As most of you know, I'm not the smartest stock investor around and that's why most of my money is in diversified ETFs and you rarely see me write about stock investing.
My saving grace is that I try to surround myself with smart people who are knowledgeable in the topic of investing. That allows me to pick their brain and learn more about investing through their lens.
In this article, I interview Jason, the man behind Data Science Investor to learn more why he started his website and how he combines data science and investing to make data-driven investment decisions.
Who are you and what is Data Science Investor?
Hi guys, I’m Jason (the creator behind Data Science Investor). In my day job, I deal with business development and innovation topics for an MNC. In my spare time, I used various data science tools and my financial knowledge to create interesting models, visualisation tools and informative articles for readers on my site to better understand key information in the areas of stocks and Singapore properties.
Data Science investor was created with the objective to provide more data-based articles on general investing in stocks and properties for the general public. I notice that there aren't many blogs which focus on deriving data-based insights for the general public in the area of investing. In the area of stock investing, opinions are plenty but not many actually deep dive into the data (prices, technical indicators, sentiment analysis etc) to provide data-based analysis/insights for the general public to aid them in their decision making.
This probably even more true for the property scene where I personally think that is a huge information gap for the buyers/sellers. As buyers/sellers, most of us do not really have a lot of data or data-based insights with us when we are purchasing/selling/investing in properties. Most of the information available is usually focused on selling you the property instead of feeding you data-based insights.
Seeing there is a noticeable gap there, I then decided to create Data Science Investor late last year.
What’s in your investment portfolio?
Currently, my investment portfolio could be divided into two portions, with a 70/30 split. The first portion with a 70% allocation (let’s call it “Permanent Portfolio”) is mainly on indexes, ETFs and some high dividend yield blue chips. The movement for this “Permanent Portfolio” is more limited as I don’t trade them often. The second portion with a 30% allocation (let’s call it “Trading Portfolio”) is based on a trading algorithm which I created based on value/quality investing principles and trend investing methodology to attempt to beat the market. Movement within this “Trading Portfolio” is more as trades are done more frequently based on the pricing momentum of the stocks.
What are your immediate investment plans?
I’m not making too many changes here due to the current Covid-19 situation. My Trading Portfolio remains the same as I still stick to the same trading algorithm. I might be making a bit more changes on my Permanent Portfolio to have more cash on hand instead as there is a bit too much optimism and “hopium” in the market right now and that doesn’t seem to be a positive sign to me.
Describe your investing strategy
Unlike most investors, I don’t usually spend too much time deep diving into a particular stock, determining its fair price and making trades based on it. I prefer to spend my time sharpening my trading algorithm (with my investing principles based on fundamental analysis built in) to try to achieve a higher alpha and a lower drawdown in my portfolio. I’m a believer that a company changes all the time, but fundamental principles don't hence I’m more focused on ensuring that I get my algorithm right rather than spending too much time analysing a single company. Sometimes, refining such algorithms could involve the use of data science tools too to better understand when to buy, sell or hold.
Tell me more about how you are using data science to make data driven investing decisions
In the area of stock investing, data science tools could be used to create classification models (such as random forest etc) to help individuals better understand when to buy, sell or hold. Information used to build such models could be features such as technical indicators (SMA 20, 50 etc), price actions etc from the historical information of indexes or stocks. The general and simplified idea behind it is that the use of such classification models could help to unravel several significant statistical relationships between the features and the correct trading decisions such that the model could eventually aid you in having a higher chance of making the right trading decision when presented with the current values of the features.
In the area of property investing, data science tools could be used to create regression models so that you could obtain data-inferred prices of a particular property you are interested in so that you will know if you are overpaying for a certain property. Similar to what was mentioned above, you will also need historical information as features to build such models. In this case, the information used here could be town, remaining lease, storey range etc. While such methods might not and should not replace professional valuation, it sure does at least give you some sort of price indication to ponder upon.
Data science works well when you have a rich source of data to analyse. Where are you get your data sources from?
Data sources for stocks are quite plenty. If you want historical pricing information, you can easily obtain them from Yahoo Finance etc. If you want some specific information such as P/E ratio, P/B ratio etc, there are several sources such as Morningstar, YCharts so you will never run out of data sources for stocks.
For Singapore property, you could also easily obtain transactional information from the URA website or data.gov. Thankfully, the government has made such data very accessible to the public and that makes a lot of the analysis I’m doing possible.
