What was bought, and when
Broadcom ($AVGO) completed its acquisition of VMware ($VMW) on 22 November 2023 and, the same day, "borrowed the full $28,390,625,000 available under the Credit Agreement to fund the Cash Consideration." That is Broadcom's own 8-K. The VMware deal was paid for with bank loans. No bonds.
The bonds came later. On 8 July 2024 Broadcom priced $5.0 billion of senior notes in three parts, settling 12 July 2024. The prospectus says where the money went: to "prepay a portion of the term A-2 loans under our term loan credit agreement." The VMware debt, termed out.
This page follows the longest piece: the 5.150% senior notes due 15 November 2031, CUSIP 11135FBY6, $1,500,000,000, sold at 99.838 to yield 5.178%, 95 basis points over the seven-year Treasury.
Against it, $10,000 of $AVGO stock at the same day's close. Income is cash on both sides, no reinvestment on either.
One bounded claim, because this series does not make unbounded ones. Between the VMware close on 22 November 2023 and this deal, Broadcom Inc. filed no prospectus supplement or free writing prospectus for a notes offering. EDGAR's filing index for CIK 0001730168 is the source, the bound is those two dates, and it covers SEC-registered public offerings only. It says nothing about private placements, which would not appear there.
The split that would have wrecked this page
At 4:30 p.m. Eastern Time on 12 July 2024, the same afternoon the bonds settled, Broadcom's ten-for-one forward stock split took effect. The 8-K says so in those words.
So the $AVGO close that day, as it actually printed, was about $1,700.67. Split-adjusted it is $170.067, and that is the number this page uses, because the shares a buyer held that afternoon became ten shares the following week.
Two checks that the adjustment is right:
- The daily series runs 170.595 on 11 July, 170.067 on 12 July, 171.42 on 15 July. A series that had not been adjusted would show a ninety percent gap across that weekend. There is no gap.
- Broadcom's dividend went $5.25 a share with an ex-date of 24 June 2024 to $0.53 a share with an ex-date of 19 September 2024. Broadcom did not cut its dividend by ninety percent. It split the stock ten ways.
Use the unadjusted $1,700.67 and this page would report that $AVGO fell about 79% over a stretch in which it did the opposite. That is what a split error looks like from the inside: every individual number true, the answer completely wrong.
Both legs marked on one date
Both legs are marked Wednesday 30 September 2026.
The bond comes off a FINRA TRACE print read first-hand: $300,000 par at 97.161, yield 5.798%, at 16:00:23 Eastern. The stock is that day's close, $351.19. Twenty-three seconds apart.
That is not housekeeping. The bond's prints the next morning are lower, 96.820 at 09:13:55 on 1 October, and the equity market had not opened. Marking the two legs on different days would produce a comparison in which every number was accurate and the conclusion was not.
The result
| Bought 12 Jul 2024 | Bond | Stock |
|---|---|---|
| Price then | 99.838 | $170.07 |
| Price on 30 Sep 2026 | 97.161 | $351.19 |
| Price change | -2.7% | +106.5% |
| Income collected | $1,143 | $323 |
| $10,000 becomes | about $10,875 | about $20,973 |
| Total return | about +8.8% | about +109.7% |
| Bought 31 Dec 2025 | Bond | Stock |
|---|---|---|
| Price then | 103.930 | $346.10 |
| Price change | -6.5% | +1.5% |
| Income collected | $372 | $56 |
| $10,000 becomes | about $9,720 | about $10,203 |
| Total return | about -2.8% | about +2.0% |
The stock won both entries. On the 2024 entry the dollars gained differ by about twelve times.
Where each return actually came from
This is the part worth keeping.
On the bond, income was the whole return and more. The holder collected $950 of coupons across four payment dates, starting with a short first stub on 15 November 2024, plus $193 of interest accrued and owed at the mark. That is $1,143 on $10,000 put in. The price took 2.7 points back out. Income was more than 100% of the return.
On the stock, income was almost nothing. Nine dividends at $0.53, $0.59 and $0.65 came to $5.49 a share, about three points of a 110-point return. The last of them paid on 30 September 2026, the mark date itself, so it counts. The price was everything.
That is the inverse of what this series found at Oracle, where the two legs fell by nearly the same amount and income was the entire difference between them. Same question, different issuer, different window, different answer.
Rates or the business
The bond's yield went from 5.178% at issue to 5.798% at the mark, up 62 basis points. The obvious reading is that the market got more worried about Broadcom. The data says otherwise, and it is worth doing properly.
Take the Treasury that matched the bond's remaining life on each date, from the Treasury's own daily par yield curve:
| 12 Jul 2024 | 30 Sep 2026 | |
|---|---|---|
| Bond's remaining life | 7.3 years | 5.1 years |
| Bond yield | 5.178% | 5.798% |
| Comparable Treasury, interpolated | about 4.14% | about 5.10% |
| Spread | about 104bp | about 70bp |
The matched Treasury rose about 96 basis points. The bond's own yield rose 62. So the credit spread did not widen. It came in by roughly 34 basis points.
That is not evidence that the market thinks Broadcom is safer, and this page will not say it is. At least two things narrow a spread with nobody changing their view of the issuer:
- The bond got shorter. It carried 7.3 years of credit risk at issue and carries 5.1 now. Spreads normally narrow as a bond rolls down its own curve. Some unknown share of the 34 basis points is time passing, nothing more.
- New issues price with a concession. The deal came 95 basis points over the seven-year Treasury in order to place $5 billion in an afternoon. That concession is not meant to survive into the secondary market.
A third possibility, that investment-grade spreads tightened across the board, is plausible and unmeasured here, so it stays an open item rather than an explanation.
What can be said plainly: every point of this bond's price loss traces to the Treasury curve, not to Broadcom. The spread leg worked in the holder's favour. The stock's 106% is the business, and nothing in the bond's price says anything about that either way.
The second window does not flatter the bond, and is worth stating for that reason. From 31 December 2025 the comparable Treasury went from about 3.82% to about 5.10%, and the spread went the other way, from about 54bp to about 70bp. Over those nine months both legs moved against the bondholder.
The debt, from the filings
Broadcom's 10-Q for the quarter ended 2 August 2026 carries total debt of $59,419 million, down from $65,136 million at the start of the fiscal year. Over nine months Broadcom issued $4,500 million of notes and repaid or repurchased $10,541 million. The 2031 notes sit at their full $1,500 million, at an effective rate Broadcom books as 5.30%.
The same filing carries something larger and newer. In that quarter Broadcom "arranged for a financial partner to take on certain agreements to purchase AI racks based on custom AI accelerators designed by us and the related lease agreements with a customer." Broadcom then "entered into a backstop agreement with the financial partner for the customer's lease obligations over the 5-year lease terms." Its own words for the size: "Our maximum potential liability under the Backstop upon the deployment of all AI racks, on an undiscounted basis, was approximately $29 billion." Separately, that customer "may, under certain circumstances and if needed, issue to us convertible promissory notes up to an aggregate principal amount of $42 billion." As of 2 August 2026, none had been issued.
Those are quotations from Broadcom's filing, not from reporting about it. They are not borrowings, they are not on the balance sheet, and this page is not going to claim they are priced into a bond trading 70 basis points over Treasuries. They are simply what a reader of Broadcom's debt should know exists.
What this does not say
It does not say stocks beat bonds. On the Nike page in this series the bond finished roughly three times ahead of the stock, and that was the same exercise with the same rules. It does not say anything about what either security does next, and there are no targets or positions here.
It rests on two entry dates and one mark date, all of them stated. A buyer from a different month gets a different answer. That is the whole reason this series keeps running.