Alpha began with a quick introduction to Stradegi, “We are a consulting and solutions company established with the sole objective of helping investment managers plan and implement transformational change. We use an industry-specific approach to overcome challenges, made possible by our unique mix of specialist consultants, with domain experience and financial technology expertise. We have a growing data science team and generally focus on factors that impact investment products, distribution, research, process, operations and governance.”
She explained that Stradegi was recently awarded the management consultant of the year, for the third consecutive year, by Asia Asset Management. “But the cherry on top,” she said, “has been for us to be recognised as one of the Top 10 Hottest APAC AI Consulting Companies by CIO Advisor.”
Baid explained that AI is an enabler for processing the vast amount of data, not humanly possible to either gather or absorb in today’s landscape. “Even for research analysts, it is impossible to track, filter, process and take action on everything. There’s just too much information”
The answer is 42
Baid said she would first somewhat demystify some of the more common buzzwords that form part of AI jargon.
“AI involves systems that try to mimic human intelligence,” she reported. “Machine Learning is a subset of AI and involves coding that encourages the computer software to learn by itself. And Deep Learning is a subset of Machine Learning inspired by the information processing patterns of our brains. And finally, Natural Language Processing, or NLP, is basically the ability of the computer to understand and contextually process language.”
She explained that a trifecta of factors is necessary for AI to work at scale - flexible mathematical models, a lot of data and computational power. The first of which has been around since the 1940s and only in recent years have the other two made it possible to implement AI at scale.
Technology surges ahead
“As both data and computing power progress at an accelerating rate, wealth managers are benefitting. In fact, a global Swiss private bank is already testing voice-controlled AI enabled portfolio updates and one of the largest banks in Singapore is performing sentiment analysis on annual reports, news and blogs, so relationship managers can relay this compact but rich information to their clients.”
Building the future
She explained that to benefit from these advances, the wealth management industry must understand what they are aiming to achieve. Most of what we are discussing today requires three components: an NLP engine that allows the specifics of the finance industry to be accurately represented, a data lake that combines and organises your structured and unstructured datasets, and a platform that allows you to visually analyse this integrated information. “What can this achieve for us?” she asked, rhetorically. “In brief, we can augment traditional datasets with insights extracted from text at incredible scale. We can connect different datasets and explore and forecast trends for an individual company, compare against a competitor, or scale our research to get a peer groups view, a competitor view, a sectoral view or a country view. And then we can uncover hidden correlations and quantify the level of historical impact to help with investment decisions. All of this in just seconds.”
Visualisation is vital
Baid explained that these insights should be in the form of visualisation and configurable dashboards. “We all know that humans can process images and visuals better than a lot of numbers in Excel spreadsheets,” she reported, “so powerful visualisations help us to absorb information faster. And after all, we build these tools to help you make decisions much quicker, make you more productive, provide richer insights, and thereby serve clients better.”
She elucidated, explaining that if an RM spends all their time on portfolios for every individual client, there is little time remaining for value added services. Here, the AI tool is able to give them specific information for each of those different clients, meaning they can then significantly enhance their client service quality and accuracy. “You can obtain better insight because you are looking at multiple sources of information, you’re looking at a lot of data, and you can do it at scale.”
How it works…
She went into considerable detail about how this works and how RMs, servicing multiple clients, can practically leverage technology like this. “We incorporate flexible settings that allow RMs to control and filter their research insights and alerts by portfolio, its constituent parts, or even by client” she said.
“Imagine if you had to do all that yourself, just for one client,” she said. “Now just imagine if the computer did all of that for you and only gave you what you consider most relevant or vital.”
Such is the power…
“Previously, you had to research, identify, gather and analyse a lot of information to arrive at a hypothesis. Now all you need to do is pull out the company, select the metrics that you want to use and you can very quickly and easily see, for example, why a stock price has moved in a certain period. Even better, imagine the NLP engine reading through the entire annual report and telling you this metric, for example EPS, has moved and the reasons for this are A, B and C. Without you even needing to read the annual report! Such is the power of the NLP engine, and by the way this is only going to get better and better with time, as machine learning begins to understand what you consider important, or vital.”
Baid explained that such is the power of Machine Learning, that once you train the machine to identify an event, it does that not just for one company, it can do that for eight companies, it can do that for a hundred companies and it can do that in seconds. The RM or analyst can use that time saved for more value-added tasks.
Connecting the dots
She moved on to the understanding of important relationships that exist in the RM’s investment universe, for example the directors of a company and their other business involvement and relationships. “With the wonderful visualisations available,” she reported, “you can see the most important and influential people, their relationships to other companies and to other directors, and all this is updated in real-time.”
Baid closed her fascinating Workshop by noting that she had given only a few examples, but that there is vastly more that can be achieved with these tools. “In short, AI and Machine Learning can make a huge difference in the wealth industry. It can make you so much more productive, it can make you more efficient, and ultimately you can service your client better,” she told the delegates.
She explained that the wealth management business of the future will require fewer advisers with better tools. “We will be using the strengths of machines to complement irreplaceable human intelligence. It is an exciting and more profitable future ahead for the industry, if you use all these technologies to your great advantage.”