Alphabet becomes first tech company to issue 100-year bonds in nearly three decades.
See full article...
See full article...
The longest portion of the offering, a 40-year bond, is expected to yield 0.95 percentage points over US Treasuries, down from 1.2 percentage points during initial talks, the people said.
Who's the poor schmuck who bought a hundred-year IBM bond?
This is more alarming. Bubbles tend to transition from "idiot FOMO investors lose their shirts, haha" to "everybody loses their shirts, oh shit" once huge amounts of debt and leveraged investments get involved.Big Tech companies and their suppliers are expected to invest almost $700 billion in AI infrastructure this year and are increasingly turning to the debt markets to finance the giant data centre build-out.
Flashbacks to AIG in 2008. Hopefully the bubble popping doesn't result in a bunch of life insurance companies and pension funds going bankrupt.While a century bond is “highly unusual” for tech companies, it could appeal to buyers such as life insurance companies and pension funds, which have a mandate to buy long-term assets
My only concern is that there's a shell company who gurantees the bond and not Google themselves. So if the investment turns sour, they'll collapse the shell company and wipe you out.That is actually pretty solid. 40 years is a bit more useful for someone like me looking at early retirement than 100 year bond. 30 year treasury is yielding 4.8% so we are talking about ~5.7% for 40 years.
The bad news is the declining gap between high grade corporates and treasuries likely has more to do with falling confidence in treasuries than rising confidence in corporates.
Scraping the bottom of the investor barrel, huh?
My only concern is that there's a shell company who gurantees the bond and not Google themselves. So if the investment turns sour, they'll collapse the shell company and wipe you out.
I believe Meta or Oracle is doing something like this on their AI bonds. Yes 5.7% yield from Google would be nice but only if they're the ones backing it.
Alphabet is also quite profitable. Why aren't they funding the AI themselves? I can understand Oracle and the AI startups borrowing to cover costs, because they have no cash. Alphabet is sitting on plenty of money; cash on hand for the quarter ending December 31, 2025 was $126.843B, a 32.6% increase year-over-year.IBM is still around and quite profitable. Given the sheer number of other companies that don't survive to be a century old and still profitable that likely is a poor example.
The amount they're planning to invest this year will probably exceed their profits by a fair margin, so borrowing is covering the difference.Alphabet is also quite profitable. Why aren't they funding the AI themselves?
Why use your own money when someone will give you their money for basically free. A lot of people worry about debt (and they should) but cash flow is just as if not more importantAlphabet is also quite profitable. Why aren't they funding the AI themselves? I can understand Oracle and the AI startups borrowing to cover costs, because they have no cash. Alphabet is sitting on plenty of money; cash on hand for the quarter ending December 31, 2025 was $126.843B, a 32.6% increase year-over-year.
That's assuming, of course, that the company will be around and solvent in 40 years to return the principal. That seems inevitable today for Google, but I think an investor looking at GM in 1970 would have never expected them to file bankruptcy within 40 years, either.That is actually pretty solid. 40 years is a bit more useful for someone like me looking at early retirement than 100 year bond. 30 year treasury (I assume the 30 is the reference because there is nothing longer) is yielding 4.8% so we are talking about ~5.7% for 40 years.
The bad news is the declining gap between high grade corporates and treasuries likely has more to do with falling confidence in treasuries than rising confidence in corporates.
Alphabet is also quite profitable. Why aren't they funding the AI themselves? I can understand Oracle and the AI startups borrowing to cover costs, because they have no cash. Alphabet is sitting on plenty of money; cash on hand for the quarter ending December 31, 2025 was $126.843B, a 32.6% increase year-over-year.
It's not locked up, you can sell it to someone else, like pretty much all bonds.Gotta be a little crazy to lock up your money for 40+ years in a bond paying out in USD. The dollar has depreciated significantly in the past year, and that is likely to continue as our debt load takes up an increasing percentage of GDP.
And then buying a bond to fund AI bubble spending, whew.
That's assuming, of course, that the company will be around and solvent in 40 years to return the principal. That seems inevitable today for Google, but I think an investor looking at GM in 1970 would have never expected them to file bankruptcy within 40 years, either.
With the US national debt well over 100% of GDP and looming problems like the Social Security funding shortfall it's no surprise that confidence in treasuries has declined, but the U.S. government still has an army and the ability to print money...
He doesn't care about their redemption. He wants the cash today, with full cognizance Alphabet may not survive to see the redemption anyway.Why not go all the way and sell perpetual bonds? It's highly unlikely that Sundar will be around to see the bond redemption date anyway.
Yeah, that's ok for bonds but does that even keep up with inflation?That is actually pretty solid. 40 years is a bit more useful for someone like me looking at early retirement than 100 year bond. 30 year treasury (I assume the 30 is the reference because there is nothing longer) is yielding 4.8% so we are talking about ~5.7% for 40 years. Would be nice if it was 5.7% for 40 years in Swiss Francs to hedge dollar decline but can't have everything.
The bad news is the declining gap between high grade corporates and treasuries likely has more to do with falling confidence in treasuries than rising confidence in corporates.
If it's an individual buying, it's for their kid's kids and they are probably already quite well off. If it's a pension or hedge fund... No idea. Doesn't seem smart to me.Any one one who buys this and thinks they are getting any of their money back is a fool.
Average of the last 100 years was 3.3%, so 5.7% would be well above historical inflation rates. Obviously the past does not predict the future, but betting that the next 100 years won't be as bad as the period covering the World Wars isn't ridiculous.Yeah, that's ok for bonds but does that even keep up with inflation?
Sure, but if the dollar continues to depreciate or if we go through another stretch of high inflation, then the market isn't going to cover your par value.It's not locked up, you can sell it to someone else, like pretty much all bonds.
The risks are related to whether Google will go bust before maturity, and what the currency is going to do.
They will respond in a timely fashion, 100 yearsAlphabet did not immediately respond to a request for comment.
No they do not have to be, it's really in the details (which haven't been published). They could form a company to take on the debt, assets and keep the debt off their balance sheet. This is what I was thinking Meta did (Meta is only behind 20% of their debt for AI datacenter build outs).Not sure what you mean by shell company. Alphabet is the debtor. That is google. They just felt the need to rename once they expanded beyond search. All existing "google" debt is alphabet debt.
That's not how insurance works. Premiums are based on loss exposure. Insurance--as defined in standard terms--at best breaks even on the net of premiums and losses over time. Insurance companies are expected to make their money on the float--what they earn investing the pooled premiums reserved against future losses.Who is going to profit from this? Oh, insurance companies have to invest in them. Cool. So I'll see a drop in my rate...when?
Gotta be a little crazy to lock up your money for 40+ years in a bond paying out in USD. The dollar has depreciated significantly in the past year, and that is likely to continue as our debt load takes up an increasing percentage of GDP.
And then buying a bond to fund AI bubble spending, whew.