Meta Platforms priced $25 billion of investment-grade bonds on Friday May 1 across as many as six tranches, anchored by a thirty-year piece at 165 basis points over Treasuries and a forty-year long bond that closed inside the company's existing curve. The bookrunners — JPMorgan, Morgan Stanley, Goldman Sachs and Bank of America — built an order book of roughly $96 billion before final allocations, which made the issue four-times oversubscribed in the standard syndicate vocabulary. [1] The equity, in the same five trading sessions, lost 9.5 percent of its market capitalization. The bond market and the equity market priced the same balance sheet in opposite directions.

The paper registered the print yesterday and noted that the buyback line had gone silent — Meta's first-quarter repurchase line was, by the company's own 10-Q, not executed. Today the silence is the artifact of a week that produced four discrete corporate-finance answers across the cohort: Apple at the $100 billion buyback, Berkshire at the $397.4 billion cash hoard, Amazon at $44.2 billion of first-quarter capital expenditure, and Meta at $25 billion of new senior unsecured debt. The four firms used to live inside the same valuation paragraph. They are now in four different paragraphs.

The capex envelope is the part that matters. Meta's full-year 2026 guide, raised on the April 30 call, runs $125 to $145 billion. The midpoint, $135 billion, is more than four times the company's full-year 2024 capex of $32 billion. Mark Zuckerberg told analysts on the call that he had "no precise plan" for how each AI product would generate the revenue that justifies the build, but that the build had to happen anyway because the alternative was being late to a market the company believes is determinative. [2] The phrase "no precise plan" is the kind of remark a chief executive does not deliver to an investor base he expects to clap. The equity declined the same evening, took two more percentage points off Friday, and finished the week down the 9.5.

The bond market read the same script and reached a different verdict. With $96 billion of orders against a $25 billion deal, the buyer base saw a balance sheet that, even with $135 billion of capex, prints something like $200 billion of operating cash flow this year and carries an investment-grade balance with a payment-coverage ratio that supports significantly more debt than the company has chosen to issue. The thirty-year tranche cleared at a spread inside Microsoft's, which is, in the bond-trader vernacular, a tell. The bond market thinks Meta's leverage is fine; the equity market thinks the return on incremental capital is bad. Both markets can be right. They are pricing different things. [3]

The buyback omission is what holds the two views together. Meta executed roughly $30 billion of repurchases in 2025; the first-quarter 2026 print of zero, with the second-quarter buyback authority still untouched as of the proxy date, is a corporate-finance signal that management has reweighted toward the capex line and away from the equity-return line. The signal is unspoken — Meta has not formally suspended the buyback program — but it is, in the language balance sheets speak, the news. The company is going to spend the cash on data centers and not on its own shares. The bond market financed the choice. The equity market priced it. [4]

Inside the cohort, the contrast with Apple's $100 billion buyback printed forty-eight hours earlier is the part that does not require commentary. Apple is returning capital to shareholders. Meta is borrowing capital to deploy on infrastructure. Both companies have ample free cash flow. Both companies have similar revenue growth profiles. Both companies are run by founders who control voting equity. They have made opposite choices. The market has given Apple a three-percent reward and given Meta a nine-and-a-half-percent rebuke.

The next test is the proxy. Meta's annual meeting falls in late May; the buyback authorization vote will be a routine line on the ballot, but the proxy advisory firms — ISS and Glass Lewis — have begun, in their preliminary memos, to flag the capex-versus-buyback question as a discussion topic. The advisors do not, in the ordinary course, recommend against ratification. They recommend that boards explain themselves. The May proxy is the place that explanation will live. [5]

Until then the financing surface and the equity surface are the two ledgers showing the same balance sheet. Both are correct. Neither agrees with the other.

-- THEO KAPLAN, San Francisco