Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/index.php?threads/intel-will-bring-in-other-capital-holders-to-help-with-fab-expansion.16578/page-2
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2021370
            [XFI] => 1050270
        )

    [wordpress] => /var/www/html
)

Intel will bring in other capital holders to help with fab expansion

Sounds good to me:

The partnership is expected to enhance Intel's balance sheet with a new pool of capital that is below its cost of equity (~8.5%) and above its cost of debt (~4.5%). We are estimating it's near the midpoint of ~6.5%. As stated at the February Investor Day, Intel has assumed that capital offsets from government support, customer participation, and private investment would offset 5-yr expected capital spending by at least 10%, with upside to 20-30%, which now appears more likely. For example, Intel expects offsets in CY22 to be >10%.

The funding reduces upfront CapEx by ~$15bn and enables cash flow break even earlier than the traditional model. Accordingly, adjusted FCF will be $15bn higher and EPS will be accretive during the investment phase, while also protecting cash/debt capacity for future investments and the dividend. In return, Intel will need to provide cash outflows to Brookfield at the ~6.5% rate over the life of the project.

The cost of ~6.5% of this financing arrangement is purely the author's own speculation. There is no way to tell how creditable it is because during the conference call, Intel CFO David Zinsner specifically stated that Intel won't reveal it.

David Zinsner did say it's between 4.4% and 8.5% during yesterday's conference call. Without knowing the full details, it is difficult to measure what the true cost of this deal is.

Replay of the conference call:
 
When your strongest engineering department is your financial engineering department.

Is it a $30 billion co-Investment deal as Intel announced or a $15 billion sale and leaseback financing arrangement in disguise?
 
Last edited:
When your strongest engineering department is your financial engineering department.
I think Intel did a case study on IBM and decided "yeah we can burn out brighter and faster than those bozos". But hey good on Brookfield for getting the opportunity to bleed this pig dry, I'm sure there will be several accounting jobs created to launder CHIPS money into offshore bank accounts. Maybe GF can even put them in touch with someone on the Cayman Islands!
 
So would you recommend they axe their dividend so that their stock can crater to 30% of its current value? Because if that happened they would not be able to borrow money for new nodes and fabs. As backwards as it may sound I think those dividends are the best way for intel to finance its expansion.
How did you calculate a 70% share price decline?
 
How did you calculate a 70% share price decline?
Nobody knows how far it would fall, so I just made up a number (I know, not super scientific but it is the best I've got). However the point still stands the dividends are majorly propping up the stock (just look at any power company's stock to see how powerful a strong dividend is). As it currently stands intel has an uncertain future, and while AMD CSPs and ARM can only realistically take so much market share, I don't think many sensible investors would touch intel with a 10ft pole until we see them meeting their roadmap and getting more IFS contracts. However if I am being paid 4.3% of what my shares are worth, then the risk of intel not bouncing back after this poor quarter is not so bad.
 
Nobody knows how far it would fall, so I just made up a number (I know, not super scientific but it is the best I've got). However the point still stands the dividends are majorly propping up the stock (just look at any power company's stock to see how powerful a strong dividend is). As it currently stands intel has an uncertain future, and while AMD CSPs and ARM can only realistically take so much market share, I don't think many sensible investors would touch intel with a 10ft pole until we see them meeting their roadmap and getting more IFS contracts. However if I am being paid 4.3% of what my shares are worth, then the risk of intel not bouncing back after this poor quarter is not so bad.

The other side of the thinking is a 10% ~ 20% price drop for Intel stock is not unthinkable. It can easily wipe out two to three years worth of Intel dividends. For institution investors it is a dangerous trap.

The question is at what cost should Intel spend on protecting its dividends and stock price?
 
