager. Man Group 是一家另类投资管理公司。 We manage over $200 billion of assets. 我们管理着超过 2000 亿美元的资产。 Our clients are pension funds, sovereign wealth funds, and large institutions. 我们的客户是养老基金、主权财富基金和大型机构投资者。 We manage real people's money, thousands of people's pensions and investment capital. 我们管理的是真实的钱，成千上万人的养老金和投资资本。 So when we think about AI, the stakes are high for us. 所以谈到 AI，对我们来说风险很高。 Our clients are real people from their teachers in Canada, their metal workers in Japan. 我们的客户都是真实的人，有加拿大的教师，有日本的钢铁工人。 So, we really need to get AI right. 所以我们必须把 AI 做对。 If we get this wrong, we could lose real money. 一旦出错，我们会损失真实的钱。 One of the really large parts of our business is systematic trading. 我们业务中极重要的一块是系统化交易。 And that represents a huge opportunity to be aided by AI. 这也是 AI 能大力赋能的巨大机遇。 By systematic trading, I mean algorithmic trading capabilities that look across thousands of securities and hundreds of markets to make investment decisions. 所谓系统化交易，我指的是算法交易能力，它横跨数千只证券和数百个市场，自动化地做出投资决策。 So systematic trading is really about trading signals. 系统化交易的核心，其实就是交易信号。 And you can think of a signal like a fantasy football team. 你可以把信号类比成梦幻足球队。 You can think that we want to pick the best players for our squad based on some intuitive strategy. 我们想基于某种直觉策略，为阵容挑选最优秀的球员。 So the green bars here would be a starting lineup. 图里绿色柱代表首发阵容。 The red bars would be the reserve squad, the people that you don't want in the team. 红色柱是替补席，是你不想上场的球员。 And the people in the middle, they might come in, you might kind of transfer them in, but they're they're not the star players at the moment. 中间的球员随时可能入队，你可以换他们进来，但他们目前还不是主力。 Maybe a subs bench. 也许是替补席。 So, a signal is really about thinking about a striker maybe hitting form, but you need to transfer them in before the Friday deadline when the price might go up. 信号的本质，就是判断某位前锋可能迎来状态爆发，然后在周五截止前价格上涨之前把他换进来。 And then you have savvy managers who are really thinking about form fixtures and what might happen to get the most points in the season. 聪明的经理会认真研究状态、赛程，判断怎样在赛季里拿最多分。 And they want to transfer the right players in at the right time and at the right cost. 他们想在正确的时间、以合适的价格换入对的球员。 So this is quite similar to systematic trading. 这和系统化交易非常相似。 A trading signal is really just this with stocks. 交易信号本质上就是把这套逻辑用到股票上。 So the bars here would represent company stocks. 图里的柱子代表各公司的股票。 We want to back the ones that would make money and we want to short the ones that won't. 我们做多看好的标的，做空不看好的标的。 So in this example, we've ranked the stocks by the past 3 month returns and we run that through history to see if it would have made money. 在这个例子里，我们按过去 3 个月回报率对股票排序，再把这个策略跑一遍历史数据，看它是否能赚钱。 The interesting question is always what is that factor that you want to rank things by? 关键问题始终是：你想用什么因子来排名？ What's the strategy to get the right stocks in your team? 什么策略能帮你把对的股票选进阵容？ And does it work? 那它有效吗？ And how do you know whether it works or not? 怎么判断它到底有没有用？ Well, the truth is you never really know. 说实话，你永远不可能真正知道。 I'd love to be able to tell the future, but I can't. 我很希望能预知未来，但做不到。 So, the best thing that we can do is look at what happened historically. 所以我们能做的最好的事，就是看历史上发生了什么。 We run that strategy, we codify it, and we run it against 15 years of history or even longer. 我们把策略跑通，将其代码化，然后跑 15 年甚至更长的历史数据。 And what that does is it runs that strategy through lots of macroeconomic environments, through lots of stresses. 这样做，就能让策略经历各种宏观经济环境和各种压力测试。 And we can see how it performed. 然后看它的表现如何。 And that back test produces lots of statistical factors. 回测会产出大量统计指标。 Some examples might be how much money did it make? 比如，它赚了多少钱？ So what's the annualized return? 年化收益率是多少？ When it lost money, and they always do at some point, how much did it lose? 当它亏钱的时候，而它总有亏钱的时候，亏了多少？ And we call that a draw down. 我们把这个叫做最大回撤。 And we look at some even more complex statistical factors. 我们还会看一些更复杂的统计指标。 One's called a sharp ratio which compares the volatility of that strategy versus how much it returned. 有一个叫夏普比率，它把策略的波动率和收益率放在一起比较。 