I opened Oracle's earnings report early this morning and stopped scrolling for quite a while. Both revenue and profit exceeded expectations, but an unfamiliar line at the bottom of the cash flow statement caught my eye. My conclusion is Hold, with a 12-month baseline target of $158; let me explain, step by step, why this complete earnings victory did not lead to an upward revision of the target price.
The analysis reference date is September 11, 2026, and the stock prices are based on the close of the U.S. regular trading session on September 10. The KRW amounts in the text have been calculated using a standard exchange rate of 1,338 KRW per dollar.
- Oracle (ORCL) Q1 Earnings: Simultaneous Revenue and EPS Surprise
- Why You Should Not Trust the 30.9 Trillion Won Operating Cash Flow at Face Value
- What kind of company is Oracle now?
- Oracle (ORCL) Valuation: Is 17.6x Cheap?
- Oracle (ORCL) Target Price Scenarios and Expected Returns
- Cut in half in a year, what happened in the meantime?
- Oracle (ORCL) Risks and Bear Triggers
- Personally, this is a point that bothers me.
- Oracle (ORCL) Investment Opinion Conclusion
- Frequently Asked Questions
- Reference materials
Oracle (ORCL) Q1 Earnings: Simultaneous Revenue and EPS Surprise
First, let me summarize the three-line conclusion. Oracle reported revenue of $19.345 billion (approx. 25.9 trillion won) for the first quarter of fiscal year 2027 (June–August 2026), growing 301 TP3T from a year ago, and non-GAAP earnings per share of $1.92 (approx. 2,569 won), exceeding market expectations of $1.74 by 111 TP3T.
However, capital expenditures in the same quarter amounted to $28.499 billion (approximately 38.1 trillion won), which was more than quarterly revenue, and free cash flow was recorded at minus $5.396 billion (approximately -7.2 trillion won).
So, my 12-month baseline target price is $158 (approx. 211,404 KRW), and my rating is Hold (Medium confidence). That is just over 31 TP3 T from the current share price of $152.94.
If you look at the numbers one by one, the performance itself is hardly flawless. Cloud revenue increased by 621 TP3T to $11.607 billion (approx. 15.5 trillion won), and among them, Infrastructure as a Service (IaaS, a business that leases servers, storage, and networks) jumped 1211 TP3T to $7.388 billion (approx. 9.9 trillion won). This means it grew a little over double in just one year.
| item | Q1 FY2027 | 1 year ago | Increase/decrease |
|---|---|---|---|
| Total sales | $19.345 billion | $14.926 billion | +30% |
| Cloud (IaaS + SaaS) | $11.607 billion | $7.186 billion | +62% |
| └ Infrastructure (IaaS) | $7.388 billion | $3.347 billion | +121% |
| └ Application (SaaS) | $4.219 billion | $3.839 billion | +10% |
| Software License & Support | $5.55 billion | $5.721 billion | -3% |
| GAAP operating profit margin | 35% | 29% | +61 TP 3T points |
| GAAP earnings per share | $1.56 | 1.01 dollars | +55% |
| Non-GAAP earnings per share | $1.92 | $1.47 | +30% |
The GAAP operating profit margin increased from 291 TP3T to 351 TP3T. This means that out of every 100 won in revenue, 35 won remains after deducting labor costs, R&D expenses, and marketing expenses. Here, GAAP refers to figures calculated strictly according to accounting standards, while non-GAAP refers to figures adjusted by excluding items that the company considers "one-time or non-cash items," such as stock compensation expenses.
The contract backlog was also a hot topic. Remaining Performance Obligations (RPO), which refer to contract volume not yet recognized as revenue, rose to $664 billion (approximately 888.4 trillion won). This is $209 billion more than a year ago and an increase from $638 billion in the previous quarter. It effectively means securing work equivalent to nearly 10 times annual revenue in advance. During the quarter alone, new AI cloud contracts exceeded $30 billion, 300,000 GPUs were delivered to clients, and 850 megawatts of new data center capacity was put into operation.
The company's forecasts—or guidance—were not bad either. They projected Q2 revenue growth of $30–$341 and non-GAAP earnings per share of $1.85–$1.93. For the full year, they set a target of over $90 billion in revenue and $8.10 in non-GAAP earnings per share. However, these annual figures are almost identical to the $89 billion and $8.06 figures that the market had already expected. The first key point is that while the quarterly figures were significantly higher, the annual expectations remained largely unchanged.
How was the stock price react? On the day of the announcement, it closed at $152.94 during regular trading, down 5.41 TP3T, but rose to the $159–$163 range in after-hours trading, up 4–61 TP3T. You can view the original press release directly on the Oracle Investor Page.
