Flow Analysis on Korea Invest Insights

Smart Money Is Returning to KOSDAQ: Prices Still Look Weak, but Flows Are Moving First

Fri, 22 May 2026 09:55:00 +0900

📚 Related Series Pearl Abyss Flow Bottom Test / Pearl Abyss Hub / May 17 Market Snapshot / Complete KOSDAQ Guide

The past two months were a decisive KOSPI large-cap win and an extreme KOSDAQ relative slump. On a market-cap-weighted proxy, KOSPI gained +49.8% while KOSDAQ gained +11.4%, leaving KOSDAQ -38.4 percentage points behind KOSPI. But flows are starting to move differently from prices. Over the past five sessions, KOSDAQ recorded +KRW 707.9bn of foreign-plus-institutional buying and +KRW 410.4bn of program buying. KOSPI, over the same period, recorded -KRW 14.058tn of foreign-plus-institutional selling and -KRW 8.157tn of program selling. Price says KOSPI won. Fresh money is starting to look back at KOSDAQ.

Key Takeaways

Two-month price action is a KOSPI victory. On a market-cap-weighted proxy, KOSPI returned +49.8% versus KOSDAQ +11.4%. Over the past 20 sessions, the gap was also extreme: KOSPI +27.0%, KOSDAQ -0.1%.
Equal-weight returns tell a different story. On a two-month equal-weight basis, KOSPI returned +14.5% and KOSDAQ +17.4%. The breadth of the overall KOSDAQ universe has not collapsed. The market was pulled upward by mega-cap AI infrastructure names in KOSPI.
The KOSDAQ/KOSPI relative ratio is near a two-month low. The current relative ratio is 74.1, down -29.5% from the two-month high of 105.1. The 40-session standard score is -1.88σ, which is already an oversold zone in relative-strength terms.
Flows are improving first in KOSDAQ. Over the past five sessions, KOSDAQ foreigners plus institutions bought +KRW 707.9bn, while program flows added +KRW 410.4bn. KOSPI, by contrast, continued to see foreign-plus-institutional and program selling.
The institution mix is different. KOSPI institutional buying is mostly financial-investment flow, which can include index, derivatives, and arbitrage activity. KOSDAQ has smaller but visible investment-trust buying of +KRW 124.9bn.
Pearl Abyss ranks 7th among the top 100 KOSDAQ names by drawdown. It fell -41.3% from its two-month high of KRW 77,400 to KRW 45,450, roughly 19 percentage points deeper than the top-100 median drawdown of -22.0%.
Pearl Abyss looks closer to an investable rebound candidate than a simple oversold bounce. It has 2026E OP of KRW 476.9bn, 2026E PER of 8.0x, 2026E ROE of 36.9%, five-session foreign-plus-institutional buying of +KRW 22.08bn, and individual selling of -KRW 22.33bn.

The conclusion is simple. KOSDAQ has not yet been confirmed as the new market leader. But flows are starting to turn before prices confirm it. Within that setup, Pearl Abyss is one of the most interesting oversold large-cap candidates.

1. Data Basis and Caveats

The numbers in this post are not official KRX index values. They are market-cap-weighted and equal-weighted proxy series built from individual-stock data in the Research OS local database.

Item	Basis
Data	Research OS local DB
As of	2026-05-21
Analysis period	2026-03-21 to 2026-05-21
Trading sessions	42 sessions
Index method	Individual-stock-based market-cap-weighted proxy
Caveat	Not official KRX index data

The purpose is not to replicate the official index perfectly. The purpose is to find the point where relative price strength and flow direction diverge.

2. Price Says KOSPI Large Caps Won

Looking only at price over the past two months, the conclusion is clear: KOSPI large caps won decisively.

Category	KOSPI	KOSDAQ	Interpretation
2M market-cap-weighted return	+49.8%	+11.4%	KOSPI wins decisively
2M equal-weight return	+14.5%	+17.4%	Breadth is not bad for KOSDAQ
Recent 20D return	+27.0%	-0.1%	KOSDAQ was fully sidelined in the past month
Recent 5D return	+4.8%	-0.5%	KOSPI still leads short term
2M average advancer ratio	43.1%	41.5%	Breadth is weak in both markets

The key is the gap between market-cap-weighted and equal-weighted returns. On a market-cap-weighted basis, KOSPI crushes KOSDAQ. But on an equal-weight basis, KOSPI is +14.5% and KOSDAQ is +17.4%.

