AI Is Growing, So Why Are Semiconductor Stocks Crumbling?
Conditions for holding or trimming Samsung Electronics and SK hynix, based on the October 20 KRX·NXT close and U.S. premarket, weighing AI investment returns, memory competition, and capital allocation.
Read as text
GOGUMA TEAM · SEMICONDUCTOR STRATEGY
October 20, 2026 · Korea Market Close and U.S. Premarket
AI Is Growing
Semiconductor Stocks
Why Are They Crumbling?
Reassessing Samsung Electronics · SK hynix
Should You Sell Now, or Hold While Watching for What?
Samsung Electronics 005930 · KRX / NXT _₩209,000 NXT 20:00 close · −4.13% from previous close _ KRX 15:30 close ₩211,500 · −2.98%
SK hynix 000660 · KRX / NXT _₩1,190,000 NXT 20:00 close · −6.81% from previous close _ KRX 15:30 close ₩1,223,000 · −4.23%
**There are still reasons to hold. They are not reasons to hold on at all costs.**Maintain unleveraged stock holdings at a level you can afford, but hold off on adding. Act without waiting for actual loss announcements if orders, earnings outlook, or value accruing to shareholders deteriorates.
Reflects Korea's October 20 KRX and NXT market close
U.S. premarket prices at 21:47 KST · Excludes results from the U.S. regular session
Goguma Team · Conditional analysis based on public information
1. The Bottom Line — Hold, but change your reasons for holding
Maintain existing unleveraged stock holdings in Samsung Electronics and SK hynix at an affordable position size. Hold off on additional purchases. Selling everything solely because of this decline, or mechanically trimming SK hynix first, is not the default course. Conversely, if you have borrowing or collateral shortfalls, need the money soon, or are overconcentrated beyond your loss tolerance, reduce exposure without waiting for a rebound.
The October 17 report reviewed the sharp decline in the U.S. market and concluded that “strong earnings alone cannot guarantee that customers will have the funds to invest or that share prices will recover.” U.S. semiconductor stocks rebounded on Monday, but the Korean market fell sharply again on Tuesday, and SK hynix traded even lower overnight than during the regular session. It has become clearer that this is not a problem that can be solved simply by waiting for a one-day rebound in the U.S.
That said, there is still insufficient evidence to give up on the industry. Actual memory earnings and near-term supply constraints remain in place. The market is asking tougher questions—not whether AI is being adopted, but whether the capital and power needed to support that growth will be available on time, how long today's high memory profits will last, and how those profits will be returned to shareholders.
This decision to hold does not mean we have called the bottom. Selling may avoid adverse announcements, but could mean missing a rebound; holding carries the opposite risk. We have not been able to calculate reliable numerical probabilities of returns for either choice. However, considering the currently confirmed business fundamentals alongside prices, we prefer to maintain the holding thesis and respond to specific deterioration rather than abandon the position entirely now on the assumption that the industry will collapse.
| Situation | Action now | What not to wait for |
|---|---|---|
| You hold the stocks without borrowing and can withstand further volatility | Maintain existing Samsung Electronics and SK hynix holdings; hold off on adding | Material changes to orders, contracts, or earnings outlook |
| Borrowing, collateral, or upcoming cash needs make further declines difficult to withstand | Reduce excess exposure | A return to breakeven or a strong earnings announcement |
| Changes in orders, delivery schedules, or pricing from key customers lower the outlook for recurring earnings | Reassess company value and reduce the position size | A subsequent announcement of actual losses or cash shortages |
| The Solidigm transaction takes shape on economic terms unfavorable to the parent company's shareholders | Reassess SK hynix separately and reduce the position | A broad statement that “the AI business is doing well” |
The precise amount to trim can only be determined after considering total investable assets, cash, borrowing, and loss tolerance together. This public report does not assume an individual's account or prescribe a blanket sell percentage.
Check your account constraints now. If applying the additional downside you set in advance to the NXT closing price would breach even one of your collateral maintenance requirements, near-term cash needs, or acceptable loss limits, reduce your position in the next regular session to a level at which you can meet those constraints. A price decline alone does not change the business outlook, but a reduction may be needed immediately to manage account risk.