Why You Should Not Trust the 30.9 Trillion Won Operating Cash Flow at Face Value
This is exactly where I stopped scrolling in the early morning. Oracle announced that its cash flow from operating activities for the first quarter increased by 1,841 TP3T to $23.103 billion (approximately 30.9 trillion won). Since this means cash flowed in equivalent to 4.85 times net profit, the numbers alone are fantastic.
However, while reading through the cash flow statement line by line, I discovered an unfamiliar item. It was the line titled, “Increase in customer advances with significant financial elements: $11.363 billion.” This was an item that was zero in the same quarter a year ago. Simply put, it is money that customers have sent in advance for construction costs to build data centers. Although classified as an operating activity under accounting rules, its nature is closer to funds borrowed from customers than money earned by Oracle through business.
| division | The announced figures | Excluding customer advance payments |
|---|---|---|
| Cash flow from operating activities | $23.103 billion | $11.74 billion |
| Capital investment | -$28.499 billion | -$28.499 billion |
| Free cash flow | -$5.396 billion | -$16.759 billion |
| Korean Won Conversion | Approximately -7.2 trillion won | Approximately -22.4 trillion won |
Free Cash Flow (FCF) refers to the actual cash remaining in hand after deducting capital expenditures from earnings. It is money that can be used to pay dividends, repay debt, or buy back treasury shares. A negative figure means that construction costs exceeded the profits generated from business operations. For the entire fiscal year 2026, Free Cash Flow was minus $23.686 billion (approximately -31.7 trillion won), while capital expenditures were $55.663 billion (approximately 74.5 trillion won).
Where did those construction costs come from? There are three sources. First, debt. As of the end of August, borrowings stood at $125.337 billion (approximately 167.7 trillion won), and quarterly interest expenses alone amounted to $1.428 billion (approximately 1.9 trillion won), an increase of 551 TP3T compared to a year ago. Second, the method of issuing and selling new shares. This quarter, $20 billion (approximately 26.8 trillion won) worth of shares were sold through an ATM (Available Market) program. The third is the customer advance payments we just saw.
When new shares are issued, the share for existing shareholders decreases. In fact, the diluted number of shares increased by 3.11 TP3T from 2.909 billion to 3 billion. This means that even if the company's total profit remains the same, the share taken by each share has decreased by 31 TP3T. A structure where negative cash flow and debt mask strong earnings is Patterns seen in Coreweave (CRWV) analysisIt is very similar to [another brand]. It is just much larger in scale.
What kind of company is Oracle now?
Many of you probably still remember Oracle as a “company that sells enterprise databases.” That is only half right. Looking at the revenue breakdown for this quarter, cloud accounted for 601 TP3T, traditional software for 291 TP3T, and hardware and services for the remainder. Since cloud accounted for 481 TP3T a year ago, the company’s center of gravity has shifted significantly in just one year.
Let me give you an analogy. Oracle was originally a shop in a prime location that had been serving loyal customers for over 20 years. It was the kind of shop where the clientele rarely changed, and after paying the rent, there was a steady cash surplus every month. However, the owner has now bought three or four adjacent buildings at once and is simultaneously constructing branches 2, 3, and 4. The construction costs are being covered through loans, the sale of equity, and the promise, "If you pay a deposit in advance, we will secure a table for you." Business at the main branch is still going well. The problem is that the construction needs to finish on schedule, and the customers who made reservations need to actually show up.
What is important here is the list of booked clients. Oracle's $664 billion contract backlog is heavily concentrated in a small number of large AI companies, with OpenAI reportedly accounting for a particularly large share. OpenAI is a company that continues to raise external funds while still operating at a loss. The fact that the financial situation of clients who have paid deposits directly impacts Oracle's stock price is the most peculiar aspect when looking at this company.
Oracle (ORCL) Valuation: Is 17.6x Cheap?
Valuation is the process of determining whether a stock is cheap or expensive. The most commonly used metric is the Price-to-Earnings Ratio (P/E), which is an indicator showing how many times the stock price is compared to one year's earnings. A P/E ratio of 20 means that it would take 20 years for current earnings to accumulate to reach the stock price.
I calculated it myself. I set the projected earnings per share for the next 12 months at $8.68. This figure is derived by subtracting the already reported first-quarter figure of $1.92 from the company's stated annual figure of $8.10 and adding the estimate for the first quarter of the next fiscal year of $2.50 to fill the gap. Dividing the current price of $152.94 by this gives a forward P/E ratio of 17.6x.
