Whether you're a new investor, or one that's been in the trenches for a while, one of the most important things you can do for your portfolio is dig deeper into the data behind your favorite companies. This is where data analytics begin, but they can go pretty deep and give you lots of useful information to work with.
Definition
What is data analytics?
Put simply, data analytics is the discipline of analyzing data using specific techniques. Often, data is sliced in very specific ways to answer questions you might have about the information you have in front of you. For example, you may want to know how frequently your favorite index fund experiences a correction, and how deep the typical correction goes.
Although this would be an incredibly basic data analysis, it is data analysis. You can also use data analytics to look at all kinds of things, such as comparing two seemingly unrelated variables to look for a mathematical relationship (see our article on The Monday Effect, which explains some pretty wild stock price correlations discovered using data analysis).
Types
Types of data analytics
There are four main ways to describe the types of data analytics. Many times, the actual practice will overlap several categories, but this is a good guide to help you better understand how data is examined:
- Descriptive analytics. This is the most common type of data analysis, and most people regularly do it without even knowing it. In descriptive analytics, you simply describe what's happening in the data. Is the price of your stock increasing? Are same-store sales down year over year? Those are descriptive analytics.
- Diagnostic analytics. Diagnostic analytics explain or try to explain why something happened. For example, you could easily see how vaccine manufacturers started raking in huge profits immediately after the COVID-19 vaccine was available.
- Predictive analytics. Predictive analytics try to guess what's going to happen next. Technical traders use predictive analytics very often to try to guess when a stock will drop based on various signals, or when it will rise again so they can sell it for a profit.
- Prescriptive analytics. Prescriptive analytics are actionable and often used in manufacturing or other industries where staying ahead of the game is vital. In prescriptive analytics, a homebuilder may run the numbers, see that this will be a year of higher interest rates and increasing lumber prices, which will reduce the buyer pool by 30%, so they will build fewer homes at a higher price point to avoid oversaturating their market.
Techniques
Data analytics techniques
Just like there are several types of data analytics, there are many techniques that are popular among data analysts. These five are the ones you're most likely to encounter:
- Regression analysis. In regression analysis, you use one or more independent variables to try to explain a dependent variable. For example, if the summer temperature exceeds 90 degrees for 100 days, it may cause people to buy more beer, which then increases the sales of domestic light beer and raises the profits of beer companies.
- Factor analysis. Factor analysis is a bit more complicated and is primarily used by technical traders. In this analysis, a complex dataset is reduced to a few variables to try to discover hidden trends. This is sort of the heart of things like technical trading, really.
- Cohort analysis. In cohort analysis, you'll break your data into related groups of similar data, like customer demographics or elements that certain businesses or products have in common. This way you can look at that specific subset of data and try to tease apart the things that make them the same and that make them different.
- Time series analysis. If you track cyclical companies, you've probably done some kind of time series analysis. In this technique, you look at data over time and spot the trends that emerge. It's great for understanding businesses that have seasonal cycles like real estate, or longer cycles like certain durable goods, including automobiles.
- Monte Carlo simulations. Monte Carlo simulations are complicated processes that do a great job at forecasting risk. Incorporating many pieces of data, the analysis simulates outcomes and can help in areas like risk mitigation and loss prevention.
Related investing topics
Why they matter to investors
Why data analytics matter to investors
Data analytics are vital tools for investors. They not only allow us to see what our investments are doing, but they also help experts in the field to better understand and describe what's happening within their companies or help improve processes within their businesses.
It's likely you'd not have modern investing without data analytics, at some level. Sure, you can buy good companies that you believe in, but without data to back up your choices, you're mostly just choosing those companies on vibes. How would you know a company is outperforming its competition if you had no way to compare the industry to itself?
