Predicting the future for semiconductor revenues is a risky business
1 min read
The semiconductor industry has shown remarkably consistent growth over the last 50 years. Plot a graph of sales revenue against time and the slope shows an increase in sales of around 15% a year. Of course there have been a few bumps in the road along the way – huge booms in revenues caused by under supply, followed by equally large slumps due to over capacity – but growth in sales has remained fairly predictable.
But during the last few years, revenue growth has slowed considerably. There's a number of reasons, one of which is nervousness caused by the state of the global economy. Semiconductor revenues in 2012 declined by 2.2% compared to 2011, yet remain in excess of $300billion a year. While a couple of larger organisations saw their sales decline by more than 10%, it's hard to agree with one analyst's view that 2012 was 'a terrible year' for the semiconductor industry.
With a level of enthusiasm found only amongst semiconductor market analysts, predictions are now being made that sales will break the $500bn barrier in only a few years – that's 60% more revenue than in 2012.
Hitting that target requires two things: a large increase in the volume of chips sold and a significant increase in the price for which they are sold.
There is the dilemma for the industry. Demand is there – the underlying trend shows unit consumption has grown by 10% a year over the long term – so growing revenues requires customers to pay more per unit. Aligning these two vectors is the issue.
Predicting the future performance of the semiconductor industry is a risky affair – for a number of reasons, it defies logic. I remember listening to the chief economist of a leading semiconductor manufacturer confidently predict that market revenues would grow by at least 30% in the following year just weeks before the industry went off a cliff.
The one thing of which you can be sure when it comes to semiconductors is that 'stuff happens'.
