A colleague recently asked me to verify that the DRAM business has had zero net profits over its entire history. This is something he had heard at a technology event that really surprised him.
I have often heard this story myself, for DRAM as well as for flash (both NAND and NOR) but I had never put in the time to test the assertion.
This statement is certainly attention grabbing, and because of that presenters everywhere will find some way to include this “fact” into their slideshows. “But is it true?” he asked me.
Well, I can’t call myself “The Memory Guy” without having an answer to this question, so I ran it through Objective Analysis‘ historical DRAM profit model.
Objective Analysis doesn’t track each company’s DRAM profits individually. DRAM maker profits often must be estimated since a large share of the product is manufactured by conglomerates who don’t report profits down to that level.
Instead we keep a DRAM profit model that implies costs based on market behavior. The model was originally derived through a very large historical database of highly-detailed information: manufacturer costs for each device density for individual manufacturers as well as each manufacturer’s surveyed quarterly unit shipments, going back to the dawn of the DRAM market. This great collection of numbers produced a surprisingly uniform result that we found could be modeled very easily using a more general market behavior method. The model’s output compares well against combined company profit estimates that certain stock analysts have been kind enough to share with us from time to time.
The chart in this post shows aggregate profits for the industry from 1991 to 2011. Clearly the average over time is greater than zero. In fact, the model shows that the aggregate profit over the past 21 years has been nearly $200 billion. This is out of $477 billion in revenues! It’s a pretty good market for those who can stay in it.
Staying in the DRAM market is indeed a difficult trick, a challenge that is spelled out in detail in the Objective Analysis Brief: Why the DRAM Market Must Consolidate, which can be purchased for immediate download from the Objective Analysis website.
Great post!
Indeed, the memory market is tough. I have been in the market since 1997. My email spawned from the warfare climate of the market. So many variables are in motion in the memory market and it does become hard to make profits. Demand must be there or the supply must be short. Memory and technology often are always in decline in cost and resell price. Over supply generally is the culprit that causes prices often to go below DRAM maker costs. Of course the true cost is never known. Memory can be a very lucrative business if you have a great brand and are not greedy. Speculation causes many companies to over stock and often times they get hurt with inventory that is high cost and under valued. DRAM makers often hold prices up to keep the market stable and when quarter or year end come they dump, punishing the market and all levels of the channel. Companies with strong branding and low inventory thrive.
You can make money but not all the time, great companies survive during the lean times and have cash and power to make it. These same companies feast in the good times, Samsung is one such company.
Scott, Thanks for the comment. You really understand.
I should be more careful, though, to make it clear that I am talking about chip makers. Module makers like your company have a different concern: “How do I make sure NOT to get caught with inventory at a time when prices are sliding?”
This can be really tough, because the steepest price slides occur immediately after a critical shortage. Intel learned this lesson in late 1995 when it held a large DRAM inventory for which it had to write off something like $12 million. At the time Intel was making motherboards and grew its DRAM inventory a lot during the shortage that ran from mid-’92 to the end of 1995.
Something that Objective Analysis specializes in is the timing of oversupplies and shortages. Our clients can save money if they can avoid repeating Intel’s mistake.