2. How much of today's decline has been confirmed?
The KOSPI fell 3.73% to 5,552.74. It recovered somewhat from its intraday low of 5,493.24, but decliners outnumbered advancers by a wide margin, 773 to 118. It was a day when broad market risk reduction accompanied the decline—not an event limited to semiconductors.Market Close
| Category | Samsung Electronics | SK hynix |
|---|---|---|
| October 20 KRX close at 15:30 | 211,500 won (−2.98%) | 1,223,000 won (−4.23%) |
| NXT close at 20:00 | 209,000 won (−4.13%) | 1,190,000 won (−6.81%) |
| Difference between the NXT close and the KRX close | −1.18% | −2.70% |
Figures in parentheses are the provider's daily percentage changes from the previous day. The final row shows the difference between the two markets' closing prices for the day. The KRX figures were cross-checked against closing reports and separate price data; the NXT figures are the provider's 20:00 closing prices, retrieved at 21:45. In the after-hours lookup, the displayed representative price switches to the NXT price, so it was not used as the KRX closing price. Sources: Seoul Shinmun, Cross-check against KRX price data, Naver market-by-market price data.
According to reports at the regular-session close, individuals were net buyers of KRW 898.1 billion, while foreign investors and institutions were net sellers of KRW 530 billion and KRW 1.0048 trillion, respectively. Other corporations were net buyers of KRW 636.7 billion. Therefore, it is also inaccurate to say that “individuals alone absorbed all the selling.” Other corporations as a group cannot simply be assumed to represent share buybacks.Closing flows by investor type
We have not estimated net buying by individual investors in each stock, as we lack a reliable cross-check against the latest raw data. Market-wide flows should not be interpreted as flows in the two stocks.
More importantly, SK hynix fell an additional 2.70% on NXT relative to its regular-session closing price. Ignoring this, the small afternoon rebound cannot be described as a recovery signal. However, a price decline alone does not confirm order cancellations or permanent impairment of corporate value.
3. Why did Korea fall again while the U.S. rebounded?
In the U.S. regular session on October 19, the Nasdaq rose 1.11%, the semiconductor index gained 1.61%, and Micron advanced 2.97%. However, CoreWeave fell 0.44%, and the U.S. 10-year yield rose from 5.098% to 5.136%. A rebound in some AI-related stocks did not mean that overall financial conditions had normalized. Many stocks outside the technology sector were also weak in the U.S. market.U.S. Market Close, Interest Rate Time Series
In Korea, foreign and institutional selling, concerns about the profitability of AI investment, global interest rates, and geopolitical pressures all converged. Other sectors, including defense, were also notably weak. In a market with a high weighting in semiconductors, trading to reduce exposure to the same industry risk may have amplified the index decline. However, we have not established the contribution of each factor to the decline or confirmed forced liquidations by any particular fund.Domestic Market Close Commentary
As of the report's writing on the night of October 20, the sustainability of the rebound remained unconfirmed. In U.S. premarket trading at 21:47 KST, Micron was at $886.37, about 2.05% below the previous regular-session close; NVIDIA was at $215.89, down 0.55%; and CoreWeave was at $59.02, down 3.44%. This is a premarket snapshot, not the closing result of Tuesday's U.S. session.Micron Premarket Raw Data, NVIDIA, CoreWeave
The 10-year yield checked that same night was 5.081%, below the previous close, but AI stocks were weak. WTI futures were about $100.07 per barrel at 21:37, according to the provider's timestamp.**A simple assumption that semiconductor stocks will immediately recover whenever interest rates fall for a day does not hold.**Customer financing, earnings sustainability, and oil price pressures must all be considered together.Pre-Market News
4. The weakest link in the AI industry is not “use” but “return on investment”
There is evidence that AI services are actually growing. On its July earnings call, Microsoft said Azure had surpassed $100 billion in annual revenue and that Microsoft 365 Copilot had more than 30 million paid seats. Alphabet's Cloud revenue and operating income also rose sharply in Q2. However, that does not mean all Cloud revenue can be classified as AI revenue or that the returns on all AI infrastructure investment have been proven. Existing services and new AI businesses are combined in the figures.Microsoft Official Earnings Call, Alphabet SEC Earnings
Customers differ in their financial capacity. The following figures are from the latest quarterly reports released in July and August, not events that occurred today.
