$GOOG is raising $80B of equity while trading at a 3-4% earnings yield.
If Google can reinvest that $80B at 20% ROIC, that is $16B of incremental earnings power.
At a 25x multiple, that is $400B of value to shareholders.
@YaelC_03 With a projected 2027 run-rate of $471M per month and an inability to manufacture 800G modules fast enough for existing hyperscaler orders, is the real risk a foreign price war, or simply executing on an overflowing backlog? $AAOI
If Credo’s AEC/ALC tech physically caps out at 7m and 30m for short-range in-rack links, how does that disrupt $AAOI $324M hyperscaler backlog?
Isn't 100% of that backlog explicitly for long-reach single-mode and 1.6T connections spanning across massive data halls? When a cluster needs to route data well beyond 30 meters, who else fills that order today?
If $GOOGL earns 20% ROIC and trades at 33x earnings, its equity cost is roughly a 3% earnings yield.
They can raise capital at a 3% implied cost and reinvest near 20%, that spread creates value.
$1 of capital earning 20% = $0.20 of earnings.
At 33x, that $0.20 is worth $6.60 of market value.
If $GOOGL earns 20% ROIC and trades at 33x earnings, its equity cost is roughly a 3% earnings yield.
They can raise capital at a 3% implied cost and reinvest near 20%, that spread creates value.
$1 of capital earning 20% = $0.20 of earnings.
At 33x, that $0.20 is worth $6.60 of market value.
$SNDK and $AAOI are the same type of setup.
Demand inflects because AI infrastructure needs more capacity.
Supply is constrained.
Revenue ramps faster than investors expected.
And if the company already has the assets / manufacturing base in place, incremental revenue can fall through at very high margins.
You are betting on a supply-constrained bottleneck where every incremental dollar of revenue could produce an abnormal ROIIC.
Thank you, @TimSeo282482
$AAOI updated model from $378M/month to $471M/month from the Q1 call.
$5.65B annualized revenue.
At a 20% EBIT margin, $1.13B of EBIT run-rate.
10x EBIT at $11.7B EV.
$AAOI filed a $600M ATM.
At $150/share, that 5% dilution.
ROIC Math:
$5.7B annualized revenue.
15% EBIT margins, $850M EBIT.
If the $600M raise is the incremental capital needed to unlock that scale, the ROIIC is absurd:
$850M EBIT / $600M capital = 140% ROIIC.
$AAOI filed a $600M ATM.
At $150/share, that 5% dilution.
ROIC Math:
$5.7B annualized revenue.
15% EBIT margins, $850M EBIT.
If the $600M raise is the incremental capital needed to unlock that scale, the ROIIC is absurd:
$850M EBIT / $600M capital = 140% ROIIC.
At $AAOI $378M/month mid-2027 revenue framework = $4.5B annualized revenue.
20% EBIT margin, $900M EBIT run-rate.
Against $12.5B EV, you are paying 2027 14x EBIT while the company is growing revenues 30%.
$PATH may still work.
The issue is not demand.
The issue is implementation.
There is clearly demand for automation. But if the customer has to redesign workflows, integrate systems, train teams, and fight internal adoption, the sales cycle and value capture are harder.
That is a different setup than the current opportunities where demand is overwhelming and supply is the actual constraint.
In those cases, the customer already wants the product. They just cannot get enough of it.
I would rather own the bottleneck where demand is obvious, implementation is simple, and supply is scarce.
@Rahmteenn UBS projects $400B of FCF from 2027 to 2029 for $MU.
Call it $200B from 2027 to 2029.
Then assume from 2030 forward, only generates 1/10th of that per year.
$20B of FCF per year.
At $20B per year, MU would generate another $800B over 40 years.