Neowiz Stock: Three Sell-Side Cuts in Ninety Days on One Reason

Neowiz stock traded at KRW 18,210 (about USD 12.89 at ₩1,413/$) on Thursday, August 20, 2026. Line up three sell-side prints from the last ninety days and the targets read KRW 26,000 (NH Investment & Securities, cut April 27), KRW 28,000 (Eugene Investment & Securities, Analyst Jung Eui-hoon, cut May 12), KRW 23,000 (Shinhan Investment Corp., cut early July). Three cuts in a ninety-day window, each by -25.71%, -20.00%, and -32.35%, with headlines that read “a bit more time to wait,” “product-cycle gap,” and “strong momentum when 2027-2028 tentpole titles arrive.” Three headlines pointing at one thing. That is where this writeup begins.

Contents15 min read

Where the three sell-side cuts converge

Read the three cut headlines side by side and they are one sentence. NH’s “a bit more time to wait,” Eugene’s “product-cycle gap,” and Shinhan’s “when 2027-2028 tentpoles arrive” all point at the Lies of P sequel. More specifically, the next soulslike console title from Round8 Studio (a Neowiz wholly-owned subsidiary). Three desks name it differently but the branch each one is standing on is the same branch.

Here is how I map what the multiple hangs on. By the end of the first half of 2026 the actually shipping products were Lies of P original (September 2023), the Overture DLC (June 2025), and the Switch 2 Complete Edition (August 6, 2026). What the sell-side is drawing for the second half of 2027 is one thing: the sequel. If that sequel ships on time and reproduces original-scale sales (company-official 4 million cumulative as of March 19, 2026, roughly 90% overseas revenue), the three sell-side valuations hold together. If it slips or under-runs the original, the three valuations collapse together. What I would be buying here is not a games company; it is the probability of reproducing sequel unit sales.

The Neowiz stock multiple right now: P/E 7.96, P/B 0.67

Kiwoom-desk screen basis: P/E 7.96, P/B 0.67, market cap KRW 393.99 billion, roughly USD 0.279 billion (KRW 18,210 x 21,636,463 shares = KRW 393,999,991,230, checked to the won). FY2025 owners’ net income KRW 49.98 billion, revenue KRW 432.73 billion (YoY +17.97%), operating profit KRW 59.98 billion (13.86% operating margin). The screen ROE of 9.80% is not heavy for the KOSDAQ gaming and entertainment group average. On the surface this reads as a cheap name and a fair valuation for a Korean small-cap. The reason not to stop reading at that face is in those three sell-side prints from earlier.

FY2024 came in at KRW 366.81 billion in revenue and KRW 32.92 billion in operating profit, but consolidated net income was a KRW 6.95 billion loss while owners’ net income was a KRW 2.39 billion profit, meaning non-controlling losses were large. FY2023 posted revenue of KRW 365.59 billion, operating profit of KRW 31.59 billion, and owners’ net income of KRW 48.68 billion. Revenue was flat for three years and then popped in 2025 as the Lies of P original (released September 19, 2023) and the Overture DLC (released June 7, 2025) reignited paid installs. The screen multiple stands on that pop.

Where I walked the Neowiz stock math differently from the sell-side

My arithmetic ran like this. Q1 2026 quarterly revenue KRW 101.42 billion and operating profit KRW 6.95 billion (6.85% operating margin); Q2 2026 quarterly revenue KRW 102.07 billion and operating profit KRW 7.94 billion (7.78% operating margin). H1 2026 cumulative revenue KRW 203.49 billion (YoY +2.23%) and operating profit KRW 14.89 billion (YoY -48.31%). Revenue barely moved, but operating profit dropped by more than half against the year-ago half. A profit line falling behind its own revenue line is a shape I walked through in PharmaResearch, where Q2 2026 profit grew slower than sales. The Overture DLC piled into June 2025, and there is no equivalent event in the first half of 2026.