| Company / Reporting Period | Cash Flow and Investment Burden | What Memory Investors Should Note |
|---|---|---|
| Microsoft FY26 Q4 | OCF $55.4 billion, cash capital expenditures $35.8 billion, FCF $19.6 billion | A customer that retains cash even after substantial investment |
| Alphabet 2026 Q2 | OCF $39.069 billion, cash capital expenditures $44.924 billion, FCF −$5.855 billion | Cloud growth coexists with short-term cash burn |
| Meta 2026 Q2 | OCF $31.86 billion, capital expenditures $31.08 billion, FCF approximately $784 million | Investment plans were maintained and expanded, but the cash surplus is slim |
| Amazon 2026 Q2 | OCF $45.387 billion, net cash capital expenditures $53.076 billion, calculated FCF −$7.689 billion | AWS growth coexists with a substantial cash investment burden |
| CoreWeave 2026 Q2 | Revenue $2.575 billion, GAAP operating loss $49 million, net interest expense $640 million | Revenue growth alone does not eliminate financing risk |
OCF means operating cash flow, and FCF means free cash flow. Figures are not aggregated because company definitions differ; for example, Meta includes repayments of finance lease principal. Amazon's quarterly FCF was calculated from reported line items and is distinct from the company's stated TTM FCF of −$7.604 billion. All figures are company-wide and do not represent AI-only results. Sources: the Microsoft and Alphabet reports cited above, Meta, Amazon, CoreWeave.
Therefore, both extremes should be avoided. The claim that all Big Tech companies are short of cash and will halt investment, as well as the claim that every customer's investment is funded solely by quarterly operating cash flow, is inconsistent with the data. Some customers rely on cash reserves or external financing. CoreWeave's multiyear contract backlog is not a cash balance, and rising yields on its bonds in the secondary market do not mean the interest rates specified in existing bond agreements change immediately.
Power and permitting bottlenecks are connected to this financial burden. If data center completions are delayed, equipment installation and revenue generation are pushed back even if end-user demand for AI remains intact, extending the payback period on capital already invested. For memory companies, this may show up as deferred orders, customer inventory, or changes in payment terms. The IEA has also highlighted these physical and financial constraints. However, claims such as “65% of planned data centers have been canceled” were not included because the original report's denominator and actual cancellations could not be verified.IEA Analysis
5. The Future of Memory Cannot Be Explained by HBM Alone
Near-term results remain strong
Samsung Electronics' preliminary Q3 operating profit was KRW 107.4 trillion, up about 20% from KRW 89.5 trillion in Q2. Micron reported FY26 Q4 revenue of $54.229 billion and adjusted FCF of $33.199 billion, and forecast next-quarter revenue of $61.5 billion ± $1.5 billion. TSMC also reported actual Q3 revenue of $46.72 billion and guided for Q4 revenue of $52.3 billion–$54.3 billion.This is evidence against the claim that a full-scale collapse in AI orders has already begun. However, these figures are for different companies and products, and do not guarantee future margins for Korean memory makers. Samsung preliminary results, Micron results, TSMC results and outlook
Conditions vary by product. TrendForce forecasts 4Q conventional DRAM contract prices to rise +10~15% and NAND prices +15~20%, while also describing reduced purchases and capacity cuts for consumer products. Enterprise SSDs, by contrast, are benefiting from demand for AI inference, caching, and search. It is not enough to note that prices are rising; investors need to see whether this is due to reduced supply or whether actual purchase volumes are also increasing. Product-level survey
Long-term contracts improve visibility but do not guarantee a boom
Micron's 26 strategic customer agreements cover more than 35% of projected revenue through 2030. Pricing terms are set for about 75% of the revenue under those agreements, mainly through price floors and ceilings. This does not mean that 75% of total revenue is guaranteed at fixed prices. Minimum purchase volumes, cancellation rights, delivery schedules, and counterparty creditworthiness matter. The same terms cannot simply be assumed to apply to Samsung and SK Hynix contracts. Micron's official explanation
Also, forecasts of a sharp rise in HBM average selling prices in 2027 include a shift in product mix toward HBM4 and HBM4E. This does not mean that the price of the same product will rise uniformly by that amount. If customers change stack height or capacity per device due to cost, the relationship between bit demand, revenue, and margins will also change. TrendForce HBM outlook
Efficiency is both a threat and a condition for market growth
For AI handling long contexts and repeated tasks, the key is not only computation but also where information is stored and for how long. Rather than keeping all data in the most expensive memory, systems are evolving to use fast memory, DRAM, and SSDs in combination, while compressing caches to reduce costs.
