I stopped at a single sentence while rereading the Nevius conference call transcript. It was a remark by the CEO stating that, under current conditions, they could sell their entire 2027 capacity today. However, I have lowered the 12-month target price from $290 in August to $215, and my rating remains Hold. Let me explain, step by step, why I lowered the target price despite the good news.
The analysis reference date is September 23, 2026, and the stock price is the previous US market closing price of $236.12. All KRW values in the text were calculated at 1,386.0 KRW per dollar.
- Conclusion First: Nevius (NBIS) Investment Opinion and Target Price
- What kind of company is Nevius?
- Three catalysts driving Nevius's stock price right now
- The numbers left by Nevius (NBIS) Q2 earnings
- Navius (NBIS) Valuation: Is an ARR of 8.4x expensive?
- Nevius (NBIS) Target Price Scenarios and Expected Returns
- Nevius (NBIS) Risk and Bear Trigger
- I saw things like this too, besides numbers.
- Nevius (NBIS) Investment Opinion: Why We Maintain Hold
- Nevius (NBIS) Frequently Asked Questions
- source
Conclusion First: Nevius (NBIS) Investment Opinion and Target Price
The 12-month baseline target price is $215 (approx. 297,990 KRW), the investment opinion is Hold, and the confidence level is Low.
The business has improved since I saw it in August. The capacity auction was won at a price 151 TP3T higher than the previous highest bid, and the year-end ARR (annual recurring revenue like a subscription) target of $7 billion to $9 billion (approximately 9.70 trillion to 12.47 trillion won) was maintained.
However, this year's capital investment plan is between $20 billion and $25 billion (approximately 27.7 trillion to 34.7 trillion won). This means that a company with a market capitalization of $64.19 billion (approximately 88.97 trillion won) is spending more than one-third of its market cap on investments alone in a single year. Therefore, I believe that the current stock price has already largely reflected a bullish scenario.
What kind of company is Nevius?
Simply put, Nevius is a company that purchases AI computing equipment in bulk and rents it out on an hourly basis. It is similar to a landlord building apartments and collecting monthly rent, but there is one difference. Since these 'apartments' are NVIDIA GPUs, their value drops significantly after 3 to 5 years. Therefore, the speed at which the rent is collected determines the company's survival.
Its headquarters are located in Schiphol, Netherlands, and it has 1,543 employees. Its original name was Yandex NV, but it changed to Navius in August 2024. In addition to its core AI cloud business, it also owns an edtech subsidiary, Triple Ten, and an autonomous driving subsidiary, Avride.
What sets this company apart from other AI cloud providers is its relationship with Nvidia. Nvidia has directly invested $2 billion (approximately 2.77 trillion won) and formed a strategic alliance. When a chip manufacturer holds a stake, the likelihood of being pushed to the back of the order allocation list decreases. Original text of NVIDIA's announcementIf you read it together, you will get the context.
Three catalysts driving Nevius's stock price right now
| Point of view | event | Impact on stock price and earnings |
|---|---|---|
| August 12th | Q2 Earnings Announcement: Revenue +4541 TP3T, Adjusted EBITDA Turns Positive | It surged 341 TP3T for the day to close at $259.20. |
| August~September | Four large deals exceeding $1 billion each, a contract backlog of $37.5 billion | Advance payments recover 50~601 TP3T of the facility investment first |
| September | Capa sold for 151 TP3 T higher than the highest price just before the auction. | Price-setting power confirmed, yet the stock price remains 211 TP3 T below the peak |
| September | Inclusion in FTSE All-World Index / Rothschild Redburn Sell · Open Target at $84 | Supply and demand are favorable, while credit risk warnings are moving in the opposite direction. |
You should read this table from left to right. The key point is that two completely opposite signals appeared together within the month of September.
Catalyst 1. The price is rising.
The winning bid in the capacity auction conducted by the company came out 151 TP3T higher than the previous record high. Short-term premium contracts are being negotiated at the $40 million to $50 million per megawatt level, whereas long-term core contracts were $20 million to $25 million, meaning this is nearly double the price. Since this means receiving double the value for the same amount of power, I viewed this as the strongest signal of this quarter.
