Just how bad would the embedded SIM be for the smart card industry?

SIM shipment volumes grew by a whopping 30% in 2010, more than four out every five SIM cards in use were replaced over the last 12 months, and the SIM card installed base recently broke the five billion barrier. Things are looking pretty good, but could the party come to an abrupt end if the removable SIM model was replaced with an embedded model?

By Alex Green, Senior Research Director, IMS Research August 30, 2011

OK, so SIM shipment volumes grew by a whopping 30% in 2010, more than four out every five SIM cards in use were replaced over the last 12 months, and the SIM card installed base recently broke the five billion barrier…. Things are looking pretty good, but could the party come to an abrupt end if the removable SIM model was replaced with an embedded model?

It’s a scenario that has received a lot of press over the last year. A number of smart phone manufacturers, such as Apple, have threatened to go down this route, and it appears to be the preferred route in a number of M2M applications. That said, a rear guard action from the GSMA, SIM Alliance and a number of MNOs seems to have stayed the execution. However, I have recently explored what impact this would have on the SIM market and wider smart card market should it occur.

To understand, let’s first look at what’s driving the SIM market. Of the 4.2 billion SIMs sold in 2010, 85% were to existing SIM card users, i.e. existing users of cellular handsets. This means that more than 80% of SIM-based cell phone users replaced their SIM cards last year. This may seem a little surprising as the proportion of cell phone users that replaced their actual handsets last year was much less than this. In fact at a global level, only 24% of cell phone users replaced their handsets with brand new handsets last year.

So why did so many more replace their SIMs?

Well there are a number of factors that drive this. These include the fact that in some circumstances when the services in a subscription are changed a new SIM is provided; there is a large second hand cell phone market and typically these still require a new SIM card; in some circumstances when a subscriber changes operator/service package without receiving a new handset they may still need a new SIM; in some cases pre-paid SIM cards expire meaning a new ones are required; certain users use more than one SIM card in order to get the best location/time specific tariffs and some handset allow for more than one card to facilitate this; and new SIM cards are require due to damaged or loss.

Now what would happen if the SIM market moved from a removable model to an embedded model? How many of these additional replacement drivers would still exist? The answer, because of demand or technical reasons, is none.

The SIM replacement market would become tied to just new handset replacements. With the new handset replacement rate being less than a third of the SIM replacement rate and SIM replacements being by far the most important part of the SIM market, the impact would be hugely negative.

With the current removable model, the volume of SIM cards sold in 2016 is set to pass 6 billion. If all cards were embedded by this time the number would be nearer 2 billion, around half what it was last year! As the SIM card market represents more than 80% of smart card volumes, this would be very bad news for many card and semiconductor suppliers.

The good news for them is that we think the chances that this scenario will come to pass is low….. Well, at least in the next five years. Embedded SIMs are forecast to be used in some tablet devices and some M2M applications in the short- to medium-term and, the chances are, in some smart phones eventually. However, for the vast majority of cellular applications it’s my view that the removable model will be here for some time yet. That being said, the industry needs to recognise this long-term threat and start to look at spreading the risk associated with so much dependence on this application. The scenario painted here may have a low chance of occurrence in the next five years, but my view is that over a 10 year period, the chances are much higher.