Hello there. In this post, I talk about what is seasonality from statistics, finance and economics. Seasonality refers to time indexed data that captures regular and predictable changes that recur in a calendar year or cycle.
Data is all around us. However I do think that we overlook that time aspect with data. A student average of 80% in grade six in 2014 is much different to a student of average of 80% in grade twelve in 2020. Both averages are 80% but the later average in grade 12 is more recent and more impressive as grade 12 is tougher than grade six.
A low grade from 20 years ago does not hold much weight now compared to a good work performance from a week ago. I am going with the philosophy that recent data holds more weight relevant to the present versus data from a while back.
A common example of seasonal data is weather data. We have the seasons of fall, winter, spring and summer along with their associated common temperatures. Temperatures in general will follow a cycle where there are rising temperatures, decreasing temperatures or temperatures that remain at their maximum or minimum or average. We do have to factor in some randomness with the temperatures. The average temperature for a Winter day in the Toronto area will not always be at -5 degrees Celsius (as an example). There are some days that are colder than -5 degrees and other days that are warmer -5 degrees. You can even compare across different winter years where the average Winter season temperature may not be always -5 degrees.
In the business and finance setting, we do have seasonal data in the form of quarterly data. Quarterly data is represented as Quarter 1 (Q1) to Quarter 4 (Q4). Each quarter refers to a 3 month period. Some businesses earn more revenues in certain quarters than others.
| Quarter Number | Months |
|---|---|
| Q1 | January, February, March |
| Q2 | April, May, June |
| Q3 | July, August, September |
| Q4 | October, November, December |
Reference: https://www.investopedia.com/terms/q/quarter.asp
Major shopping events such as Black Friday, Back To School Season, Boxing Day and Valentine's Day are contributing factors for increased revenues and profits in certain seasons. As an example, technology companies do use the Back to School Season and holidays to launch new products to the public around those times. The travel & tourism industries along with the airline industry do not earn much revenues in seasons under cold weather compared to seasons with more warmer weather.
Seasonality in data is useful for analysts to spot patterns and trends in time-based data. The field of statistics and econometrics that deals with time-based data for analysis and prediction is called time series analysis. A pre-requisite for time series analysis is linear regression as there is statistical modeling involved.
Given historical time-series data, a statistician, an economist, data scientist or a finance practitioner would use statistical tools to determine a best fit model for that data. Model selection involves the usage of the Autocorrelation Functions (ACFs), Partial Autocorrelation functions (PACFs) and model fit measures. Different people may choose the different model based on philosophies and model fits criteria.
The fitted model for the data can be used to make predictions for prices, number of expected customers, or any measure over a certain time period. Time periods for predictions could be over one day, one week, one month, one quarter, one year and so on. Predictions that are farther away from the present time do contain more variance/error. (It is difficult to predict far out in the future.) The time-series predictions does its best with the predictions without given financial nor business contexts.
Side Story
I once did a time series prediction for an escape room games company that had actors as add-ons to the escape room experience. The historical data was based on about 2 or 3 years of data with quarters. I used a time-series model called the Holt-Winters model for capturing the seasonality in data and the overall trend for the company's revenues. If I remember correctly there was something weird looking in the data and in the predictions. The representative for this company at the presentation mentioned that the weird looking part of the line graph was because of a new escape room release that I did not know about. The point of the story is that you may not get great predictions or data due to missing contextual information.
This section has more of a focus on retail shopping. I would not know much about business strategy.
Seasonality can be used to time purchases. You could avoid paying more than necessary during an in-demand season on certain goods and services. Get more for less and know your prices.
Groceries & Food
Right now it is winter outside the Toronto area in Canada. We do get imported fruits from the USA, Latin America and Asia. Strawberries are less common and are priced higher in the winter versus than in the spring and early Summer.
There are restaurants that do change their food offerings and prices on their menu based on the season. Example includes seasonal pricing on seafood items and cold ramen soup and noodles during the summer.
Flights
Christmas Day is a day when families come together and people are opening presents. However, there was this one guy I knew who traveled instead. He booked a flight from Toronto to China on Christmas Day 2018 as the tickets were cheaper.
Back To School Season
The back to school season is filled with sales on school supplies, furniture, laptops, computers and technology. Although the sales are targeted towards students, non-students can take advantage of the sales too. Walmart and Staples (Business Depot) typically have very low prices on paper, pencils, pens and paper notebooks. There are phone makers that use the back to school season to launch new phones along with advertising campaigns that aim to entice people to get "the latest and greatest".
Black Friday & Holiday Seasons
There are good sales during Black Friday and the holiday season on select items especially technology. You do have to careful with some of the sales as some sale prices are actually the same prices as original prices from a different time period. You can use a price tracker website such as camelcamelcamel.com for Amazon prices.
Real Estate
When I refer to real estate, I am talking about the residential real estate market. Even though location is one big factor in housing prices, the prices are also affected by seasonality due to weather. In the Toronto area here, there is not much activity from buyers and sellers in the winter versus the spring and summer when the temperatures are warmer. Holidays and back to school seasons are also things to consider when buying and selling a home.
More Details On Real Estate Seasonality in This Page
Thank you for reading.