In certain experiments, Google's TurboQuant substantially reduced KV cache memory. NVIDIA and Huawei have also proposed approaches that distribute context information across multiple storage tiers. However, an experiment compressing a particular cache by a factor of six is not evidence that total HBM orders will fall to one-sixth. Model weights, other workloads, and real-world service conditions differ. Google research, NVIDIA context memory architecture, Huawei product announcement
Even if improved efficiency enables more AI workloads on the same budget, investors should not assume that rising usage will necessarily outweigh memory savings. What investors should track is actual bit purchases, product mix, realized prices, and cash after investment—not model performance rankings. Even if some HBM demand shifts to SSDs, there is no guarantee that the margins on the resulting revenue will be the same.
Both exaggerating and ignoring Chinese competition are risky
In September, CXMT announced its G5 DRAM platform and mass production of 24Gb LPDDR5X. It is inaccurate to claim that competition remains only at the research-paper stage. However, an increase in dies per wafer compared with its own previous generation cannot be equated with a yield or cost advantage over Korean manufacturers. Announcing next-generation server products and obtaining customer qualification for HBM are also different stages. CXMT official announcement
In NAND, Counterpoint estimates YMTC's 2Q revenue share at 14%. This should not be read as its share of the highest-capacity enterprise SSD market, but competition cannot be dismissed as a distant prospect either. Replacing Korean products in China's domestic market alone can affect Korean manufacturers' sales opportunities and bargaining power. Counterpoint survey
Therefore, supply constraints over the next 3~6 months and normalized profit margins over the next 12~24 months should be assessed separately. Breaking ground on a factory or producing the first wafer is not the same as commercially viable mass supply, but share prices can discount future margins before actual capacity expansion is complete.
6. Samsung and SK Hynix: same industry, but not the same stocks
Samsung's mobile, foundry, and other businesses must be considered alongside memory earnings. The release of preliminary Q3 operating profit does not eliminate all risks ahead of the earnings announcement. Segment mix, net income attributable to owners of the parent, cash flow, and the outlook remain to be seen. Nor should a margin of safety be declared based solely on net cash and preliminary operating profit while ignoring competition in commodity memory.
In addition to its business competitiveness in HBM and server memory, SK hynix's capital allocation at Solidigm must be assessed separately. Q2 operating profit of 60.54 trillion won and HBM4 shipments are tangible evidence. However, net income includes a large gain related to investment assets, so annualizing it as though it were recurring earnings would distort valuation. SK hynix official results, Quarterly results
Solidigm's external financing is neither “a guaranteed loss if it goes public” nor “automatically positive because it receives growth capital.” For example, if a subsidiary valued at 100 receives 25 in new capital at fair value, the parent company's stake is worth 100 even if its ownership falls to 80%: 125×80%=100. Calculating a loss based only on the lower ownership percentage overlooks the cash received. Conversely, a loss could arise if the stake is sold at a low price or the proceeds are used for low-return investments. This is an illustrative assumption, not the terms of an actual transaction.
The company's October 1 statement was that it was reviewing various financing options and had made no decision. The issuance price, mix of new and existing shares, use of proceeds, transfer of debt, and cash attributable to the parent company therefore need to be confirmed. Parliamentary audits and regulatory discussions are important, but do not by themselves establish a new dilution rate or loss amount. Solidigm official statement
There is also countervailing evidence. On August 19, SK hynix approved a 40 trillion won share buyback and cancellation of all repurchased shares, and said it would seek to expand shareholder returns based on cumulative FCF. However, the scale and execution of any additional returns remain subject to board approval and further announcements. A shareholder return plan does exist, but a special dividend and a positive share-price reaction are not confirmed. Official shareholder return announcement
For these reasons, we have not adopted a default recommendation to sell only SK hynix first. Transaction uncertainty warrants further monitoring, but has not been established as an economic loss that calls for selling at the current price. Switching into Samsung would still leave exposure to the shared memory cycle. A switch trade requires separate reasons for selling SK hynix and buying Samsung.
7. Instead of saying the price is cheap, calculate the earnings required
This report does not force a new price target. That does not mean we are giving up on assessing the price. How much recurring earnings are needed to justify the current priceWorking backward reveals the assumptions behind the assessment.
