MUSE

Less than a week after earnings, META moved to execute its 'cash grab' plan. It announced two concurrent underwritten offerings pre-market totaling approx. $35bn:

$22.5bn Class A common: Issued via forward sale agreements with Morgan Stanley and Goldman Sachs, with no upfront cash to the company. Settlement will be timed at META's discretion by Nov 19, 2027.$12.5bn mandatory convertible preferred: 250mn depositary shares under ticker METAM, mandatorily converting into common around Aug 15, 2029 at a floating rate. The dividend rate will be set at pricing.

Each tranche also includes a greenshoe of $3.375bn and $1.875bn, respectively. NVIDIA, AMD and Broadcom intend to subscribe up to $5bn of common, on a non-binding basis. After Class C begins trading (Q1 2027, post the Jan 19 1-for-1 stock dividend), the company plans an approx. $60bn Class C ATM program (shares sold into the market over time, not an overnight secondary; shares are freely tradable upon sale), with no banks yet mandated.

Dolphin view: 'Alternative pricing' in the money-raising phase

In our earnings take, we noted the Class C setup is essentially pre-arranging a financing channel for future spend. Intermittent issuance should prevent excessive bullish froth.

Today puts numbers to that placeholder, but near-term real dilution is modest, easing some nerves. However, it does not remove the valuation overhang from ongoing financing expectations.

(1) Size: Assuming ~$650 per share, the full $95bn implies ~146mn shares, ~5.7% of the current 2.556bn shares outstanding. Including greenshoes, roughly ~6% (est.).

(2) Pace: This will not hit all at once. The $22.5bn Class A forwards settle in tranches over 13 months, the $60bn Class C ATM would be dribbled out through 2027, and the $12.5bn preferred does not convert until 2029.

(3) Overhang: Versus historical $20–30bn annual buybacks, from 2027 there will be sellers with $60bn of stock to place. It is not a dump; spread out, daily flow impact at ~2.5% of volume is not large, but sentiment could still be capped.

How big is META's funding gap?

Quick math: we model 2027/2028 revenue at $312bn/$368.6bn (+22%/+18% YoY). At a 55% OCF margin, that implies OCF of $171.6bn and $202.7bn.

Capex in 2027/2028 could reach $220bn/$250bn, implying FCF gaps of roughly -$50bn and -$50bn, or -$100bn combined.

META currently holds $82.5bn cash, with no short-term debt and only long-term bonds outstanding. On a conservative view, near-term funding needs are close to $20bn.

In other words, the announced $35bn A + the planned $60bn C, $95bn in total, would be sufficient to cover the next two years' gap.

One caveat: if 'Muse' blows out demand (as we also flagged in our earnings take), the compute deployed over the next two years may prove insufficient. Capex could be lifted above the assumptions noted.

Valuation update: If the full $95bn is issued and converted, dilution would be ~6%. We keep our valuation framework from the earnings review: add near the bearish case, take profits near the bullish case.

Total equity value unchanged; per-share adjusted for dilution, our bear/base/bull cases are $508/$705/$790, respectively. Given the resilient ads base with relatively low AI cannibalization risk, and Muse potentially reinforcing the ecosystem entry point, we would look to add on pullbacks.

Risk disclosure and statements:海豚研究免责声明及一般披露