To The Moon

Same company’s massive debt.

Some see pure crisis.

Others see pure opportunity.

That’s $Oracle(ORCL)$   right now.

Quick Translation + Reality Check (with the absolute latest numbers after Sept 10 earnings)

Oracle reported Q1 FY2027 (ended Aug 31, 2026) on Sept 10 after the close:

Revenue: $19.3 billion (+30% YoY) — beat estimates

Cloud Infrastructure (OCI): $7.4 billion (+121%) — absolute monster growth

Total Cloud: +62%

Remaining Performance Obligations (RPO / backlog): $664 billion (was $638B at end of FY2026) — still insane

Non-GAAP EPS: $1.92 (beat $1.74)

Free cash flow burn: only –$5.4 billion (vs expected –$9.56B) — much better than feared

Raised full-year FY2027 adjusted EPS guidance to $8.10

They also delivered 850 MW of new data-center capacity and over 300,000 GPUs to customers since last quarter.

The Big Debates — Updated with Latest Facts

1. The Debt & Cash Burn Scare

Yes, free cash flow is still deeply negative because of huge data-center and GPU spending. S&P previously cut them to BBB (lowest investment grade). Some voices (like Finance Lancelot on X) still call Oracle the company most at risk of being crushed by AI infrastructure costs.

But the latest quarter showed the burn is improving and a big chunk of new contracts come with customer prepayments or customer-supplied GPUs — reducing Oracle’s own cash outlay.

2. Capacity Already Locked at Old Prices

This remains the awkward part. ~98% of AI capacity was previously contracted. Industry GPU rental prices are rising hard right now, but Oracle can’t fully enjoy the new higher prices on most of its existing capacity. New contracts and any pricing flexibility will matter a lot going forward.

3. Gross Margin Pressure

Expected. New capacity comes online with costs (power, people, depreciation) hitting before full revenue ramps. This is classic expansion pain, not necessarily proof the business is unprofitable long-term. Watch whether older capacity is already generating solid returns.

4. Fundraising & Dilution

Oracle has a formal $40 billion fundraising target for the current fiscal year (debt + equity).

They already completed a $20 billion At-The-Market equity raise in Q1, issuing roughly 141–158 million new shares (~5.5% dilution). More capital is still needed.

5. Larry Ellison’s Sudden Reversal (Sept 12 news)

Just one day after disclosure of a Rule 10b5-1 plan to sell up to 50 million shares (~$7.5 billion at ~$150), Ellison cancelled the entire plan.

Zero shares were sold. He has no current plans to sell any of his 40%+ stake.

People close to him say he views the stock as undervalued. This removes a huge potential supply overhang and is a strong alignment signal with public shareholders while the company digests heavy infrastructure spend.

6. Wall Street Split

Citi: Still bullish, $330 target (based on 2030 earnings power). Sees the prior sell-off as an opportunity.

Others (e.g. Morgan Stanley style): Acknowledge short-term rebound potential but stay neutral until multiple quarters of proof.

Delivery on time + improving cash generation = stock can re-rate fast even if debt is still high. Stock prices the future, not just the current balance sheet.

Bottom Line Prospects

Bull case getting stronger evidence:

Massive backlog converting, triple-digit infrastructure growth, better-than-feared cash burn, raised guidance, and founder skin fully in the game.

Still real risks:

Ongoing heavy capex, potential margin compression, dilution from further raises, and the lag between spending and full profitability.

The game remains the same: Can revenue and collections catch up to the spending before the funding pressure becomes critical?

Investment decision framework still applies perfectly after these results:

Delivery & guidance continuity — strong so far

Funding not spinning out of control — improving signs + $20B already raised

Don’t chase pure multiple expansion without real earnings power behind it

Ellison cancelling the $7.5B sale is the loudest “I’m not running away” signal possible.

AI compute demand is completely unhinged. Oracle just resold 100% of its renewal GPU capacity at a 20% premium—and get this: the majority of that hardware was four years old or older.

​Pour one out for Michael Burry’s rapid GPU depreciation thesis. The market isn't dumping old silicon; it's practically fist-fighting over it.

​Meanwhile, Dario, Sam, and Elon are out here whispering about "slowing down" frontier AI. Pull the other one. There is zero chance Anthropic or OpenAI pumps the brakes and politely waits for the competition to catch up. That talk isn't safety—it’s textbook regulatory capture.

​Sure, the pearl-clutching might trigger some short-term panic and red candles, but let’s be real: that’s just a discount code for the smart money. Because the alternative to the US slowing down isn't a utopia; it's handing the keys to China while they quietly leapfrog the entire ecosystem.

​Compute is king, the frontier is wide open, and nobody is stopping.

Would you rather bet before the next proof points… or wait until cash flow turns and the story is fully de-risked?