How we cracked the code of private banking
Two weeks ago, we achieved first place in the Delen Private Bank ‘crack the code’ hackathon by creating a tool that helps stock analysts assess sentiments regarding IPO’s in a quick and reliable manner, allowing them to create more value. In 24 hours, we analysed the bank’s inner workings, came up with an idea to improve these and built a first working prototype.
For every IPO, there are many different influences on the stock price. On one hand, we have the numbers, such as financial statements and earning forecasts. These are thoroughly analysed and interpreted by experts at financial institutions, but oftentimes this is not sufficient to accurately predict share prices.
Stocks are traded by humans, who are biased and make irrational decisions — Think Facebook’s IPO. Sentiments and news coverage can heavily influence stock prices, and as such being aware of them is essential to make the right investment decisions. However, there is so much data out there that analysts are often unable to fully and accurately grasp what is being said and felt around the globe. American coverage about a certain event might be optimistic, whereas Chinese coverage might me more reluctant. And how does Twitter react? What about investment fora? On top of the obvious language barriers, there simply isn’t enough time for a single analyst or team to comprehend everything.
That’s where our product comes in.
During the hackathon, we chose Surface Oncology, Inc (SURF) as the IPO we would first use our analysis and prediction technology on. Surface Oncology is a clinical-stage immuno-oncology company developing next-generation antibody therapies that target the tumor microenvironment. The company went public April 19th 2018, one week after the hackathon took place.
Numerous data streams were collected (Bloomberg, Reuters, Financial Times, China Today…) from which articles about Surface Oncology and the biotech industry in general were scraped. We leveraged IBM Watson’s sentiment analysis to assess which (often subtle) emotions were linked to these articles.
Once these sentiments were collected, our engine searched for similar past IPO’s.
In the case of SURF, the engine dug after similar biotech companies that went public in a time of increasingly positive biotech news coverage and positive company coverage.
Using these similar cases, possible stock price evolutions, as well as an expected price can be shown to the analysts.
This use case is only the tip of the iceberg. Not only can our application be expanded to cover all kinds of price evolutions, but it could also be used to track events such as mergers, the appointment of a new CEO and even improve public relations, marketing…
We would like to thank Delen Private Bank, Exellys and Board of Innovation for having us and hosting the event! We tremendously enjoyed our stay, as well as the interesting workshops on business modeling, experiment design, the future of chat in financial applications & more!
