Back in September I had forecasted AMD would be heading for a GAAP loss in Q1 or Q2 2007:
http://sharikou180.blogspot.com/2006/09/ati-revises-quarterly-guidance-down.html
Having seen Bob Rivet's presentation at their recent Analysts Day, I think it's time to reiterate that prediction for the following reasons:
1. While AMD eked out a 56 million $ profit in Q306, ATI lost almost the same amount. And...it does not look like ATI is going to win back market share from Nvidia in a hurry.
2. Their Q306 adjusted debt to capital ratio is at 37%. Of their plans to bring this down, I think "Improvements in working capital" & "Using excess cash flow" are questionable as they seem to make the assumption that they will continue to enjoy the same kind of profits they did in 2006. We shall have to wait and see on this one.
3. There is a 380 million $ acquisition related (intangibles/synergy related costs) write down which Rivet says will be front end loaded in 2007.
All this is going to impact their bottomline. I repeat - AMD is heading for a GAAP loss in either Q1 or Q2 2007. Intel on the other hand seems to have dodged the bullet the good Doctor Sharikou has been asserting and it looks like yet again one of his predictions is going to fall flat...since it is highly unlikely Intel will have a GAAP loss in Q406. Which incidentally was the fall back position when they did not have a GAAP loss in Q206 as he had originally predicted.
Subscribe to:
Post Comments (Atom)
10 comments:
I think this prediction is a bit iffy... AMD has been growing mobile pretty well (and losing desktop share) - this is essentially replacing low end margin with mid-level margin. (growth in mobile is what has been sustaining Intel for sometime)
That said I believe AMD stated they do not expect the graphics side to be profitable and revenue from chipsets will entirely be tied to AMD PC market share going forward while still competing (theoretically at least) with Nvidia in this area.
Couple this with increasing equipment depreciation and payments on debt, I think AMD's Loss/no loss is dependent on whether Intel eats back into the server market share.
This seems like a distinct possibility but the server market just moves so damn slow it may not happen right away... just look at how long it took Opteron to get share.
I would say 40% on Q1/Q2 GAAP Loss.
PS - I also think AMD had some one time benefits (maybe tax benefit if I remember correctly?) which helped the Q3'06 #'s
My $0.02.
Very good points anonymous, and I'd have to say that the server market definitely wont change enough to make AMD specifically suffer a gaap loss, especially if they finally have their laptop platform down and running by q1, which will easily give them a huge boost.
By the way, nice job leaving that out 180, unless that's coming out later than I thought (didn't they say it would be q4 06?).
Guys - are you not reading the post. Did I say anything about market share? The math is simple. ATI losses will close to wipe out AMD quarterly profits until they can fix that biz. Then, AMD has to take a 350 million USD write for the acquisition which Rivet says will be loaded in 1H07.
Let me also put some perspective on the server share issue. AMD was growing their server biz by close to 20% QoQ till the Woodcrest launch. In Q3 that growth slowed to under 10% (BTW - Q3 is when seasonality kicks in and the PC industry sees growth). While you're right that server business takes longer to swing, Woodcrest is strong in the lower 2P end of the market that has shorter lead times. Intel has also been sampling large customers much prior to the launch...which is why they were able to ship a million units and it went to 40% of the 2P market in just the first 3 months.
(FYI - Intel saw double digit server unit growth in Q3...after a long time)
Yes, AMD will see some upside through their mobile business definitely. But not enough to stomach a 350 million $ write down.
Rivet has also budgeted a higher rate of tax for 2007 in case you missed that.
Greg - winning a multi-thousand mobile platform deal in a large enterprise takes time from tender to delivery. The point of platforms is longevity/consistency for the customer. You don't launch a platform on day 1 and expect everyone to fall at your feet until you can convince them there is a stable path for the platform over time man.
Bottomline - the server/mobile share arguments are marginally relevant because they cannot make up a 350 million $ write down through these.
I don't think it will matter. I'm still guessing that during the next several quarters Intel will have an average loss to AMD of 0.6% volume share.
Well...I don't share your view on that. I think AMD will gain some share in mobile - but primarily due to Dell. However, they will lose share in servers and desktop.
I'm talking about unit share - not revenue share.
cheers...
That doesn't seem to reflect the definition of GAAP loss 180. At least, not as I understand it. Could you elaborate a bit more then.
Also, I'd like to point out that if ATI is losing share, that doesn't necessarily mean AMD will be losing a lot of money, due to the low margins of the graphics market and, according to you, as I really don't know, the very low margin chipset department.
Even then, with a platform deal, ATI should gain marketshare in at least chipsets (because Intel wasn't really making a huge push for their chipsets anyway) and their graphics department should do fairly well if they can up their 80 nm production with ddr4.
Seeing as dx10 games are not out yet, and how extremely hard to use and expensive the g80 is, ATI could make a small gain just before the first dx10 games come out. Nvidia needs something that's easier to use than their current setup if they want to compete with what ATI has now.
Still, I suppose that Intel has a lot to do with what graphics platforms are most acceptable by the market and thatTATI wont regain their former "glory" for quite some time (part of the reason AMD specifically mentioned their graphics department would no longer make as high of margins).
However, making as sure a claim as you have about GAAP loss does not seem to me to be warranted.
That's what I'm talking about. I see a steady decline of 0.6% volume share per quarter from now until the end of 2008.
Unfortunately, this could be masked in Q4 and Q1 so we may have to wait until Q2 07 to see what is really happening. Basically, if Intel takes volume share in Q2 07 or holds onto whatever apparent gains it had in Q4 and Q1 then that is a genuine gain and AMD's loss.
"Let me also put some perspective on the server share issue. AMD was growing their server biz by close to 20% QoQ till the Woodcrest launch. In Q3 that growth slowed to under 10% (BTW - Q3 is when seasonality kicks in and the PC industry sees growth)"
The thing you are missing is that the revenue was still growing - unless that turns negative (and eats into net profit) AMD will continue to get the revenue/profit from this section of their business.
Keep in mind the overall semi market will grow so even if both AMD and Intel are flat from market share perspective they both will earn slighty additional revenue (and presumably profit) as well. I think last estimate was about 10% (I could be wrong on this) for 2007, so even if market share is flat that's still 5% growth in H1'07 for each company.
So again I think whether AMD has a loss will be dictated by whether they actually see server revenue/profit growth turn negative - assuming 5% industry growth in H1'07 this would mean AMD could lose as much as 5% of their server market share (~1-2%?) and still be revenue neutral in the server area.
I also think mobile share will continue to grow a bit over the next few quarters so this should help AMD's margins a bit as well; losing desktop biz may be a blessing in disguise but unfortunately for AMD they are seemingly taking the biggest hit on the high end desktop areas.
Finally it will be interesting to see if Nvidia sells many chipsets for AMD CPU's or if AMD starts to freeze them out - the ATI chipset revenue is taking a huge hit as all their supporting Intel chipsets are or soon will be gone, and the AMD CPU share is not enough to make up for this, especially if Nvidia is also competing in this space.
All that said, I think with 65nm ramping up (should help cost picture) and continued strength in server, AMD will eke out some Q1/Q2 profits.
Now Q3/Q4 will be interesting as conversion to Core2 will be complete and I don't think K8l will be out in substantial enough volume to noticeably impact these quarters
Anonymous - yes...they may be revenue neutral on the server biz. But that's a reversal from seeing QoQ growth. However, all that does not take away from the fact that ATI is going to be a drag on profits and they have a 380 million $ write off in 1H07 as per Mr Rivet.
I think what will help them is an improvement in the pricing environment. But even that is not enough if they take the write down.
Post a Comment