The KRW 40.9 billion 2026E operating-profit estimate that Eugene’s Jung Eui-hoon carried in her May 12 note sits in this same spot. Against FY2025’s KRW 59.98 billion, that is -31.81%. With KRW 14.89 billion already recorded through H1, H2 needs KRW 26.01 billion to close the gap; H2 2025 (Q3 KRW 26.47 billion + Q4 KRW 4.70 billion = KRW 31.17 billion) would have to be nearly matched, and the only fresh product pushing into that H2 is Switch 2 Complete Edition early revenue. That is the thin spot the three desks priced into their cuts.

Three ninety-day events that Neowiz stock has to carry

1) Switch 2 Complete Edition global release (August 6, 2026)

Base game + Overture DLC + performance tuning, bundled. Physical package (iam8bit collaboration) follows October 2, 2026. This isn’t the sequel; it’s one more sales window for the original. The sell-side did not use this event as valuation ground, and all three cut notes wrote instead that “the sequel has to come” underneath.

2) Developer-CEO transition (August 12, 2026)

Lies of P development head Vice President Park Sung-jun was named co-CEO alongside Bae Tae-geun. This is Neowiz’s first developer-CEO structure. With the sequel and six projects on the roadmap, putting the IP creator on the executive line reads as a signal, and it reads that way naturally. The counterweight is on the same screen: Q2 2026 results announced the very next day showed revenue -6.6% YoY and operating profit -57.3% YoY.

3) Q2 2026 results (announced August 12, 2026)

Quarterly revenue KRW 102.07 billion (-6.6% YoY) and operating profit KRW 7.94 billion (-57.3% YoY). By segment, PC and console fell -35% YoY on the prior-year Overture base effect while mobile rose +21% YoY on BrownDust2 third-anniversary and the Cats & Soup Sanrio collaboration. Quarterly net income was KRW 18.58 billion (+44% YoY), inflated by financial-asset valuation gains. H1 owners’ net income landed at KRW 34.79 billion (+63.44% YoY).

Line the three events up and what happened in these ninety days is: the revenue-adding event is an extension of the original sales window, the P&L slipped to the right, and the org realigned toward the sequel. The three time axes underneath “buy now” do not point at the same date.

Neowiz stock hangs on one sequel: why the three desks cut

The sequel framework Kiwoom Securities’ Kim Jin-gu drew in his July 30, 2024 note still gets used as a market benchmark. On top of assumed 2 million cumulative original units through Q2 2027, the frame carries sequel Q3 2027 release, unit sales +50% versus original, price +25%. Two things have moved since that frame was set.

First, official unit sales ran faster than modeled. On March 19, 2026 the company confirmed original + DLC combined had crossed 4 million units (official X account, overseas revenue roughly 90%). Sensor Tower’s separate count reads “11 million cumulative units, 2.7 million in the five months after the DLC, 0.43 million MAU,” but the metric definition gap versus company-official 4 million (download and bundle inclusion possible) means I list both and lean on neither. Second, the sequel schedule per company IR sits at “five of six projects in full production, sequel through prototype and vertical slice.” Unreal Engine 5 adoption is confirmed. No dated release announcement.

If the sequel firms up as Q3 2027, the three sell-side desks’ valuation ground lines up. If it slides to “first half of 2028 or later,” the three cuts come again. I hold that first branch at lower probability than the three desks do; big soulslike console titles rarely ship on their first announced window.

Capital return and three share-count series: things I looked at

The January 23, 2026 capital-return framework

20% of operating profit, KRW 10 billion minimum per year (KRW 5 billion treasury buyback and cancellation + KRW 5 billion cash dividend), guaranteed across 2025-2027. The war chest is the KRW 50 billion pulled from capital reserves through an extraordinary shareholder meeting. FY2025 declared DPS was KRW 303 (up 23.7% YoY, 1.59% dividend yield), and total cash dividend was KRW 5.99 billion. FY2024 paid no dividend (consolidated net loss of KRW 6.95 billion). FY2023 DPS was KRW 245 with a KRW 4.99 billion total. This registers as a resumption; a streak it is not yet.