The issue is the benchmark for determining whether "17.6 times is cheap." I did not use the current multiples as they are, but rather combined two. One is Oracle's own 5-year average forward P/E, which is approximately 21 times. The other is the median of similar large infrastructure software companies.
| enterprise | Preceding P/E | Sales growth rate | Operating profit margin | Free cash flow |
|---|---|---|---|---|
| Oracle (ORCL) | 17.6 times | +30% | 35% | -$24.5 billion |
| Microsoft (MSFT) | 20.9 times | +17.7% | 45.1% | +$16.6 billion |
| SAP | 21.3 times | +9.4% | 27.6% | +$9.1 billion |
| Amazon (AMZN) | 24.2 times | +19.6% | 13.7% | +$3.2 billion |
| IBM | 17.8 times | +1.1% | 16.5% | +$12.1 billion |
| Coreweave (CRWV) | Deficit | +112.5% | -1.9% | -$9.1 billion |
The median for the four companies excluding Coreweave is 21.1 times. This falls almost exactly at the same level as our own average of 21 times. Here, I apply a discount of approximately 51 TP 3 T to determine the appropriate multiple. 20 timesI set it at that level. There are two reasons for the discount. First, credit rating agency S&P downgraded Oracle's credit rating to BBB-, just above speculative grade, and second, stock dilution is ongoing. It is natural for the market to assign a lower multiple to a company with high debt and an increasing number of shares, even with the same earnings.
Since it is currently 17.6 times compared to the appropriate 20 times, 11.9% DiscountIt is in that state. If you look only at the multiple, it is a bit cheap. Therefore, the target price calculated using this method is $8.68 × 20 times = $173.5.
If we recalculate using Discounted Cash Flow (DCF),
DCF is a method of calculating a company's future cash earnings at their current value and adding them up. Since 1 million won in the future is worth less than 1 million won today, it is deducted at a certain rate; this rate is called the Weighted Average Cost of Capital (WACC). This method is essential for companies like Oracle that burn cash now to earn money later.
| DCF assumption | 값 | reason |
|---|---|---|
| WACC (discount rate) | 11.0% (Range 10~12%) | Beta 1.73, Risk-free Rate 4.21 TP3T, Equity 781 TP3T · Borrowing 221 TP3T structure |
| Permanent growth rate | 3.0% (Range 2.5~3.5%) | Long-term nominal economic growth rate level |
| forecast period | 7 years | From the completion of data center construction until normalization |
| Free Cash Flow Path | FY27 -$32 billion → FY29 +$8 billion → FY33 +$68 billion | Capital investment peaked in FY27 and gradually decreased |
| Calculation result | $102–$180 per share | The median is about $134 |
The value calculated using the central assumption (WACC 11%, perpetual growth 3%) is $133.6 per share. This is about $40 lower than the $173.5 calculated using multiples. The reason the two methods diverge is clear: P/E looks only at accounting earnings, whereas DCF considers the point at which those earnings are converted into cash.
I have two values Multiple 60% : DCF 40%I mixed them. In the DCF calculation, the enterprise value's 851 TP 3 T is derived from the perpetual value after 7 years, but I reduced its weight because that part fluctuates too significantly depending on the assumptions. Conversely, the multiples are much more verifiable figures because they use the annual earnings guidance directly provided by the company. The result of mixing them is... $157.6And, rounded, we set the baseline target price at $158.
Oracle (ORCL) Target Price Scenarios and Expected Returns
| scenario | Target price | Won | Compared to the current price | weight |
|---|---|---|---|---|
| Bull | $208 | Approximately 278,304 won | +36.0% | 25% |
| Base | $158 | Approximately 211,404 won | +3.3% | 45% |
| Bear | $115 | Approximately 153,870 won | -24.8% | 30% |
The bullish $208 figure is based on a forward P/E ratio of 24x. The scenario envisions contract backlogs turning into revenue as scheduled, capital expenditures peaking in this fiscal year and then declining, and free cash flow returning to surplus around 2029. If that happens, I believe the market will treat Oracle again at the industry average multiple.
A bearish target of $115 represents a forward P/E of 13.3 times. This scenario involves financing costs jumping as a major AI client delays or adjusts a contract, capital expenditures exceed plans, and the credit rating is downgraded one notch to speculative grade. Coincidentally, this figure is nearly identical to the 52-week low of $114.50 (approximately 153,201 KRW) actually hit this July. It also implies that the market is at a level it visited just two months ago.
