Shaw Communications (SJR) has segments that include Wireline customers for businesses and consumers and Wireless customers for prepaid and postpaid plans. The  company has recognized around 1.3  billion (CAD) for recent quarters, of which around 300 million (CAD) is earned from the Wireless segment.

An interesting pattern emerges from the customer segments. Shaw is losing Wireline customers but gaining Wireless customers, and most of the decline in the Wireline segments are from consumers rather than business customers. What does it mean for expectations of future revenue if a company is losing customers in the segments that make up around three quarters of their revenue and gaining customers in the segments that make up around a quarter of their revenue? To attempt to answer this question, I ended up:

  • Hand-collecting 16 quarters of revenue, customer, and segment data from SEC filings.
  • Created data visualizations of total revenue, segment revenue, and numbers of customers over time.
  • Split the data into a “training set” for parameter estimation for quarters 2/28/2017 through 11/30/2019 and an out of sample “test set” for quarters 2/29/20 through 11/30/20.
  • Estimated the average revenue per customer based on univariate OLS models, excluding constants, for each segment before combining the estimates into predictions for total revenue vs. an OLS model with total revenue as the dependent variable and including the total number of Wireline and Wireless customers as the independent variables.
  • Evaluated the relative performance of each of the above models before selecting the total revenue model as the superior model.
  • Used time-series models of Wireline and Wireless customers, where the estimation models were OLS models including linear time trends, to predict the number of customers in each segment for four forward quarters.
  • Used the forward estimates in the sixth bullet combined with the parameter estimates in the fourth bullet for the total revenue model to predict four forward quarterly revenue estimates.






My opinion is that even though Shaw is losing customers in the segments responsible for approximately 75% of their revenue and gaining customers in the segments responsible for approximately 25% of their revenue that this is not as ominous as it may appear at first glance. The relatively positive total revenue trends, in light of shrinking numbers of customers in particular segments, may be explained by: pricing behavior, bundling, consumer behavior, or perhaps the type of customer being retained vs. lost. This last point is something that has been discussed in recent earnings calls. Alternatively, a more nefarious explanation would be earnings management, but hopefully that is not the case due to management’s integrity and/or oversight by the auditors.

You can find the hand-collected source data, a PDF of the combined results, and the Python code in both .ipynb and .py files in a GitHub repository.

Disclaimer, Caveats, and Methodological Assumptions:

This analysis contains my own opinions and assumptions. This analysis is not investment advice and should not be relied upon for investing or trading decisions. This analysis and write up are for informational purposes only. I have no business relationship with SJR and am not being compensated for this article. I have no investments in SJR or related to SJR as of 3/3/21, but I am planning on taking an equity long position in SJR in the near future.

The impact of bundling services or potential overlap between Wireless and Wireline is something I thought about but didn’t have a solution for. My thoughts are that the total revenue model may account for the impact of bundling (if any) by including the respective customer totals in the model.

I did not separate out business vs. consumer customers due to the small sample size used in the analysis. Estimating a third parameter with such a small sample was something I ultimately decided could do more harm than good.

The sample size is small, but the tradeoff is the estimates are derived from the most recent data.

All customer revenue is deemed to be variable via the exclusion of constant terms in the OLS models. The reasoning here is, even if you consider fixed contract prices, no customers still means no revenue for the company.

The four forward quarterly revenue forecasts assume SJR will continue to gain Wireless customers ending at 2.1 million wireless customers for the final quarter. This may be unrealistic given the size of the market or for various other reasons.