Three share-count series (a §111 recurrence)

The share count for this company comes out differently on three different screens. The market-cap engine on the metrics server implies 21,636,463 shares (which reconciles to the KRW 393.99 billion market cap), the paid-in-capital screen at KRW 11.0 billion divided by the KRW 500 face value implies 22,000,000 shares (which reconciles to paid-in capital), and DART’s cash-dividend total of KRW 5.99 billion divided by DPS 303 implies 19,755,776 shares (which reconciles to dividend-eligible float). Each figure ties out on its own screen. Given treasury shares of 2,069,446 (WISEfn; 9.41% of the 22 million issued), the gap between dividend-eligible 19.76 million and issued 22 million is close to the treasury series, and the metrics-server 21.63 million sits at a mid-point date. I could not pull the DART semi-annual filing receipt number myself, so I could not close the timing on all three series. In my arithmetic the sell-side upside percentages, dividend yield, and P/E ratio all drift in the third decimal depending on which of the three you use. I ran the multiple math on 21.63 million (market-cap-consistent) but wrote dividend-related lines against the DART cash-dividend total directly.

Things I did not use

The metrics-server `ebitda` field returned KRW 6.95 billion, matching Q1 2026 quarterly operating profit of KRW 6.95 billion to the won, meaning the depreciation add-back fell through and the field simply echoed operating profit. This is the same vendor-fallback trap that has now shown up five sessions in a row across sectors (DART’s fnlttSinglAcntAll response carries no depreciation account for many issuers), and it just recurred in gaming. Dividend-payout vendor field of 11.9 versus `_raw` 52.2 also excluded (the underlying ratio basis is opaque). Interest-coverage 2.59 (Q1 2026) excluded because I cannot pin the interest column. `total_equity` at KRW 555.81 billion versus DART’s owners’ equity of KRW 531.16 billion (a 4.64% gap that reflects non-controlling interests of about KRW 24.65 billion) sent me back to DART for the by-hand calculation. `revenue_cagr_3y` at 13.68 is distorted by an FY2022 low base. The benchmark tool timed out during this session, so I did not run the market-versus-sector excess-return decomposition here; §88 breaks its six-session streak and moves to the next writeup.

Neowiz stock in its recovery band: -42.28% below 250-day high

KRW 18,210 sits -42.28% below the 250-day high of KRW 31,550 and +13.11% above the 250-day low of KRW 16,100. Three-month return is -12.45%, six-month is -30.89%, twelve-month is -29.42%. The 20-day moving average (KRW 18,609) is -2.14% above spot, and the 120-day moving average (KRW 21,128) is -13.81% above spot. The last close sits slightly off the trough. Reaching the KRW 25,667 sell-side average requires +40.95%, and the ground under that +40.95% is the one sequel I outlined above.

Why no peer table for Neowiz stock in this writeup

Among Korean listed console-IP names the field is Krafton (259960, PC and mobile service model), Nexon (Tokyo listing), Shift Up (462870, Stellar Blade), and Pearl Abyss (263750, Black Desert plus Crimson Desert pending). But Krafton and Nexon are 10 to 30 times larger service-game businesses, Shift Up carries a single-console-title valuation stacked directly on top so the weight is heavier than this company, and Pearl Abyss’s Crimson Desert release-slippage arc mirrors this company’s sequel risk closely enough that lining the two side by side would mix the causal variables. I did not build a peer table in this writeup, and I am leaving that fact on the page: no Korean listed company is a clean pair for a name working through “the gap after a single hit IP.” The one Korean gaming name I have split open from the balance-sheet side rather than the IP side is Netmarble, whose debt fell while its financing cost doubled.