And it's this process, this systematic trading workflow that we think that we can use AI to really enhance to come up with those ideas to run the back tests and that has been our focus. 正是这个流程，这套系统化交易工作流，让我们相信 AI 能真正提升它，帮我们产生想法、跑回测，而这也一直是我们的重心。 So there are trading signals running right now in production at Mang Group, a regulated investment firm running real capital that were researched, back tested and proposed by AI. 现在 Man Group 有交易信号正在生产环境中运行，这是一家受监管的投资公司，运作着真实资本，这些信号是由 AI 研究、回测并提出的。 By that I mean humans came up with the sorry AI came up with the idea. 我的意思是，AI 想出了这个主意。 AI got the data. AI 获取了数据。 AI ran the back test. AI 跑了回测。 AI then wrote up the strategy proposal and AI productionized the signal. AI 撰写了策略提案，并将信号产品化。 Humans of course reviewed all of the output to make sure that it was sensible. 当然，人类审核了所有输出，确保它是合理的。 But a AI was at the center of that process. 但 AI 是这整个过程的核心。 And I'm sure you want to know what was that signal? 我肯定你们想知道那个信号是什么。 What was that investment idea? 那个投资想法是什么？ How much money did it make? 它赚了多少钱？ How can I use it? 怎么用它？ Sorry, I'm not going to tell you that today. 抱歉，今天我不会告诉你们。 That's our IP. 那是我们的知识产权。 What I'm here to tell you about today is our journey. 我今天要讲的是我们的探索历程。 What was the foundation that allowed us to do that? 是什么基础让我们做到了这一点？ And how can you apply those learnings in your company? 你如何把这些经验用到自己公司？ And it really starts with AI understanding our workflows. 一切都始于让 AI 理解我们的工作流。 And to do that, we use skills. 为此，我们使用技能。 Can I have a show of hands in the audience as to who's written a skill? 请问现场写过技能的人能举个手吗？ Okay, that's great. 好，太棒了。 Most of you. 大多数人都写过。 So, coming up with the signal is the quick bit. 提出信号，其实是最快的部分。 The hard part is everything that you need, everything that's underneath it, all of the workflows that make it happen, that allow you to act on it. 难的是它背后所有的东西，所有支撑它运转的工作流，让你真正能采取行动的那些流程。 Think of it like an iceberg. 把它想象成一座冰山。 The signal is the tip. 信号只是冰山一角。 Underneath it are all of the workflows that make it possible. 冰山水面以下，是让它成为可能的所有工作流。 How do you clean the data? 数据怎么清洗？ How do you stitch prices? 价格怎么拼接？ How do you detect outliers? 异常值怎么识别？ How does it run? 它怎么运行？ What's the infrastructure it runs on? 它跑在什么基础设施上？ How do you run those back tests? 回测怎么跑？ And this is where it can quickly go wrong. 这里是容易出错的地方。 If different teams are running different versions of those workflows, you get different answers. 如果不同团队在跑不同版本的工作流，得到的结果就会不一样。 One team's back test looks amazing. 某个团队的回测看起来非常出色。 and other teams looks average. 另一个团队的却只是平平。 And because they're using different workflows, you don't really know whether it was the idea that was better in one team than the other or whether they're just measuring things differently. 由于用的是不同工作流，你根本无法判断是一个团队的想法更好，还是他们只是在用不同的方式衡量。 Shared workflows fix that. 共享工作流能解决这个问题。 One common foundation means that effort isn't duplicated and you have consistency. 统一的基础意味着不重复投入，并且保持一致性。 The outputs are comparable. 输出结果才具有可比性。 And that's extremely important in systematic trading when we're comparing signals. 在系统化交易中比较信号时，这极为重要。 Out of the box, Claude is an amazing general purpose tool. 开箱即用，Claude 是一款出色的通用工具。 It does a lot, but it doesn't know us. 它能做很多，但它不了解我们。 It doesn't know our data. 它不了解我们的数据。 It doesn't know our systems. 它不了解我们的系统。 It doesn't know how we work. 它不了解我们的工作方式。 And it's the same for everybody in this room. 在座各位都面临同样的问题。 So the first thing that we had to do was teach it. 所以我们首先要做的，是教会它。 Not by retraining it, not by doing fine-tuning, but by giving it access to our data, our capabilities, and our workflows. 不是重新训练，不是微调，而是让它访问我们的数据、我们的能力和我们的工作流。 That's our superpower. 这是我们的超能力。 We have decades of institutional knowledge in systematic research and some of the best technical capabilities on the street. 我们在系统化研究领域积累了数十年的机构知识，以及行业内顶尖的技术能力。 And if we can connect that with AI, then AI can leverage that superpower. 如果能把这些与 AI 连接起来，AI 就能发挥这一超能力。 Skills are the connective layer that allow AI to leverage that superpower. 技能就是那个连接层，让 AI 得以调用这一超能力。 So getting them right is paramount and that was our focus. 把技能做好至关重要，这也是我们的重心所在。