Catalyst 2. The customer pays first.
At the end of the second quarter, deferred revenue (prepayments for which contracts have been signed but have not yet been recognized as revenue) increased to $6 billion (approximately 8.32 trillion won). This means that over 8 trillion won in money paid in advance by customers alone is currently in the company's coffers. It surged by 251 trillion won in just one quarter from $4.78 billion in the previous quarter. Since it is essentially receiving rent before purchasing equipment, it is a structure that significantly reduces the financial burden.
Catalyst 3. However, the financing method is a concern.
Advance payments alone are insufficient to cover capital expenditures. The company replenished its coffers with a $775 million asset-backed loan and a $2.8 billion ATM capital increase (a method of selling new shares in the market in small increments). The problem lies with the capital increase. As the number of shares increases, the share per existing share decreases. In fact, the number of issued shares increased by more than 131% over the past year, rising from 239.8 million to 271.9 million.
The numbers left by Nevius (NBIS) Q2 earnings
Revenue for the second quarter was $582.3 million (approx. 807.1 billion KRW), an increase of 4,541 TP3T from a year ago. Gross profit was $448.7 million (approx. 621.9 billion KRW), resulting in a gross profit margin of 77.11 TP3T. This means that for every 100 KRW sold, 77 KRW remains after deducting direct costs, which is very high for a cloud business.
However, the company recorded an operating loss of $175.9 million and a net loss of $190.4 million (approximately 263.9 billion won), resulting in a loss of $0.68 per share. The reason for the deficit, despite good margins, is depreciation expenses. Quarterly depreciation expenses amounted to $259.7 million, consuming 451 TP3T of revenue. This means that the amount of purchased GPUs that are written off as expenses on an accounting basis accounts for nearly half of revenue.
The financial condition was more solid than I had feared. With cash at $8.042 billion and total borrowings at $10.171 billion, net borrowings stood at just $504 million. This means that debt and cash are closely intertwined. The current ratio is also 4.03, indicating that assets available within one year are four times greater than the money due for repayment.
However, free cash flow (the actual cash remaining after deducting capital investments from earnings) is minus $9.6 billion based on the past year. That is because we spent $5.7 billion (approximately 7.90 trillion won) on capital investments in the second quarter alone. I Q2 earnings review written in AugustI wrote that “the axis of judgment is financing, not the contract,” and even now, a month and a half later, that thought remains the same.
The original performance figures are Company Q2 Earnings Press ReleaseYou can check it directly at.
Navius (NBIS) Valuation: Is an ARR of 8.4x expensive?
Since Nevius is operating at a loss, we cannot use the P/E ratio (price-to-earnings ratio, indicating how many times the stock price is compared to one year's earnings). The forward earnings per share estimate is minus $3.29. The trailing P/S ratio is 47.4 times, but this is also meaningless. Applying the past year's revenue to a company whose sales jump every quarter distorts the figures.
So, I divided the enterprise value (EV, market capitalization plus debt minus cash) by the projected revenue for the next 12 months. With an enterprise value of $66.87 billion (approximately 92.68 trillion won) and projected revenue of $8 billion for the next 12 months, taking into account the median year-end ARR target and ramp-up speed, the multiple comes out to 8.4 times.
| item | Nevius (NBIS) | Coreweave (CRWV) | IREN |
|---|---|---|---|
| stock price | $236.12 | $86.76 | $48.55 |
| Enterprise Value (EV) | $66.87 billion | $93.93 billion | $20.56 billion |
| Lagging sales growth rate | +454% | +112% | -27% |
| EV / Revenue for the next 12 months (my estimate) | 8.4 times | 6.7 times | 8.2 times |
| Adjusted EBITDA margin | 41% | 49.9% | 5.4% |
| Net borrowings | $500 million | $46.07 billion | $1.94 billion |
It is best to start looking at this table from the bottom two rows. Although Nevius has a higher multiple than Coreweave, its net debt is on a completely different level—$46 billion versus $500 million. As for Coreweave's debt structure, I A separate postI took a close look at it. Airen has a similar multiplier, but since the margin is only 5.41 TP 3T, it is in a different class. Airen's contract ARR structure is This articleI wrote it down.