The reference prices are the October 20 NXT closing prices: Samsung 209,000 won and SK hynix 1,190,000 won.
| Stock | Required annual EPS at assumed P/E of 4x | At assumed 6x | At assumed 8x |
|---|---|---|---|
| Samsung Electronics | 52,250 won | 34,833 won | 26,125 won |
| SK hynix | 297,500 won | 198,333 won | 148,750 won |
The formula is current price ÷ assumed multiple. The 4x, 6x, and 8x figures are arbitrary sensitivity assumptions, not validated fair multiples or probabilities. EPS means earnings per share. This table is neither a price target nor a trading threshold, and does not prove that the required earnings are sustainable.
We also calculated a reference estimate for core business earnings. Assuming a 25% tax rate and approximately excluding Q2 non-operating income, annualized core-business EPS is about 40,684 won for Samsung and 255,095 won for SK hynix. However, this assumes that Q2 pricing, shipments, and costs remain unchanged for four quarters—it is not normalized earnings or a forecast for the next 12 months.
| Earnings retention rates relative to the Q2 annualized core-business estimate | Samsung: current price/earnings | SK hynix: current price/earnings |
|---|---|---|
| 100% retained | 5.14x | 4.66x |
| 80% retained | 6.42x | 5.83x |
| 60% retained | 8.56x | 7.77x |
Samsung approximation: Q2 EPS 10,849×(operating profit 89.5/pre-tax profit 94.4)×[0.75/(1−tax 22.8/pre-tax profit 94.4)]×4. This assumes the allocation of existing EPS between common and preferred shares remains unchanged. Hynix: operating profit of 60.5426 trillion×0.75×4÷Q2 diluted weighted-average shares of 712 million. This is not a finalized calculation based on current share count and after-tax adjusted net income attributable to owners. Net cash was not added separately. Sources: Samsung Q2 materials, Hynix Q2 materials.
Samsung disclosed only preliminary operating profit for Q3, so pre-tax profit and net income attributable to owners could not be confirmed and were not included in this calculation. Operating profit was higher than in Q2.
This table makes two points. First, assuming a multiple of 4–8x, the current price may be justified even if Q2-level earnings do not last indefinitely. However, the validity of this multiple range has not been verified. Second, if earnings fall sharply and the market-applied multiple also declines, the share price could fall further. A low current P/E is not proof of a safety net; it is a prompt to analyze how long earnings can last. A 20% decline in product prices should not be treated as equivalent to a 20% decline in earnings.
8. What to do tomorrow and over the next 3–6 months
Tomorrow, distinguish between two things rather than focusing on whether the U.S. market rebounds
First, watch the price reaction. After weakness on NXT is reflected, see whether selling continues on KRX and whether any recovery holds. But do not add to core positions just because the price briefly recovers above the previous day's low or the opening price. The U.S. premarket tonight remains unsettled, and the regular-session outcome is not covered in this report.
If your account's risk limit has already been exceeded, reduce the excess without waiting to see whether the price recovers. If you are within the limit and the business conditions below have not changed, do not reverse your baseline holding decision based solely on price weakness.
Second, focus on business information. What matters more is whether customers' procurement or installation delays have actually affected memory orders, deliveries, and prices. Do not treat a rebound in one item, such as interest rates or share prices, as proof that business risks have been resolved. Conversely, if a material contract change or company guidance cut emerges, do not wait until the scheduled earnings announcement to respond.
Two conditions for switching from holding to reducing
The first is a downward revision to the outlook for recurring earnings from core memory. If original sources such as company or customer disclosures or official statements confirm reduced orders from major customers, meaningful delivery delays, or unfavorable changes to pricing or payment terms—and that information alone lowers estimates for next-quarter sales volume, margins, or cash collection in core server memory—reassess and reduce the position size. Do not wait for the company to officially lower its guidance; treat reports without an original source as items to verify.
However, do not make a single sentence about routine price renegotiations or a small consumer-product adjustment an immediate reason to sell everything. Consider product mix, shipments, and costs together. For increases in inventory or accounts receivable, distinguish among costs, production preparation, and payment timing.
The second is damage to the value attributable to Hynix's parent-company shareholders. Reduce Hynix if the transaction is unfavorable even after considering the cash inflow, debt, and investment impact from Solidigm's financing. Do not make an automatic decision based solely on a lower ownership stake or the absence of an announcement about additional shareholder returns.