While getting data is easy, the most often neglected part is how do you “cleanse” or preprocess the data to ensure that you have “good” and “clean” data going into the model. Otherwise, you might just be getting less than satisfactory results from your data models.
Regarding data science, I imagine it’s easy to find a rich data source containing data points of lagging indicators but challenging to find data points of leading indicators. What are your thoughts on this?
That’s actually a pretty interesting question. I will not say that it’s easy to find data points of lagging indicators or challenging to find data points of leading indicators. This is often misunderstood by most people. In fact, it’s usually equally easy to find data points for both. What makes people think that the data points of leading indicators are hard to obtain is the fact that leading indicators are very hard to be measured unlike lagging indicators. For example, it’s easy to measure your weight (lagging indicator) but difficult to measure the calories of the food you are taking in (leading indicator) to understand if you are able to successfully lose weight in a month.
Regardless, both indicators are equally important to be used as features for data modelling in data science. Depending on what you are trying to predict with your data model, one factor might have an advantage over the other but I do not think that necessarily has relevance with whether it’s a leading or lagging indicator.
Do you intend to go into building predictive models to identify stocks to invest in?
I do build predictive models to allow me to make better judgement calls in “Buy”, “Sell” or “Hold”. For identification of stocks to invest in, I do create trading algorithms with my criteria built in to identify value/quality stocks with good momentum for my trading portfolio. However, I don’t think the trading algorithm is predictive in nature in this case.
What else is in your financial plan?
Besides having stock portfolios, I’m also looking at the use of my home/property to further accelerate my retirement plans by buying and selling at the right value.
On the financial side of things, what were your growing-up years like?
I came from very humble backgrounds and there was almost no one who taught me the importance of investing or being financially savvy when I was a teenager or young adult. Most of what I understood today are self-taught with my first foray into investing coming from a chance encounter of the book “The Intelligent Investor”. That happened when I was 18 or 19 and I didn’t actually have any capital to do any form of investing until I started earning my own when I graduated. Since then, most of what guided my current investing philosophy or methodology comes from multiple lessons of making the right and wrong investing decisions in the past years. As my primary job also involves tech, I then decided to marry both tech and financial stuff which I knew to attempt to achieve better investing results. Of course, a by-product of it is Data Science Investor 🙂
How did you get interested in investing?
My main interest behind investing is the possibility of using investing to build up as much capital as early as possible to allow me to be financially free. This is a very important milestone to me as I think that there are a lot more things that could be done in one’s life once he or she is no longer held ransom by money. Coming from humble backgrounds, I understand how much of an impact not being shackled by money could have on one’s life choices and decisions and that drives me to pick up investing to accelerate the path to be financially free.
What does money mean to you?
I see money as a limited bag of seeds. If you don’t learn to grow them and just consume them, you will run out of food soon and be constantly worrying about your next meal. If you learn to grow them, you will have an abundance of fruits and they will be the least of your worries as long as you don’t over consume them. And once you have an abundance, you could then use the excess to help more people either by directly giving them the fruits from your seeds or teaching them how to grow their own seeds.
How are you planning for retirement?
Retirement will be achieved when I can sustain my ideal state of lifestyle without having to actively work for it. I’m still a long way to go for this but I’m hoping to achieve it before I’m 50. Some milestones which I will like to hit along the way will be to achieve a certain capital size of portfolio in the next decade or so, followed by achieving constant sizable passive dividends from this portfolio in the later years leading to my 50.
Who do you hope to reach with your blog?
I hope to reach out to anyone who’s interested in getting data-based insights to make stocks or property related decisions. Such insights are constantly shared in my blog and I believe they should be of use to anyone who is looking at past trends, key correlations and future possibilities in the area of stocks or property. Ideally, I also do hope more people could be able to better understand the importance of making sense out of historical data or information when it comes to investing.
What should readers be expecting to read in your blog?
My thought articles, summary of historical trends in properties/stocks and various important visualisation tools such as property maps to allow readers easily locate property of their choice. These are primarily what I have been writing/creating in my blog and they will continue to be in the future.
Do you have an interesting personal finance story to tell?
I'm always on the look out to profile financial bloggers and content creators who are doing great work to find out the story behind what they do.
Contact me with your story and if it's interesting enough, you'll earn a spot on my blog.