I set the weights at a baseline of 451 TP3T, a bullish 251 TP3T, and a bearish 301 TP3T. While this does not give the highest probability to the bearish scenario, I increased it by 51 TP3T points compared to the usual 251 TP3T. The reason I did this, even though the multiple signal indicates undervaluation, is that the company's downside risk is not "slowing growth" but "blocking financing." Slowing growth causes the stock price to fall slowly, whereas financing issues cause it to drop all at once. Since these are risks of a different nature, I added a slightly higher probability to them.
For reference, the average target price of 42 Wall Street analysts is $241.43 (approximately 323,033 KRW). That is a difference of $83 from my target of $158. However, over the past four weeks, a succession of analysts have lowered their target prices. Scotiabank lowered its target from $241 to $215, Jefferies from $320 to $290, TD Cowen from $300 to $240, and UBS from $285 to $245. CLSA went even further and assigned a "Hold" with a target of $145. While the average is still high, the trend is pointing downwards.
Cut in half in a year, what happened in the meantime?
To understand Oracle's current stock price, we need to go back to last September. When contract backlogs jumped to $455 billion during the September 2025 earnings announcement, the stock price surged 361 TP3T in a single day, soon rising to $331 (approximately 442,878 KRW). That is the 52-week high.
Over the past year, the stock price has dropped 47.51 TP3T. From its peak, it has fallen 541 TP3T. In July of this year, it fell to $114.50 but has now recovered slightly to $152.94. Earnings have improved every quarter, but the stock price has moved in the opposite direction.
There are three main reasons why the market has changed its mind. First is the scale of capital expenditure. $55.6 billion was spent in fiscal year 2026, and this is projected to increase to up to $95 billion (approximately 127.1 trillion won) in 2027. Even if up to $25 billion is recovered from customers, the remaining burden is significant. Second is the credit rating. S&P has downgraded the rating by one notch to BBB-, leaving it just one notch away from speculative grade. Third is customer concentration. It has been repeatedly pointed out that a significant portion of the contract balance is skewed toward OpenAI alone.
To summarize, the market is not doubting whether the growth is genuine. It is asking whether the price paid for that growth is reasonable. Given that growth has been confirmed but the timing of payment collection is unclear, The gap between IREN's contract ARR and actual quarterly revenueIt's the same kind of question as when looking at it. Like Nvidia Cases where profits are at an all-time high but operating cash flow fails to keep up.Several have also come out this year.
Oracle (ORCL) Risks and Bear Triggers
I will summarize the risk into three steps: “What is the problem → How it affects the stock price → What you need to check.”.
- Customer Concentration Risk If a single major AI client downsizes a contract or delays payments, the credibility of the entire $664 billion contract balance is shaken. → You just need to look at how the remaining performance obligations and advances items in the quarterly report move compared to the previous quarter.
- financing risk → If the credit rating drops to speculative grade, bond interest rates jump, and shareholders' share decreases again because more shares must be issued. → Check the rating adjustment disclosures from S&P and Moody's and the trends in quarterly interest expenses.
- Depreciation time bomb Depreciation expense for this quarter was $3.156 billion, an increase of $1,341 million compared to a year ago. This is the cost of writing off the books for built data centers every year; once the $95 billion in capital investment is recognized as an asset, it will continue to suppress accounting profits for the next two to three years. → You just need to look at the point where the growth rate of depreciation expense on the income statement exceeds the growth rate of revenue.
- dilution risk The number of shares increased by 3.11 TP due to the sale of $20 billion worth of stock. The lower the stock price, the more shares are needed to raise the same amount of money. → You can see this immediately if you compare the number of diluted shares every quarter.
- SaaS growth slows → Cloud application revenue grew by only 101 TP3T, while traditional software decreased by 31 TP3T. The engine to support it is weak if the infrastructure business falters. → Check if the quarterly SaaS growth rate maintains double digits.
If I had to pick just one bear trigger, it would be news of major clients, including OpenAI, raising funds. Not Oracle's earnings. customer's We are currently in a phase where financial conditions are driving Oracle's stock price. Since this is a variable beyond the control of Oracle's management, you need to be extra cautious.
Personally, this is a point that bothers me.
To be honest, what held me back the longest was the sale of $20 billion in stock. A year ago, the company's stock price was $331. To raise the same $20 billion at that time, just over 60 million shares would have sufficed, but at around $150, more than double that amount of stock was needed. In other words, the timing of the capital raise was the worst possible. Of course, they might not have known they would need that much money last September. However, the question remains: "Why didn't they raise it in advance then?".