Nobody knows how far it would fall, so I just made up a number (I know, not super scientific but it is the best I've got). However the point still stands the dividends are majorly propping up the stock (just look at any power company's stock to see how powerful a strong dividend is). As it currently stands intel has an uncertain future, and while AMD CSPs and ARM can only realistically take so much market share, I don't think many sensible investors would touch intel with a 10ft pole until we see them meeting their roadmap and getting more IFS contracts. However if I am being paid 4.3% of what my shares are worth, then the risk of intel not bouncing back after this poor quarter is not so bad.
I've noticed that we agree on many topics, but not on this point. Intel's defense of the dividend when they're losing money and marketshare looks out of touch with reality, not to mention seems indicative of a lack of a sense of urgency. I strongly suspect that paying a dividend while losing money and taking government subsidies will not go over well in a bipartisan way. The optics suck, especially in the light of subsidies.
 
For those who are interested in this deal between Intel and Brookfield, the link below is Intel's 8-K filing at SEC.

Source:
I like the paragraph on page 14 of the Purchase and Contribution Agreement where they're defining their pronouns.

Outside counsel for both companies must have made millions writing this stuff and setting up the LLCs.
 
I like the paragraph on page 14 of the Purchase and Contribution Agreement where they're defining their pronouns.

Outside counsel for both companies must have made millions writing this stuff and setting up the LLCs.


I think there are some strange reasons about the word "Purchase" and "Contribution". On the surface Intel (51%) and Brookfield (49%) are forming a $30 billion LLC.

Intel will contribute $15 billion worth of fab assets and Brookfield will chip in $15 billion cash to this LLC.

But Intel and Brookfield must find a way to free up that $15 billion cash and "transfer" that money into Intel's bank account. Otherwise this deal doesn't make any sense.

That's probably why the word "Purchase" comes to play. What exactly is this LLC buying?

I suspect this $30 billion Semiconductor Co-Investment Program (SCIP) is really a $15 billion Sale and Leaseback financing arrangement.

Why Intel wants to create this new terminology: SCIP?
 
I think there are some strange reasons about the word "Purchase" and "Contribution". On the surface Intel (51%) and Brookfield (49%) are forming a $30 billion LLC.

Intel will contribute $15 billion worth of fab assets and Brookfield will chip in $15 billion cash to this LLC.

But Intel and Brookfield must find a way to free up that $15 billion cash and "transfer" that money into Intel's bank account. Otherwise this deal doesn't make any sense.

That's probably why the word "Purchase" comes to play. What exactly is this LLC buying?

I suspect this $30 billion Semiconductor Co-Investment Program (SCIP) is really a $15 billion Sale and Leaseback financing arrangement.

Why Intel wants to create this new terminology: SCIP?
Intel was/is going to have capital expenditures (often called CapEx) for the Arizona fab anyway. Intel just spends as it would, but rather than using debt or cash flow for all $30B, Intel gets $15B from Brookfield to spend, and in exchange gives Brookfield a 49% interest in the fab's ownership and revenue, however revenue is defined, and I haven't figured that out. (The legalese makes me woozy.) The resulting Fab appears to be owned by a holding company called Arizona Fab LLC, which is co-owned by Intel's Arizona Fab Holdco and Brookfield's Foundry JV Holdco LLC. (If I read this legalese BS correctly.) The "purchase" is 49% of Intel's new AZ fab. It is not a sale and leaseback arrangement, it is a joint venture. As for Intel creating a new acronym, creating new acronyms is one of Intel's core competencies. :)
 
Intel was/is going to have capital expenditures (often called CapEx) for the Arizona fab anyway. Intel just spends as it would, but rather than using debt or cash flow for all $30B, Intel gets $15B from Brookfield to spend, and in exchange gives Brookfield a 49% interest in the fab's ownership and revenue, however revenue is defined, and I haven't figured that out. (The legalese makes me woozy.) The resulting Fab appears to be owned by a holding company called Arizona Fab LLC, which is co-owned by Intel's Arizona Fab Holdco and Brookfield's Foundry JV Holdco LLC. (If I read this legalese BS correctly.) The "purchase" is 49% of Intel's new AZ fab. It is not a sale and leaseback arrangement, it is a joint venture. As for Intel creating a new acronym, creating new acronyms is one of Intel's core competencies. :)

Some thoughts:

1. Intel has always told public the two new Arizona fabs cost $20 billion. This number can't be off too much because it will be tied into local, State, and Federal's subsidies and incentives. The $30 billion SCIP is much bigger than the $20 billion Intel claimed. It doesn't add up.