If an irreversible transaction is executed without disclosure of the issue price, use of proceeds, or cash attributable to the parent company, reassess Hynix separately as a governance risk—even if no economic loss has been confirmed. This is a reason for immediate reassessment, not an automatic sell signal.
If a reason to reduce arises, do not attach an open-ended condition such as “I'll sell on the first rebound.” Waiting for a rebound could prevent you from acting on your decision. However, which stock to reduce and by how much should be determined based on its actual contribution to portfolio risk and liquidity conditions.
Do not treat the 3–6-month and 12–24-month periods as the same commitment
| Period / scenario | Assessment | Action |
|---|---|---|
| Base case over 3–6 months | Supply constraints and near-term earnings provide support. Customer investment burdens and share-price volatility remain. | Maintain existing holdings and hold off on adding. Consider adding only after evidence shows that recurring earnings under the adverse scenario can meet the earnings implied by the current price. |
| Improving scenario | Actual shipments and cash collections remain solid, while concerns about customer installations and procurement ease. | Reassess the price and risk budget at that time before considering an increase. Do not chase the stock on good news alone. |
| Deteriorating scenario | Key orders, realized prices, or collection terms weaken earnings estimates, or unfavorable capital allocation becomes concrete. | Reduce the position. Do not wait for actual losses to be reported or for the share price to recover to breakeven. |
| 12~24 months | The possibility that Chinese supply, capacity expansion, and efficiency gains could change normalized margins. | Do not assume today's super-boom will last forever. Reassess the investment thesis each quarter. |
This is a conditional action plan, not a model estimating probabilities. Even if the base scenario holds, there is no guarantee the share price will recover within that period.
SK hynix's official earnings call is October 28 at 09:00 KST, while Samsung's is October 29 at 10:00 KST. We will check product-level volumes and prices, inventory, accounts receivable and customer advances, cash after investment, shareholder returns, and transaction terms. This does not mean withholding all judgment until every item is available. We will incorporate important leading evidence first. SK hynix SEC filings, Samsung announcement
Re-entry into a position that has already been reduced should depend not on whether the previous sale price has been recovered, but on whether the reasons for reducing it have been resolved and on recurring earnings and cash flow relative to the price at the time. There is no new evidence that frequently switching between the two stocks produces higher returns than holding.
9. Lessons to carry forward from this assessment
Correcting mistakes does not mean moving price targets and buy levels lower every time the share price falls. The October 17 report noted that strong earnings do not guarantee a share-price recovery. This time, we have made that warning more specific through verifiable action criteria for recurring earnings and capital allocation. Even when investing in a strong industry, significant drawdowns can occur if views change on customer investment returns, supply expansion, and capital allocation.
We assessed the views of investment gurus by the same standards. We compared NVIDIA management's optimism about demand with actual products, shipments, and revenue, and did not use Dario Amodei's long-term technology outlook as evidence of current orders. Howard Marks's February memo was a useful perspective for distinguishing technological progress from investment returns. A famous person's optimism does not establish that the current price is fair. NVIDIA official earnings, Dario's long-term outlook, Howard Marks memo
The team conducted independent analyses and then challenged one another's arguments. The initial proposal to reduce SK hynix was revised because there was insufficient basis to translate unconfirmed risks into current economic losses. The optimistic claims of a price bottom and guaranteed shareholder returns were also rejected. The fact that the team ultimately converged on a hold recommendation was not counted as evidence for the investment. Normalized earnings, fair valuation multiples, and the company-specific earnings impact of Chinese competition remain unresolved.
The principle we aim to uphold in this report is simple: Hold while assessing the company's value, but protect the account's ability to survive first, and act before actual earnings deteriorate if important leading evidence changes. This means neither holding on at all costs nor changing trading criteria every day out of fear.
Analysis as of: Korea's KRX and NXT close on October 20, 2026, and the U.S. premarket at 21:47 KST. The U.S. Tuesday regular session and subsequent news are not included. After the intraday analysis, we cross-checked closing and premarket data. Company guidance, research firm estimates, and technology experiments were distinguished from realized results. We could not obtain the latest total forced-liquidation figures by stock, finalized transaction terms for Solidigm, or new normalized EPS and fair valuation multiples. We do not provide new price targets, probabilities of successful trades, or backtest results. No personal information is included. This is a conditional analysis based on public information and does not guarantee returns.
Comments 0
to comment Login