The second thing that bothered me was the disparity in sentiment I felt while creating a table of analyst target prices myself. The average is $241, but there have been four downward adjustments in the past four weeks alone, with the lowest target price being $110. It was striking that the highest figure of $400 and the lowest of $110 were for the same company. This level of dispersion can be interpreted as meaning "no one is certain." That is why I also set my confidence level to Medium rather than High. I decided to trust it only half-heartedly until I can actually confirm that the $664 billion in contract balances turns into cash in one or two quarters.
Oracle (ORCL) Investment Opinion Conclusion
My opinion is HoldThe confidence level is Medium. The 12-month baseline target price is $158 (approx. 211,404 KRW), and the expected return is +3.01 TP3T.
Let me explain my three reasons for viewing this as a Hold. First, the valuation signals are clearly favorable. The forward rate of 17.6x is 11.91 TP3T cheaper than my fair value of 20x and lower than the industry median of 21.1x. Second, the fundamentals on the income statement are also improving. With revenue growing by 301 TP3T and the operating profit margin rising by 61 TP3T points, this is undeniable. Third, however, cash pressure offsets these two factors. Considering negative free cash flow, $125.3 billion in debt, 3.11 TP3T equity dilution, and a BBB- credit rating, the expected return is suppressed to around +31 TP3T. To raise the rating to Buy, the expected return would need to be at least +121 TP3T, but it has not reached that level.
The average Wall Street opinion is Buy, and the target price is much higher at $241. What I viewed differently is how the gap between accounting earnings and actual cash is evaluated. Wall Street seems to base its valuation on the assumption that contract balances smoothly translate into revenue, and revenue into cash. I, however, consider that path to involve at least two years and an additional $95 billion in investment, and I discounted accordingly.
If you already hold the stock, I don't see much reason to rush to sell right now. However, for those looking to enter the market, I recommend waiting to see one of two things before making a move. One is a signal that the quarterly free cash flow deficit is starting to narrow, and the other is news that a major client's financing has been successfully completed. If either of these is confirmed, the probability of my bearish scenario will drop significantly, and at that point, I intend to adjust my target price upward.
The dividend remained at $0.50 (approximately 669 KRW) per quarter. It will be paid on October 23 to shareholders as of October 9. With an annualized dividend yield of 1.31 TP3T, this is not a stock to invest in solely for the dividend. You should also keep in mind that $1.565 billion is being paid out in quarterly dividends alone, while the free cash flow is negative.
Frequently Asked Questions
What do you estimate the target price for Oracle (ORCL) to be?
My 12-month baseline target price is $158 (approximately 211,404 KRW), and my rating is Hold. I have set the bullish scenario at $208 and the bearish scenario at $115. The average of 42 Wall Street analysts is $241.43, which is significantly higher than mine, but please also note that four firms have lowered their target prices in the last four weeks.
Why has the stock price been cut in half in the past year when the performance is so good?
What the market sees is not revenue, but the costs of generating that revenue. Oracle plans to increase capital expenditures by up to $95 billion in fiscal year 2027, and it is funding this through debt and the issuance of new shares. With its credit rating downgraded to BBB- and free cash flow remaining negative, sentiment has shifted toward the view that the cost of growth is too high.
If the contract balance is $664 billion, does that mean the revenue is confirmed?
Since the contract has been signed but the amount has not yet been recognized as revenue, you should view it differently from confirmed cash. Revenue is generated only when Oracle actually builds and delivers the promised data center capacity, and the customer must also maintain the ability to pay for it. Given that the scale is nearly 10 times the annual revenue, it also takes a long time.
Is there a possibility that Oracle could be in danger due to debt?
There is no immediate liquidity issue. The company holds $36.369 billion (approximately 48.7 trillion won) in cash equivalents, and its core software support business consistently generates cash. However, you should be cautious regarding the credit rating of BBB-, just above speculative grade, and the fact that quarterly interest expenses have increased by 551 TP3T in one year. If the rating drops one notch further, borrowing costs will rise significantly.
Reference materials
- Oracle, Q1 FY2027 Earnings Press Release (2026-09-10)
- Oracle, FY2026 Q4 & Annual Earnings Press Release
- PR Newswire, Oracle Q1 FY2027 Results (Including Financial Statements)
- Yahoo Finance reports on Oracle's credit rating and debt burden
- Forbes, Oracle Stock Price Debate in Second Half of 2026
I used AI research tools during the writing process and personally verified the numbers and logic one by one before publication. Stock price, exchange rate, and earnings estimates are based on the date of writing, September 11, 2026, and are subject to change thereafter. Please remember that this article is for informational purposes only and is not an investment recommendation; the final judgment and its consequences rest solely with the investor.