With so much being talked about the ‘data economy’ , let us share a bit on our perspective of this beast called data analytics.
Firstly, let’s answer the most obvious question: Is data analytics all it is hyped to be?
To answer that question, consider what you could have done differently in 2020 if you had perfect insight into what took place over the last 8 months. So by early Feb 2020, your data analytics dashboard should have been screaming “Coronavirus approaching”. If you owned property, especially commercial real estate, you would have sold out at the top of the market. If you were in commodities, your dashboard would have said ” Negative oil prices approaching.” Since you saw the collapse in demand way back in Feb 2020, you would have ample storage and would happily take possession of WTI oil for -$37, yup – people would have paid you take the oil off their hands. A few months later, you would have sold the same barrel of oil for $40, making net gain of $77 per barrel of oil, an unheard of profit margin. Then around March, when equity markets had hit limit down, you would have seen that they would bounce right back up driven by momentum stocks, and piled into Tesla, Apple and NVDA.
And in the space of 8 months while everybody else was suffering, you would be a billionaire.
Of course, not all of that is true. There is a limit to how well you can predict the future given all available information. And that is really what data analytics is all about – the ability to learn interesting stuff that few people know at the present moment.
So how does one become a Jedi Master in data analytics?
There are several key components.
The first would be a solid grounding in Mathematics, especially Probability Theory. Harvard has an excellent undergraduate level course for Statistics called Stats 110. That course will equip with a solid foundational understanding of set theory and probability distributions. For those intending to branch into data analytics for finance (or how to gamble in the Wall Street casino), there are 2 courses that must be taken, both from MIT – Linear Algebra and Quantitative Finance. All the courses mentioned are available for free via YouTube. (Note: Chinese New Year weekend this year was spent understanding what the Null space of a matrix actually meant).
So with that out of the way, one will need to have a bit of grounding on how to get data into your system. As a confessed Pythonista, there is only choice – Python 3.8. With Python, it is possible to analyse data in a variety of formats and from a variety of sources. The thing that I would complain is that Python was maybe a bit weak on XML data sources, I had to spend the last 3 months actually building up my own Python module to handle XBRL data that you find in financial filings.
Lastly one would actually need some framework about the topic being analysed. A framework is like a body of knowledge that gives you intuition about the subject or in the case of data analysts, the data being presented. There is a framework about International Relations, Finance, Marketing – even Dam Engineering, which is currently the focus of The Rembau Times.
All in, it takes years and years to build up skills to become an effective data analyst. And once you are at the top of your game, comes the last bit – the most vital skill, the Trumpian skill. It is the ability to present the most relevant, headline grabbing fact with the supporting analyses tucked away and not getting in the way of the key take away point. After all, data analytics is about one thing – the ability to influence those who don’t understand data analytics to see things your way.