Since I considered a career switch, I started looking at what I could do. It started off naturally with Fintech so AI, Machine Learning, Big Data and then Cryptos.
In fact, a number of my friends are working on AI currently.
I also did the Oxford Fintech Program, Singularity University Programs.
As I was going through these programs, I thought about what I used to do that could be replaced or improved using AI.
The first thing that comes to mind is I used to scroll through my phone each morning trying to read all the overnight news on the stocks that I covered before I get to work. I used to feel so happy if I could do that before I reached office as that would mean I have a couple of minutes reading some football news.
There is actually so much subjectivity in this process such as I have to decide whether a piece of news is important or not, whether it is positive or negative to share price — deciding all these in a matter of within minutes as I scroll through to the next piece of news.
Where I found these news, ie, the news source are also critical. When I first started I used to get scolded at for missing key news, which back then, I didn’t know where to find. Yes, not all news are from Bloomberg. As an associate, you are expected to be the first to know all the news, which clearly is quite impossible because you are the only person gathering the news whereas if you are on the buy side, you have a whole bunch of people sending you news. Of course, this leads to information overflow, which is another problem potentially solvable by AI.
So I believe AI could help analysts by more systematically collect news, analyse and classify them as share price positive or negative and to what extent. Of course, different people would have different weightings and algorithms running so the answers are different. They may also need to continuously update these to stay ahead of the curve.
Next, a lot of our modelling are done in Excel. There are already algorithms that read financial statements and have Excel spreadsheets as outputs. Therefore, a lot of the jobs currently done by associates could be replaced.
I guess this evolution is a bit like when I heard my senior told me that when he was an associate, they didn’t have Excel. Spreadsheets were calculated on paper and they had to rub off their projections and rewrite if they changed their forecasts.
When we write reports for investment ideas, it was collating pieces of information or Mosaic theory. I think this maybe the part where AI may need more guidance but some maintenance research such as result notes and the daily/weekly/monthly, they could most likely be replaced by AI hence again impacting the jobs of associates or junior analysts.
I also think there will be more generalists analysts that cover multiple sectors or companies assisted by AI. Currently research are usually divided by industries, ie, analyst have specific industry focus. Going forward with the assistance of AI, analysts will have the capacity to cover more sectors and companies. There could also be more focus on thematic investments such as stocks that get affected by the Trade War or One Belt One Road rather than just focusing on the industry alone.
A final point is that when I first started, there are a lot of focus on fundamental analysis. However, in recent years, media and social media analysis have became much more important as they have direct impact on short term share prices. There would need to be more systemic way to analyse these information, which cannot be done by one single analyst. There is also more focus on short term share price movements as investment horizons became shorter.