The order in which I determined the appropriate multiple is as follows. The median for peers was 7.5 times, and since Nevius has much lighter financials and higher margins, I applied a premium of 201 TP3T to set it at 9.0 times. Then, I rounded it down to 8.0 times by reflecting the DCF results explained below by about 111 TP3T. The current 8.4 times is 5.01 TP3T more expensive than my fair value.
It comes out much lower if calculated using DCF.
DCF is a calculation that adds up the cash a company will earn in the future, discounted to today's value. I set the discount rate (WACC, the rate used to reduce future money to today's value) at 121 TP 3 T, the terminal growth rate at 3.01 TP 3 T, and the forecast period at 5 years. It was difficult to apply a generous discount rate because the beta was high at 1.44 and the size of the debt was large.
Based on a sustainable free cash flow of $4 billion to $7 billion in 2030, the fair value of the stock came out to be between $95 and $165. This is less than half of the current stock price. I initially thought I had made a calculation error, so I ran the analysis twice, but the results were the same. This means that for the company to justify its current value, it must continue to add over 1 gigawatt of capacity every year, even in the 2030s.
Nevius (NBIS) Target Price Scenarios and Expected Returns
| scenario | Key assumptions | Target price | Compared to the current price | Probability |
|---|---|---|---|---|
| stress | $9 billion in revenue over the next 12 months, maintaining a 10x multiple | $305 (approx. 422,730 won) | +29.2% | 25% |
| basic | $8 billion in revenue over the next 12 months, a fair multiple of 8.0x | $215 (approx. 297,990 won) | -8.9% | 50% |
| Weakness | Reducing to $7 billion in revenue over the next 12 months and multiples to 7.5x | $175 (approx. 242,550 won) | -25.9% | 25% |
When the three scenarios are weighted by probability, the expected return comes out to minus 3.61 TP3T. This means that it is considered most likely that the price will be near or slightly below its current level one year from now.
I will also note down the growth rate assumptions. The implied growth rate derived from the ARR flow was +1671 TP3T, the analyst consensus (average forecast) was +1451 TP3T, and the AI cloud industry average was +551 TP3T. I mixed these in a 4:4:2 ratio and used +135.81 TP3T. The growth rate itself is still in the top tier of the market.
The average target price of 19 analysts is $276.26. That is 221 TP3T higher than my $215. There are two points of difference. I used a fully diluted number of shares that accounted for convertible bonds, and I rounded down the multiple results using DCF. Since the consensus structurally tends to lean toward bullish, I did not follow it exactly.
Nevius (NBIS) Risk and Bear Trigger
- Dilution of capital increase → The number of shares increased by 131 TP3T due to a $2.8 billion ATM capital increase. There is also a possibility of additional capital increases if CAPEX continues to grow. → Please compare the weighted average number of outstanding shares in the quarterly reports every quarter.
- Depreciation cliff → Since the lifespan of a GPU is 3 to 5 years, only costs remain the moment revenue stops. → Check if the quarterly depreciation ratio (currently 45%) continues to rise.
- Customer Focus Four deals worth over $1 billion account for a large portion of revenue. The loss of even just one would be a significant blow. → Please check the disclosure regarding the proportion of top clients in the earnings announcement.
- Building a hyperscaler in-house If large customers who currently lease data centers build their own, demand will decrease. → Look at the quarterly capital expenditure directions of Microsoft, Meta, and Google together.
- Supply and demand risk The short selling ratio relative to outstanding shares is 19.21 TP3 T. Michael Burry also stated that he increased his short positions. → Please look at the short selling balance disclosed twice a month.
I identify three bearish triggers. First, if the year-end ARR guidance falls below $7 billion. Second, if financing stalls and the capital expenditure (CAPEX) plan is reduced. Third, if even a single major deal is cancelled. The basis for Rothschild Redburn’s "Sell" recommendation with a target price of $84 was also ultimately credit market risk. Summary of the reportIt is good to refer to this when checking opposing arguments.