What the three desks missed and what I also could not lock down

1. The metric definition of Sensor Tower’s 11 million cumulative: I could not pin the gap versus company-official 4 million. Download and bundle inclusion is plausible, but there is thin ground to declare either the right one.
2. NH, Eugene, and Shinhan reports read in the primary: I sourced the target and rating headlines through Tier-2 Korean press coverage. I did not read the reports themselves, so I did not confirm the analyst names at NH and Shinhan or the underlying estimate tables.
3. DART semi-annual filing access failed for this session: I did not sight the receipt number or the explicit Neowiz Holdings ownership share. That is why the three share-count series’ timing did not close.
4. Kingdom 2 China service permit (banho): 2026-H2 service is planned, but I could not confirm banho issuance. If issued, it adds to H2 numbers; if not, it does not.
5. Current webboard (Pimang) revenue share: A namuwiki reading puts it above 40%, but company IR segments come out three-way (PC and console / mobile / other), so I could not verify separately. This writeup’s spine is the sequel, so this number does not swing the thesis; it just does not belong in a table without ground.
6. The share-count basis under NH’s April 27 KRW 26,000 valuation: I did not confirm whether it uses pre-cancellation or post-cancellation shares. On a name with three share-count series diverging, this is not a third-decimal question but a direction-drift question.
7. Round8 Studio sequel development budget: With six projects in parallel per company IR, I did not sight the head-count and development-cost allocation table.
8. The org context behind VP Park Sung-jun’s co-CEO appointment: whether this reads as a promotion or as workload relief for Bae Tae-geun did not confirm. I left the developer-CEO structure as a fact and the read as an open question.
9. Neowiz Holdings (042420) treasury-share purchase of KRW 10 billion on May 19, 2026: this is a holdco event and not a 095660 event, so I set it aside for this writeup. That said, holdco policy filtering into 095660 valuation indirectly is a thread that remains.

Nine ways my thesis on Neowiz stock could be wrong

1. The three desks all sit at “target cut, rating kept” and nobody flipped to sell. That is a timing disagreement about when to re-rate, and not a direction disagreement.
2. Actual 095660 treasury-share cancellation activity between Q1 and Q2 2026 could not be sighted. If the KRW 5 billion minimum from the January policy has already been executed, the three-series gap has a different origin than I framed.
3. A sequel arriving earlier than expected (Q1 or Q2 2027) would make the three sell-side valuations look conservative. I may be under-modeling this company’s development pace.
4. Mobile (BrownDust2) and webboard (Pimang) cash flow works as a defense line against sequel risk. FY2025 operating cash flow of KRW 76.07 billion and free cash flow of KRW 75.15 billion prove that defense line is thick.
5. If Sensor Tower’s 11 million count is closer to the truth, the brand-strength stacking onto the sequel is larger than a company-official 4 million reading would suggest.
6. If Switch 2 Complete Edition early sales run past consensus, H2 2026 results could open above Eugene’s KRW 40.9 billion.
7. The VP Park Sung-jun co-CEO transition may pull sequel decision speed forward. That signal may still be under-reflected in the multiple.
8. If Neowiz Holdings’ (042420) KRW 10 billion buyback compounds into continued open-market purchases, 095660 gets an indirect downside support.
9. The most painful counter: I am chained to those three sell-side reports. If the three prints are one desk’s consensus copied by three, if the overlap of the three headlines is not three independent observations but three echoes of the same view, then I am counting one input three times.

Stance and three thesis-break conditions

No position. No order. A market cap of KRW 393.99 billion sits well outside the KOSDAQ top 100 and belongs in a Type-1 hold-and-watch bucket. What this company is to me: a name whose multiple, priced up by one hit IP, now hangs on a single next game’s reproducibility.

Any one of these three conditions triggers thesis-break.

1. If company IR officially references a sequel release date in first-half 2028 or later, the three sell-side valuations’ timing ground shifts, and the current multiple cannot be held up by buybacks and dividends alone through that gap.
2. If Q3 2026 operating profit (2026-11-15 semi-annual filing) prints under KRW 10 billion, the H2 sum of KRW 26.01 billion needed to reach Eugene’s KRW 40.9 billion becomes untenable, and the valuation framework goes with it.
3. If within six months of Park Sung-jun’s co-CEO appointment the sequel does not advance from prototype to alpha in company IR, the developer-CEO structure would not have moved development pace, and the “late 2027” assumption behind the sell-side valuations opens to the downside.