That means the entire KOSPI universe was not uniformly strong. It means mega-cap AI infrastructure names such as Samsung Electronics, SK Hynix, and Samsung Electro-Mechanics pulled the KOSPI index upward.

By contrast, KOSDAQ looked weak at the index level, but the broader stock universe did not completely break down. The essence of KOSDAQ weakness was not that every stock was worse. It was that market-cap leadership and flow leadership were captured by KOSPI mega caps.

3. KOSDAQ/KOSPI Relative Strength Is Oversold

If we set the KOSDAQ/KOSPI relative ratio to a base of 100, the current position becomes clearer.

Item	Value
Relative ratio at start of 2M period	99.7
2M high	105.1, 2026-03-31
Current	74.1
Versus high	-29.5%
2M change	-25.6%
Versus 20D average	-12.5%
40-session standard score	-1.88σ

KOSDAQ is now at its weakest relative level versus KOSPI in the past two months. A 40-session standard score of -1.88σ is statistically a fairly deep relative-weakness zone.

But that does not automatically mean “buy KOSDAQ now.” Relative-strength oversold conditions show rebound potential, not rebound confirmation.

The current setup is more accurately described this way:

Price: KOSDAQ at a relative low
Flows: early improvement signs
Trend: not yet confirmed
Conclusion: rebound preparation is possible, but price confirmation is still needed

KOSDAQ has entered a zone that looks cheap in relative terms. It is still too early to say the market has started buying it aggressively.

4. Flows Are Looking at KOSDAQ Before Prices Confirm It

The full two-month flow picture is not clean in either market.

Unit: KRW 100mn

Market	Individuals 2M	Foreigners 2M	Institutions 2M	Foreigners+Institutions 2M	Program 2M
KOSPI	+252,628	-555,191	+177,042	-378,149	-293,401
KOSDAQ	+4,765	+19,909	-36,542	-16,632	+6,300

Over two months, foreigners plus institutions were negative in both markets. But the quality is different.

In KOSPI, foreigners and program flows sold heavily, while individuals and financial-investment accounts absorbed the supply. The index was strong, but the quality of the flow was not particularly clean.

In KOSDAQ, institutions sold, but foreigners and program flows were net buyers. Even though KOSDAQ prices were weak, foreigners and program trading did not fully abandon the market.

The more important signal is the recent 20-session and five-session change.

Market	Foreigners+Institutions 20D	Program 20D	Foreigners+Institutions 5D	Program 5D
KOSPI	-323,604	-219,789	-140,578	-81,572
KOSDAQ	+876	+5,182	+7,079	+4,104

This is the core of the post.

Prices are overwhelmingly stronger in KOSPI, but recent flow improvement is better in KOSDAQ.

Over the past five sessions, KOSDAQ recorded +KRW 707.9bn of foreign-plus-institutional buying and +KRW 410.4bn of program buying. KOSPI continued to see foreign-plus-institutional and program selling.

Price is the result of the past. Flow is a hint about the next move. KOSDAQ is weak in price terms, but in flow terms it is starting to look like a market testing a bottom.

5. Institution Breakdown: KOSPI Is Financial Investment, KOSDAQ Shows Investment Trusts

The internal institutional breakdown sharpens the contrast.

Market	Financial Investment 5D	Insurance 5D	Investment Trusts 5D	Private Funds 5D	Pension/National 5D
KOSPI	+45,340	-623	+1,143	+3,299	-2,409
KOSDAQ	+177	+4	+1,249	+205	-690

Most KOSPI institutional buying came from financial-investment accounts. That flow can include index, derivatives, and arbitrage activity. It is risky to read it as long-duration directional capital.

KOSDAQ is smaller in absolute size, but it shows investment-trust buying. That matters when looking for rebound candidates among oversold growth stocks. Investment trusts can be the first capital to touch KOSDAQ mid- and large-cap growth baskets again.

Pensions and insurance accounts have not yet joined. So this is not a confirmed flow turn. It is an early turn. But the interpretation that KOSDAQ has been completely abandoned is wrong.

6. Pearl Abyss: 7th-Largest Drawdown Among Top 100 KOSDAQ Names

Now narrow the lens to Pearl Abyss. Among the top 100 KOSDAQ names, Pearl Abyss ranks 7th by drawdown from its two-month high.

Item	Basis
Data	Research OS local DB
Price date	2026-05-21
Analysis period	2026-03-21 to 2026-05-21
High price basis	Intraday high over the two-month window
KOSDAQ top 100	Market-cap proxy based on foreign-held shares and foreign ownership ratio
Coverage	1,688 KOSDAQ names; drawdowns calculated for top 100

Pearl Abyss’s two-month high was KRW 77,400 on April 1, 2026. Its closing price on May 21, 2026 was KRW 45,450.