2. This is definitely not a joint venture. Brookfield knew it very well the $15 billion cash will have the possibility to flow into other Intel's businesses and they might not have any control or visibility at all.

To me it's more like sale and leaseback in disguise.

3. Some mentioned about Brookfield will share 49% of the Arizona fabs' revenue. That's impossible.

Revenue is not profit. There are business partners sharing the profit but I haven't heard they can share the revenue. This makes no sense to IRS either.

Additionally from yesterday's conference call Intel CFO stated that Brookfield is promised to receive annual return between 4.4% and 8.5%. This is not the revenue sharing model at all and this is not how a typical joint venture works.


5. The Intel 8-K filing may give some additional information but IMO don't expect too much from it. There are crucial information that I don't think Intel would like to let people know at this point.

6. It's a strange deal. I think many analysts haven't gotten time to study it yet.

For example:

"Concurrent with transaction signing, Brookfield Infrastructure secured binding non-recourse financing to fund a significant portion of its capital investment in this arrangement with Intel. BIP’s equity portion for this investment will be phased over the duration of the construction period and is anticipated to be $500 to $750 million in aggregate. Funding is expected to be sourced primarily from retained operating cash flow and proceeds from our capital recycling program, with the majority funded closer to facility commissioning."

I can't understand the difference between the announced Brookfield $15 billion cash investment and this $500 to $750 million investment.

 
Last edited:
Some thoughts:

1. Intel has always told public the two new Arizona fabs cost $20 billion. This number can't be off too much because it will be tied into local, State, and Federal's subsidies and incentives. The $30 billion SCIP is much bigger than the $20 billion Intel claimed. It doesn't add up.
Probably two different figures with separate context. The agreement actually reads that the joint investment "shall not exceed $28,982,302,000". See page 5 of the Arizona Fab LLC.
2. This is definitely not a joint venture. Brookfield knew it very well the $15 billion cash will have the possibility to flow into other Intel's businesses and they might not have any control or visibility at all.
In reading over the 8K, and how the JV is structured as a separate company with its own BoD, I don't see how this can legally happen. (Of course, I'm not a corporate law attorney.) All of the money passes through the JV corporation and only the JV. There's no mention of how employees are handled that I've seen. I was an employee of a JV once, and while I had a separate JV badge, I remained an employee of my parent company. I suspect Intel will do the same.
To me it's more like sale and leaseback in disguise.
Getting a disguised operation past the SEC seems very unlikely. And you might end up in an orange suit, especially for an entity of this value.
3. Some mentioned about Brookfield will share 49% of the Arizona fabs' revenue. That's impossible.
I've seen that too, and I don't understand how that could work either. I suspect there's either a fixed return to Brookfield or some percentage of net margin, which can be calculated using Intel's internal transfer prices and the balance sheet of the JV. I think the net margin alternative would be subject to "interpretation" by Intel, so I would bet Brookfield is getting some agreed upon return.
Revenue is not profit. There are business partners sharing the profit but I haven't heard they can share the revenue. This makes no sense to IRS either.
Agreed.
Additionally from yesterday's conference call Intel CFO stated that Brookfield is promised to receive annual return between 4.4% and 8.5%. This is not the revenue sharing model at all and this is not how a typical joint venture works.
I also don't understand how a revenue sharing model can work.
5. The Intel 8-K filing may give some additional information but IMO don't expect too much from it. There are crucial information that I don't think Intel would like to let people know at this point.
I've scanned most of it, and the answers we're looking for don't seem to be in there, but I don't feel qualified to interpret a lot of it.
6. It's a strange deal. I think many analysts haven't gotten time to study it yet.

