Aksi Jual Saham Oracle Menunjukkan Backlog AI Butuh Bukti Arus Kas

Aksi jual saham Oracle setelah hasil fiskal Q4 menyoroti perubahan dalam cara pasar menilai perdagangan infrastruktur AI. Perusahaan ini membukukan angka berita utama yang lebih kuat, namun reaksi terhadap harga saham menunjukkan bahwa investor tidak lagi membayar hanya untuk simpanan AI saja. Mereka sedang menguji apakah simpanan tersebut dapat menghasilkan pendapatan dengan tingkat pengembalian yang tinggi, arus kas bebas yang positif, dan kekuatan neraca dengan cukup cepat untuk membenarkan skala pengeluaran yang diperlukan saat ini.

The operating picture was not weak. Oracle posted fiscal Q4 FY2026 revenue of $19.2 billion, up 21% from a year earlier, while non-GAAP earnings per share reached $2.11 versus $1.70 a year earlier and above roughly $1.96 expected by many analysts. GAAP net income available to common shareholders rose to about $4.223 billion from $3.427 billion. For a mature enterprise software name, those figures would normally support a constructive reaction.

The problem is that Oracle is increasingly being judged less like a high-margin software franchise and more like a capital-heavy AI infrastructure builder. Oracle Cloud Infrastructure revenue reached $5.8 billion, up 93%, and remaining performance obligations climbed to $638 billion, up $85 billion sequentially. Those numbers point to real demand, but they also raise a tougher question: how much cash, equipment, power, and financing must be committed before the demand becomes visible profit.

Itulah sebabnya garis pengeluaran mendominasi respon pasar. Perluasan pusat data AI yang dilakukan Oracle diperkirakan memerlukan belanja modal sekitar $70 miliar pada tahun fiskal ini, dan pelanggan diperkirakan akan berkontribusi sekitar $20 miliar hingga $25 miliar secara langsung. Perbandingannya sangat tajam: belanja modal pada tahun fiskal 2026 adalah sekitar $55,7 miliar, sudah jauh di atas sekitar $21 miliar hingga $22 miliar pada tahun fiskal 2025. Peluang pertumbuhannya besar, namun beban di muka juga besar.

Free cash flow is the pressure point traders should not ignore. Oracle generated $32.0 billion of operating cash flow in fiscal 2026, but free cash flow was negative by $23.7 billion after the heavy investment cycle. The company also expects fiscal 2027 financing of about $40 billion through debt and equity, including a previously announced $20 billion at-the-market equity issuance. That mix makes dilution, leverage, and funding cost part of the equity story.

The $638 billion remaining performance obligation figure is important, but it should not be treated as immediate revenue. Backlog differs by timing, margin, customer concentration, infrastructure readiness, and contract structure. A large AI customer can validate demand and still increase concentration risk. For traders, the cleanest interpretation is that Oracle has demand visibility, but the equity market wants better proof that this demand can convert into cash flow at attractive returns.

Management also kept fiscal 2027 total revenue guidance at $90 billion. That figure supports a long runway for growth, yet it did not resolve the central tension. If AI capacity ramps smoothly, Oracle can defend the idea that current spending is a necessary entry fee for a larger cloud business. If timelines stretch, power access tightens, hardware costs rise, or utilization disappoints, the same spending profile can pressure margins and shareholder returns.

The financing details also matter because AI infrastructure is not a low-commitment growth option. Debt can fund capacity without immediate dilution, but it raises sensitivity to rates and credit spreads. Equity issuance protects the balance sheet, but it spreads future upside across more shares. Customer prepayments or contributed hardware can ease the cash burden, yet traders still need to evaluate whether the economics remain attractive after construction cost, depreciation, energy contracts, and utilization risk.

Hal ini membuat penilaian menjadi lebih bersyarat dibandingkan pada fase awal perdagangan AI. Sebuah perusahaan dapat memperoleh permintaan cloud yang meningkat dan masih berhak mendapatkan keuntungan yang lebih rendah jika pasar percaya bahwa pertumbuhan tambahan memerlukan terlalu banyak modal. Sebaliknya, perusahaan yang sama dapat memperoleh kembali status premium jika terbukti bahwa skala tersebut membangun leverage operasi. Oleh karena itu, beberapa pembaruan Oracle berikutnya harus menunjukkan tidak hanya kemenangan kontrak, namun juga jalur dari permintaan yang ditandatangani ke margin dan perolehan uang tunai.

Oleh karena itu, arah pasar masuk akal meskipun pendapatan utama solid. Penurunan tajam setelah pendapatan mengalahkan biasanya berarti ekspektasi sudah tinggi dan pasar menemukan kelemahan baru dalam pengaturan ke depan. Dalam kasus Oracle, kelemahan tersebut bukanlah permintaan. Hal ini merupakan kesenjangan arus kas antara penandatanganan komitmen AI cloud dan pembuktian bahwa belanja infrastruktur dapat menghasilkan keuntungan jangka panjang.

For broader technology and index traders, Oracle matters because it sits in the same AI infrastructure chain that has driven enthusiasm across large-cap technology. A funding scare in one major cloud builder can spill into sentiment around data-center suppliers, chip demand, power infrastructure, and AI-linked equity multiples. The read-through is not automatically bearish for the sector, but it raises the hurdle for companies that need heavy capex to chase AI growth.

This is especially relevant for NAS100 exposure because the index has rewarded companies that can connect AI narratives to near-term earnings power. The market reaction suggests a more selective phase: investors may still support AI leaders, but they are likely to separate asset-light beneficiaries from companies absorbing heavy infrastructure bills. That distinction can influence factor leadership, with software margins, chip demand visibility, and data-center balance-sheet risk all competing for attention.

The near-term setup is therefore two-sided. Bulls can point to 93% OCI growth, a $638 billion RPO base, and a large enterprise customer pipeline. Bears can point to negative free cash flow, rising capex, equity issuance, and the risk that AI infrastructure returns take longer than investors expected. MC Markets views this as a quality-of-growth test: Oracle has demand, but the stock needs evidence that the next phase of AI spending can improve cash conversion rather than only expand scale.

A constructive turn would likely require clearer signs of capex discipline, stronger free-cash-flow visibility, and confidence that high-value AI contracts will convert into revenue without pressuring margins. A bearish continuation would be more likely if financing needs rise again, if guidance depends on ever-larger infrastructure commitments, or if the market starts discounting RPO because of customer concentration and execution risk. Until that balance improves, rallies may face valuation resistance even when headline cloud growth remains strong.