Drawdown from high = 45,450 / 77,400 - 1
= -41.3%

The drawdown distribution for the top 100 KOSDAQ names is as follows.

Category	Drawdown from high
Average	-22.3%
Median	-22.0%
Bottom quartile threshold	-29.5%
Bottom decile threshold	-34.3%
Down more than -20%	57 names
Down more than -30%	23 names
Down more than -40%	7 names

At -41.3%, Pearl Abyss sits exactly inside the seven-name group down more than -40%. This is not merely “somewhat oversold.” It is one of the deepest large-cap drawdowns in KOSDAQ.

7. Comparing the Largest Drawdowns

The top 10 drawdowns among the top 100 KOSDAQ names are:

Rank	Name	Market-cap rank	2M high	Current	Drawdown
1	SCD Pharm	7	1,233,000	355,000	-71.2%
2	Orum Therapeutics	82	124,800	65,500	-47.5%
3	Rznomics	63	245,250	130,000	-47.0%
4	Seers Technology	88	63,200	34,300	-45.7%
5	Woori Technology	39	29,500	16,700	-43.4%
6	ABL Bio	11	189,400	109,900	-42.0%
7	Pearl Abyss	38	77,400	45,450	-41.3%
8	LigaChem Biosciences	15	224,000	139,500	-37.7%
9	Sampyo Cement	84	19,690	12,780	-35.1%
10	Curiox Biosystems	81	126,800	82,500	-34.9%

Pearl Abyss has two distinguishing features in this table.

First, it fell alongside biotech-event names, but it is a different animal. SCD Pharm, ABL Bio, LigaChem, Peptro, and HLB are largely clinical, licensing, or regulatory event options. Even when flows improve, valuation is difficult to anchor on earnings.

Second, Pearl Abyss can be discussed with earnings numbers. It has 2026E OP of KRW 476.9bn, 2026E PER of 8.0x, and 2026E ROE of 36.9%. The debate is not “is this an option with no earnings?” The debate is “are the earnings estimates right?”

That is why Pearl Abyss is different within the oversold group. It is not a biotech-event rebound candidate. It can become an investable rebound candidate if earnings and flows confirm together.

8. Pearl Abyss Among the Top 20 Oversold Names

Looking again at the key oversold names through earnings and flows makes Pearl Abyss’s position clearer.

Rank	Name	From high	26E OP	26E PER	26E ROE	5D Foreigners+Institutions	20D Foreigners+Institutions	5D Individuals	5D Program
1	SCD Pharm	-71.2%	450	[Unconfirmed]	[Unconfirmed]	+228.8	+281.4	-42.0	[Unconfirmed]
4	Seers Technology	-45.7%	607	22.6x	86.1%	+114.5	+22.2	-121.7	+172.3
5	Woori Technology	-43.4%	[Unconfirmed]	[Unconfirmed]	[Unconfirmed]	+368.5	+9.6	+6.1	+190.1
6	ABL Bio	-42.0%	-486	Loss-making	-26.4%	+564.0	-1,026.5	-556.5	+779.1
7	Pearl Abyss	-41.3%	4,769	8.0x	36.9%	+220.8	-73.8	-223.3	+235.6
8	LigaChem Biosciences	-37.7%	268	116.1x	8.1%	-860.4	-815.0	+821.1	-149.4
13	Peptro	-33.2%	[Unconfirmed]	[Unconfirmed]	[Unconfirmed]	-263.9	-958.8	-70.3	[Unconfirmed]
15	Oscotec	-32.7%	340	46.5x	16.6%	-76.4	-237.1	+80.7	+7.5
19	HLB	-31.7%	[Unconfirmed]	[Unconfirmed]	[Unconfirmed]	-123.1	-664.5	-10.5	-18.1

Pearl Abyss has the best balance in this table.

The drawdown is deep enough.
Valuation pressure is low at 8.0x 2026E PER.
It has an earnings base with 2026E OP of KRW 476.9bn.
The five-session flow shows a handoff structure: foreigners plus institutions +KRW 22.08bn, individuals -KRW 22.33bn.
But 20-session foreign-plus-institutional flow is still -KRW 7.38bn, so this is not yet confirmed accumulation.

Pearl Abyss is therefore not a stock where the reversal is already complete. It is a stock where we need to verify whether the early reversal is real.