For example:

"Concurrent with transaction signing, Brookfield Infrastructure secured binding non-recourse financing to fund a significant portion of its capital investment in this arrangement with Intel. BIP’s equity portion for this investment will be phased over the duration of the construction period and is anticipated to be $500 to $750 million in aggregate. Funding is expected to be sourced primarily from retained operating cash flow and proceeds from our capital recycling program, with the majority funded closer to facility commissioning."

I can't understand the difference between the announced Brookfield $15 billion cash investment and this $500 to $750 million investment.


I'm not knowledgeable enough in finance matters to know what this means. It looks like Brookfield is taking a loan for the initial contribution to the JV. I didn't see these numbers in a scan of the 8K documents.
 
Last edited:
Motley Fool write up on BIP:


"Its partnership with Intel is quite different. Brookfield and its institutional partners are investing up to $15 billion for a 49% stake in two semiconductor fabrication facilities the tech giant is building in Arizona. In exchange, it will get half the revenue produced from these facilities."

So my comments about the nightmare of secured debt/leaseback finance were hasty and incorrect, appears this is a straight equity investment in a JV.

I'm less hostile to this as an equity JV, but, Intel's last JV with Micron wasn't a winner. BIP is not bringing expertise, just cash. Intel isn't experienced in foundry, but has fab operations expertise at least. The JV will begin life with huge leverage, is reliant on INTC technology transfers, and so has a hard life ahead.

On the plus side, this will be a 49% pure play Foundry with real firewalls between INTC and the JV.
 
Motley Fool write up on BIP:


"Its partnership with Intel is quite different. Brookfield and its institutional partners are investing up to $15 billion for a 49% stake in two semiconductor fabrication facilities the tech giant is building in Arizona. In exchange, it will get half the revenue produced from these facilities."

So my comments about the nightmare of secured debt/leaseback finance were hasty and incorrect, appears this is a straight equity investment in a JV.

I'm less hostile to this as an equity JV, but, Intel's last JV with Micron wasn't a winner. BIP is not bringing expertise, just cash. Intel isn't experienced in foundry, but has fab operations expertise at least. The JV will begin life with huge leverage, is reliant on INTC technology transfers, and so has a hard life ahead.

On the plus side, this will be a 49% pure play Foundry with real firewalls between INTC and the JV.

All members in a LLC can share profits, not revenue. Sharing revenue is almost meaningless. I'm not sure what this Fool's article really meant. In yesterday's conference call, Intel CFO David Zinsner did say Brookfield is promised to receive return between 4.4% and 8.5%. He would not reveal the actual number though.


"Its partnership with Intel is quite different. Brookfield and its institutional partners are investing up to $15 billion for a 49% stake in two semiconductor fabrication facilities the tech giant is building in Arizona. In exchange, it will get half the revenue produced from these facilities."
 
Motley Fool write up on BIP:


"Its partnership with Intel is quite different. Brookfield and its institutional partners are investing up to $15 billion for a 49% stake in two semiconductor fabrication facilities the tech giant is building in Arizona. In exchange, it will get half the revenue produced from these facilities."

So my comments about the nightmare of secured debt/leaseback finance were hasty and incorrect, appears this is a straight equity investment in a JV.

I'm less hostile to this as an equity JV, but, Intel's last JV with Micron wasn't a winner. BIP is not bringing expertise, just cash. Intel isn't experienced in foundry, but has fab operations expertise at least. The JV will begin life with huge leverage, is reliant on INTC technology transfers, and so has a hard life ahead.

On the plus side, this will be a 49% pure play Foundry with real firewalls between INTC and the JV.
I thought Intel said that IFS wafers would not be processed in there own special fab (ie. no foundry/IDM fabs just fabs)?
 
Back
Top