#MyDataAndI

back to the overview

Can you dig it?

Avatar Julian chats with Spyros Mesomeris – Deutsche Bank’s Global Head of Quantitative & QIS Research – about α-DIG, a new tool that uses text mining and artificial intelligence to uncover insights for portfolio managers.

75

Hi, Spyros. Can you describe α-DIG in a tweet?

75

50

OK, let me try: α –DIG is a tool that crunches big data using text mining techniques to help investors understand the financial market value of company intangibles.

75

What are company intangibles?

75

Examples of intangible assets include patents, brand recognition, copyrights; on the liabilities side, litigation risk and other potential headwinds that are not captured in the numbers. At least not yet. A product recall, supply chain issue or a corporate scandal – all of these things can have material share price implications downstream.

75

Oh, I thought investors already paid attention to things like that.

75

This is true particularly as of late, and specifically with regard to Environmental, Social and Corporate Governance (ESG) policies and issues. The problem is that you may want to invest in companies that promote and comply with best ESG practices, but you’re not very well informed about the potential financial outcomes of that decision.

When it comes to the intangibles, investors often have to rely on stale and backward-looking data, or what companies say about themselves in press releases and sustainability reports. Our analysis is more evidence-based.

75

OK, so α-DIG does text mining instead. What’s that?

75

Text mining is basically an artificial intelligence technology that empowers users to transform the content in text documents into quantitative, actionable insights. It uses machine learning techniques such as neural networks to infer the meaning behind text. α-DIG text analyses Dow Jones financial news going back almost 20 years.

75

Why not social media?

75

We aren’t saying that financial news media are always correct, but they are less susceptible to fake news than channels like Twitter and Facebook.

75

Can you give me a practical example of how a portfolio manager would use α-DIG?

75

Let’s say there has been an oil spill. That manager can look back over the last 15 or 20 years, and analyse the share price reaction and market value of oil companies that had similar issues in the past; our text mining techniques are able to identify the most similar past events for more accurate comparison.

She can then use the tool on an ongoing basis to evaluate whether this event, and subsequent issues around it, are already incorporated in the price. That helps her to appropriately adapt her exposure to companies that are affected by these situations.

75

α-DIG is part of what seems like a larger trend of mining big data.

75

We are living in a world of expanding data availability from various sources, which are potentially interconnected in a myriad of different ways. There is a wealth of information out there that the human brain alone cannot process.

It requires computer power and storage space, but also data science tools to understand which data are valuable and how they potentially connect with other adjacent data. There are many non-linear interactions between data sets that go beyond the traditional statistical techniques.

75

Data science sounds a little like alchemy: combine and mix different data to try to create gold.

75

Think of data science as a toolbox that is useful for extracting knowledge and insights from data that may come in various forms. The potential is huge. But we should also be aware that many big data sets will just be noise. They won’t be particularly useful for predicting share prices or economic growth, for example.

About #MyDataAndI

Data is the stuff that dreams are made of. On the basis of data-derived insights, big platform businesses and small start-ups alike seek to make individualised offers to potential customers. Many people are just now beginning to understand the value of their data. They want to maintain control over them and want to know what is being done with their data, by whom and why.

On this website, the avatars Amy, Julian and Marcus invite you to chat with them – about what your data is worth in the “data economy”, about the ways you can profit even from your bank data, and about data protection and data security standards. The three fictional characters assume the perspectives of our diverse customers and pose questions that we all have. Openly, curiously and critically. Talk with them!