I saw things like this too, besides numbers.
I created a table and placed Nevius and Coreweave side by side, and I noticed something interesting. The market groups the two companies together under the umbrella term "AI Cloud," but their financial structures are polar opposites. Coreweave was built on debt, while Nevius was built on advance payments and capital increases. Both are risky, but the types of risk differ. For Coreweave, interest rates and maturities are frightening, whereas for Nevius, the number of outstanding shares is daunting. Personally, I viewed the latter as a slightly easier structure to weather.
To be honest, there are some aspects that make me hesitate. The CEO’s statement that “we could sell all our 2027 capacity today” is a strong statement, but I decided to believe it only half-believe until the full contract unit price is disclosed. It reminded me of 2021, when mining companies were clamoring to switch to data centers. The decisive difference from then is that this time, the contract between Nvidia and Meta is a real one. That is why I set it to Hold rather than Sell.
Nevius (NBIS) Investment Opinion: Why We Maintain Hold
I will summarize my reasons for viewing this as a Hold in three lines. First, the valuation is 51 TP3T above the fair value, so it is neither significantly expensive nor cheap. Second, the probability-weighted expected return is -3.61 TP3T, which falls below my standard Hold range (between -81 TP3T and +121 TP3T). Third, the multiples are fair, but the DCF indicates overvaluation, so the two signals conflict. Therefore, I lowered the confidence level to Low.
Lowering the target price from $290 in August to $215 was not because the company has deteriorated. It is because I started using the fully diluted number of shares and reflected the DCF results in the target price. For the same reason, my approach differs from that of companies like Nvidia that are already generating cash. Regarding the cash flow analysis for Nvidia... This articleI've organized it there.
If you already hold the stock, it is not too late to make a decision after confirming the finalized year-end ARR guidance and the size of the fourth-quarter capital increase. If you are looking to enter the market, you will feel more at ease if you approach it in installments, considering the possibility that a bearish scenario around $175 could materialize.
Nevius (NBIS) Frequently Asked Questions
What is the target price for Nevius?
My 12-month baseline target price is $215 (approximately 297,990 KRW). I set the target at $305 for a bullish scenario and $175 for a bearish scenario. The average target price of 19 analysts is $276.26, which is higher than my figure.
Why is Nevius in the red even though its revenue increased by 4,541 TP3T?
It is due to depreciation expenses. The cost of purchased GPUs is spread out and written off as an expense in accounting; in the second quarter alone, that amount was $259.7 million, which was 451% of revenue. Although this is not an actual cash outflow, book profit decreases by that amount.
Which is safer, Nevius or Coreweave?
Looking solely at financials, Nevius is lighter. Its net debt is $504 million, whereas Coreweave's exceeds $46 billion. However, Nevius faces the burden of an increased number of shares as it raises capital through a capital increase. You can view them as having different levels of safety.
Why is it important that the Capa auction price has gone up?
That is because it means receiving more money for the same power and equipment. Short-term contracts are being negotiated at $40 million to $50 million per megawatt, while key long-term contracts were $20 million to $25 million. You should interpret this as a signal that demand is tighter than supply.
May I enter Nevius now?
My recommendation is to hold, so it is difficult for me to actively recommend it. The expected return came out at minus 3.61 TP3T. While growth will continue, I believe a significant portion of that growth is already reflected in the stock price. If you are interested, I recommend approaching it by buying in installments rather than all at once.
source
- Nebius Q2 2026 Earnings Press Release
- NVIDIA Announces Strategic Partnership with Nebius
- Nebius Signs AI Infrastructure Agreement with Meta
- Rothschild Redburn, Nebius Sell Opinion/Target Price $84
I drafted this with the help of AI research tools and personally reviewed the numbers and logic before publication. Stock price, exchange rate, and earnings estimates are based on the time of writing and are subject to change. Please remember that this is an informational post, not an investment recommendation, and that the final judgment and its consequences rest solely with the investor.