It is not unusual for people in the industry to talk about the semiconductor business model as it is something common and static for all companies in the market. While there are a lot of commonalities there is also a lot of interesting differences that can reveal a lot about what strategic choices a company is pursuing. Despite quarterly updates from the CEO, the format is dictated by Wall Street analysts trying to get numbers for their financial model so they can predict (guess) the numbers of the quarter. It is rare that something interesting is revealed. I cannot blame the leadership teams as they have lawyers and analyst hovering over them just waiting for a mistake.
What is much more useful is the financial reporting that public semiconductor companies have to deliver every quarter. As the format is reasonably stable it is possible to compare changes over time and also comparing companies to each other. This analysis is made for Semiconductor companies only – in our definition is public companies that have semiconductors as their only business and sell them to electronics customers. This excludes conglomerates as Samsung and foundries like TSMC and represents a choice taken for the availability of data and consistency.
While the manufacturing costs are hidden in the Cost of goods sold, the operating costs are more visible and interesting. Knowing where a company spends its operating dollars tells you a lot about that company.
Don’t tell me where your priorities are. Show me where you spend your money and I’ll tell you what they are. James W. Frick
Other operating costs can exist, but the two most important ones are Research and Development (R&D) and Selling, General and Administrative (SG&A). The implications of a high R&D spend is understood by all people in this industry. It is a clear indication of the level of ambition of a company. SG&A costs are slightly more complicated as it is the sum of the selling costs and the general administration costs of the company, including management, staff functions, rent and utilities. A few companies report the two separately and offer a view into their relationship. The indications are that the G&A costs are dependent upon the size of the company and are somewhere from 2.5% to 7.5% of revenue, As such the G&A costs are relatively stable over time and changes in SG&A is a result of changes in sales and marketing activity.
The Operating Model Grid
A tool to investigate the two costs simultaneously is the operating model grid. The centre of the grid is representing the averages of the companies studied. By comparing companies this way, you can eliminate much of the economic noise that affects the entire market. By comparing similar quarters one year apart, any seasonality is excluded. The periods compared in this grid is Q3-2016 and Q3-2015.
Although it is not suggested that one business model is superior to another, there are four zones in the grid representing four different operating zones when compared to average. Companies in the ambition zone outspend the average company in both R&D and Selling. Smaller companies are typically in the ambition zone as they need high R&D and SG&A just to survive, where larger companies have more options when choosing their operating model.
Companies in the consolidation zone spend less than the average company in both categories. This could be as a result of having a commodity offering where price is more important than selling and R&D is technology focused rather than product focus. It could also be a preparation for an acquisition. Companies in the selling zone prioritise selling spend over R&D. This could be a result of having products that need high support or it could be an indication that the company want to get bought. In the development zone are companies that typically have high volume leading edge products that are sold to few customers. What is more interesting than the zones is the change in operating mode over the year represented by the arrows.
The grid is not an absolute – it is a model to identify interesting questions to be researched. Both Nvidia and AMD have experienced healthy growth, and they have decided not increase the operating costs simultaneously. This could be a lack of confidence that the growth is sustainable or it could be that they are not able to find the right people and projects sufficiently fast.
Hynix and Micron have decided to handle the growth in memory differently. Where Micron keeps their R&D steady and increases their sales investment in line with growth, Hynix puts the pedal to the metal. This is a very strong indication that Hynix has a strong belief in their current strategy.
A merger will typically create a movement towards consolidation. (We compare the sum of the two entities to the newly merged company). The overhead you pay for the acquired company needs to be repaid. Operating costs are fast carbohydrates for merger integration teams.
What is interesting is the three European semiconductor companies all are quite close together and also close to TI. Intel is not surprisingly in the ambition zone but heading towards consolidation. This could be a result of the change from PC focus to data centres, IoT and AI that cannot absorb the drop in spend on PC.
The Semiconductor Operating Model Grid for Q4 will be released in a couple of weeks together with an in-depth analysis of the fourth quarter for Semiconductor companies.
Engagement Group is enabling Semiconductor, and other Hi-tech Companies grow Organically by using the Power of Social Networks.
We create content that engages customers. We create connections to new decision makers. We help salespeople get new revenue from new accounts.We make your organisation ready for the social network tsunami. You need to understand the social principles to survive and surf the wave.
Visit Engagement Group for more information or contact Claus Aasholm.
Interested in Semiconductor market research or social selling? Join our mailing list here.
You can get the free “Selling Insights for the Semiconductor Industry” report here.