As some of you have suggested, buying depressed Intel shares is dead money. I’m curious about the math behind $3B in cost savings and it’s impact on earnings, share price and dividends over the next several quarters.
15 replies (most recent on top)
Intel is a victim of Moore's law. Here is how:
- Each new generation shrinks feature size
- Each new generation costs exponentially more to bring online
- Intel innovation is stalled so, the unit volume of products is near zero growth and the mm^2 area of manufacturing has downward pressure because of smaller features size and competition bringing ARM which is more efficient that n-stage pipelines, branch predict blah blah blah.. even Intel had to throw in the towel and now has 'little cores' (sort of).
- Intel simply cannot afford to keep spending on new processes and fabs unless it could somehow grow new businesses. If Intel has to rely on outside innovation and sell fab services to scale, it has to dramatically change how TMG operates -- need lower cost, higher flexibility, level playing field vs. Intel parts, legacy process node support. It just isn't going to happen. You can teach a dog to hunt, but you can't teach a parrot to swim.
@lrg+1kqoazuf Can't just look at the cash holdings in assets without looking at the debt in liabilities. $22B in cash and short-term investments but $37B in debt on the liabilities side.
Look at 10Q at https://www.intc.com/news-events/press-releases/detail/1586/intel-reports-third-quarter-2022-financial-results
@fql+1kqoazuf They need to spend RD like TSMC and build leading edge fabs like TSMC but don’t have Apple,AMD, MediaTek, Nvidia and even their own Intel driving huge scale efficiency and learning.
All Intel fabs got are test chips and x86 tiles and itty bitty fabs.
No amount of government subsidies can make up for the advantage AMd, Apple, Nvidia or even Google, Microsoft, Amazon have over using Intel server
@pth I like to compare LT Debt to Net Income as a pseudo-payback period to see how long it would take the company to pay off all of its debt.
Currently it's $30B LT Debt / $8.2B Net Income = 3.66 years. Not too bad, but no AAPL or NVDA of course.
Suppose Net Income drops 75% in the coming years. Then you have $30B / $2.05B ~ 15 years. That's obviously bad.
@lrg+1kqoazuf Checkout their debt and than calculated the odds they will return to high margin business, any question why the stock is so cheap
Also, employees should consider if they want to risk their career and personal fortunes on ‘riverboat’ gambling Pat G.
People might not truly appreciate the quite real existential risk this company is currently in.
Good post.
It’s what Warren Buffett would throw into the “too hard” bucket.
Too much uncertainty means a hard pass.
No better than gambling in Vegas.
Compare to TSMC which has much more reliable metrics to value since it has a pretty stable moat.
@cku Cash was $22.6B on Oct. 1, 2022, according to Value Line.
Intel used to sit on $10 billion in cash, but I think that’s under $5 billion now. That means there’s not a big draw-down opportunity here anymore.
@ivx Thank you for this. Tesla on Portland Ice, haha! Sorry, can't help with, and would not want to attempt, DCF (Discounted Cashflow analysis).
Ask Pat or Dave, then watch them deflect answering the question. 😆
The $3B is sketchy and lacked specifics. Is the $3B saving vs. 2022 or is that $3B in reduction on whatever expenses had been planned? It seemed like quite a bit of slight of hand to me.
In any case, the accounting is simple:
Gross Profit = Sales - COGS
SGA = Sales General and Admin expenses
EBIT (Earnings before Interest and Tax) = Gross Profit - SGA
Earnings (Net Income) = EBIT - intertest expense - taxes
OK out of earnings you can then pay dividends (or maybe do share buybacks).
Currently, Intel dividend coverage ratio DCR ( Net Income minus Preferred Div / Common Dividend) is around ~2.22
With earnings dropping like a rock, it is possible the DCR approaches / exceeds net income and that is no good. Eventually the dividend has to be cut in this case. Companies are loathe to cut dividends because they market will punish the stock severely.
Anyway, I hope you enjoyed this little stroll down accounting lane.
For analysts on the forum, I would just add that Intel is a very difficult stock to compute a DCF on. The problem here is that the IDM 2.0 has huge variance of spending and revenue outcome. Not to mention the core business gross margin is falling like a rock and the roadmap is slipping worse then a Tesla on Portland Ice. So, I myself throw up my hands at a valuation on Intel. If you believe IDM 2.0 will work, it could be a good buy. But if you think IDM 2.0 will fail this thing could still make a great short. Anyway, if you have opinion on how to do a DCF on this thing, please opine!
Intel CapEx is almost 20B for manufacturing for process nodes that LTD might not be able to deliver, to manufacture delayed products like Meteor Lake and plummeting margins as fabs are smaller and less efficient than the ones Apple, AMD, Nvidia use.
Then we got Ohio and Germany fabs that will cost even more but no customers of any meaningful volume to justify the spends. This is a Hail Mary play where there aren’t even any of your own team at the end zone to catch the pass.
As noted the only way to save 3B climbing to 10B is people and projects.
If you benchmark TSMC and Apple or AMD or Nvidia Intel needs to double their revenue or cut their headcount in half at a minimum
With roughly 4Billion shares outstanding, $1 a year dividend will require $4B dividend payout a year. The dividend has to go down as this is not sustainable. Unless, it grows magically and returns to huge margins.