a) Progress in coding - Correlation Matrices
I've stopped Python classes at the moment to consolidate what I have learnt by writing useful code to deepen my understanding of our local markets and come up with superior lecture slides. My latest development is that I wrote a short script that can generate correlation matrices.
Apologies if the above printout does not look user friendly at all, but it has interesting insights in that it tries to arrange correlations between local banks against the STI ETF. Predictably the numbers are quite high with banks performing almost in close tandem with each other. With these numbers, I'm not too sure whether a thorough analysis of net interest margins or NPL analysis would add any value to cherry-pick one bank against the other two. May be better to buy all three banks and project their performance as a whole. This can be as basic as anticipating interest rate movements.
A different situation arises if you generate the same table for REITs that are found in the STI.
REITs have a much lower correlation against the STI and when you look at the correlation of REITs against each other, they don't seem to move in tandem amongst themselves. So based on this table, there is definitely some value-add when cherry-picking some REITs using bottom-up techniques. I suspect an industrial Ascendas REITs and Capitamall Trust would really have a different behaviour across time.
By the time I reached this level of coding skill, I can cut and paste fragments of previous code I written to come up with some new functionality that can resolve some intuitions I always had about local markets.
I just need to expand my code to allow me to group stocks into equal-weighted portfolios to see how they behave against each other.
b) Dungeons and Dragons
We played 5 hours of Dungeons and Dragons yesterday using Roll20. The software was glitchy and I did not expect to face bigger problems than my webinars. My campaign is largely homebrew with the latest rule additions from Unearthed Arcana. In this game, I play a Paladin of Conquest.
After sharing my problems with my RPG friends, I was told that some gaming groups use two web apps to run a game.
c) One Book I am reading
With course prep and coding, there is little time to read. John Kay Mervin King has written and an excellent book called Radical Uncertainty. While a little short on practical market advice, this class of books remind us to go back to first principles when thinking about randomness in financial markets.
I particularly enjoyed the author's musings that link legal judgement to statistics and thoroughly agree that law school is basically about taking a very ambiguous situation and slowly using our thinking tools to make order out of chaos - it's a powerful skillset to deal with non-quantitative aspects of investing. It helps me cope with the pain of folks telling me that I have wasted four years of my life studying for a Law degree in SMU when I am doing quite ok on the financial front.
d) Leisure
I am currently following the anime Demon Slayer and find the characters extremely relatable with regards to my RPG gaming hobby. My behaviour in a D&D game is very similar to murder hobo Inosuke shown above. The show is also quite fun given that I can map the neuroses of several demons to actual people I know in the gaming/BBFA community.
There are some pacing issues in the story and I still prefer One Punch Man, but this is something anime fans should not miss.
My class is currently broken down into four sessions that run across four different days, this means that I can actually devote some time into coding to deal with some unanswerable questions from my students and even update this blog.
There is a good chance we can catch up on Sunday evening so I will see you then.
No comments:
Post a Comment