Q&A on Neowiz stock: six questions

Q1. Is the current multiple cheap?

On the screen ratios of P/E 7.96 and P/B 0.67, yes. But the numerator of that ratio, FY2025 owners’ net income of KRW 49.98 billion, leans heavily on the Overture DLC launch of June 2025, and H1 2026 operating profit already came in more than -48% year on year. Holding the current multiple requires the sequel to land, which is what the three desks priced into their cuts.

Q2. Why do the three sell-side targets (26,000, 28,000, 23,000) with a KRW 25,667 average count as ground for me?

Three reports written at different dates (April, May, July) using different frameworks land inside a narrow target band (KRW 23,000-28,000), with cut sizes clustered in the -20 to -32% range, and headlines all pointing at the same event. “Three near-independent observations converging” builds one level of ground. But if those three are one desk’s consensus echoed three times, this is one observation and not three, which is the tenth item under my counter-list.

Q3. Is this a dividend name?

The company set up a capital-return framework on January 23, 2026 (KRW 10 billion minimum per year, three-year guarantee) and reactivated its dividend at FY2025 DPS of KRW 303. That said, a 1.59% dividend yield does not sit in the dividend-stock band, and the FY2024 DPS-zero history (loss year) is on the record. The dividend belongs in the frame as a defense reserve for the sequel thesis and it is not the thesis itself.

Q4. When does the Lies of P sequel arrive?

No official date from the company. Kiwoom Securities’ Kim Jin-gu, in his July 30, 2024 note, modeled a Q3 2027 release with unit sales +50% versus the original and price +25%, and the frame has since been used as a market benchmark. The recent company IR line reads “prototype and vertical slice cleared.”

Q5. How does 095660 sit within the Neowiz group?

Neowiz Holdings (042420) is 095660’s legal largest shareholder, and Chairman Na Sung-kyoon’s line effectively controls 042420. On May 19, 2026 042420 resolved a KRW 10 billion treasury-share purchase for cancellation. In this writeup I categorized that as a holdco-level event and not a 095660 event, but holdco policy filtering into 095660 valuation through indirect channels remains a thread.

Q6. US access — can foreign investors buy this name?

There is no Neowiz US ADR. The stock is KOSDAQ-only. Foreign ownership sits at 21.38% and the credit-margin balance at 1.10%. There is also no US-listed ETF that gives clean single-name exposure to Korean small-cap console publishers, so this is a name a US-based reader would have to access through a KRW-settled brokerage on the KOSDAQ. That is not a service note; it is a fact this writeup carries because the sequel-priced multiple is only reachable through that route.

Prices and multiples reflect the Thursday, August 20, 2026 close as checked at the time of writing (Friday, August 21, 2026, around 5:00 KST, pre-market). This piece may publish on a delay, so figures can differ from live quotes; please treat the numbers as approximations. USD conversions are approximate, at roughly KRW 1,413 per dollar on the same reference date. Korean won is the reference currency throughout. Dates for the three sell-side reports (NH April 27, Eugene May 12, Shinhan early July) and for company events (Q2 2026 results August 12, developer-CEO appointment August 12, Switch 2 launch August 6) follow the cited press wording. No fresh flow-of-numbers check should be expected before the Q3 2026 semi-annual filing due November 15, 2026.

External sources: Digital Daily Q2 2026 results and co-CEO transition; Newspim operating profit KRW 7.94 billion; Bloter Neowiz undervalued gaming stock July 9, 2026; NewsQuest Eugene Investment note by Jung Eui-hoon; Nate press on NH KRW 26,000 target; Insight Korea on Kiwoom Kim Jin-gu Lies of P sequel Q3 2027 frame; Inven Lies of P 4 million milestone March 19, 2026; Sensor Tower Lies of P 11 million cumulative count; Seoul Economic Daily KRW 10 billion capital-return policy over three years; TopDaily dividend resumption January 23, 2026; Reuters Asia-Pacific coverage.