9. Pearl Abyss Detailed Flow

Pearl Abyss’s recent five-session flow looks different from the prior two weeks.

Investor group	5D net buying
Individuals	-KRW 22.33bn
Foreigners	+KRW 20.77bn
Institutions total	+KRW 1.32bn
Financial investment	-KRW 1.61bn
Insurance	+KRW 0.01bn
Investment trusts	+KRW 0.30bn
Private funds	+KRW 2.62bn
Pension/National	+KRW 0.01bn
Other corporations	-KRW 0.04bn
Other foreigners	+KRW 0.29bn

The good part is clear. Individuals sold, while foreigners and private funds absorbed the supply. That is exactly the kind of structure investors want to see first in an oversold bottom test.

But it is not yet a complete turn. Over 20 sessions, foreigners plus institutions are still -KRW 7.38bn. The past five sessions improved, but cumulative flow has not fully turned.

The wording needs to be precise.

The flow turn is not confirmed yet.
But early signs of a flow turn have appeared.

If the May 18 note framed Pearl Abyss as a “flow bottom test,” the May 21 data updates that judgment by one step. The key now is whether the bottom test continues without failing.

10. Comparative Judgment

Pearl Abyss vs Oversold Biotech Names

Pearl Abyss is different from SCD Pharm, ABL Bio, LigaChem, Peptro, and HLB.

Most oversold biotech names are clinical, licensing, or regulatory event options. Even when flows are strong, valuation is hard to anchor on earnings. ABL Bio has strong five-session flow, but 2026E operating profit is -KRW 48.6bn. LigaChem trades at 116.1x 2026E PER and has -KRW 86.04bn of foreign-plus-institutional selling over the past five sessions.

Pearl Abyss is different. It has 2026E OP of KRW 476.9bn, 2026E PER of 8.0x, and 2026E ROE of 36.9%. So the core question is not “can it succeed?” but “how much of the 2026 earnings base can be sustained into 2027?”

Pearl Abyss vs Seers Technology

Seers Technology is the most interesting comparison. It has 2026E OP of KRW 60.7bn, 2026E PER of 22.6x, 2026E ROE of 86.1%, and five-session foreign-plus-institutional buying of +KRW 11.45bn. Its -45.7% drawdown also gives it rebound torque.

But institutions are still weak overall, and foreigners are absorbing most of the supply. Pearl Abyss has +KRW 20.77bn of foreign buying and +KRW 2.62bn from private funds. Its valuation is much lower.

On a risk-adjusted basis, Pearl Abyss > Seers Technology. But both still require additional confirmation of flow quality.

Pearl Abyss vs Woori Technology and Sungho Electronics

Woori Technology and Sungho Electronics have strong flows, but weaker earnings and consensus anchors. Woori Technology has +KRW 36.85bn of five-session foreign-plus-institutional buying, but earnings-based comparison is blocked. Sungho Electronics may have strong short-term institutional buying, but if foreign selling and individual buying are mixed in, investors need to watch for distribution.

Pearl Abyss may have weaker flow intensity than some thematic names. But it is more investable because the drawdown can be explained with earnings numbers and valuation.

11. Practical Judgment

The current market can be summarized as follows.

Item	Judgment
Price	KOSPI large-cap leadership, KOSDAQ at relative lows
Flow	KOSPI price is strong, but flow quality is weak
KOSDAQ	Price is weak, but foreign, program, and investment-trust flows are improving
Meaning	KOSDAQ rebound setup is forming, but price confirmation is needed
Pearl Abyss	Top investable oversold candidate, but flow turn is not confirmed

There are three entry triggers for KOSDAQ.

KOSDAQ/KOSPI relative ratio rebounds for three consecutive sessions.
KOSDAQ foreign-plus-program buying continues.
Oversold KOSDAQ large caps repeatedly show individual selling absorbed by foreigners and institutions.

Pearl Abyss has four separate triggers.

Defense near KRW 45,000.
Continued foreign and private-fund buying.
Recovery above KRW 47,200.
Recovery in 2Q guidance or long-tail/DLC expectations after IR/NDR.

The invalidation conditions are also clear.

Condition	Interpretation
Break below the KRW 44,000 zone	Recent flow turn likely failed
Foreign five-session buying flips back to selling	Low-price re-entry was a one-off defense
KOSDAQ/KOSPI relative ratio fails to rebound	Market-wide rotation is delayed
2Q guidance or long-tail thesis weakens	Fundamental thesis is discounted again

12. Final Line

The past two months were a decisive KOSPI large-cap win and an extreme KOSDAQ relative slump. On a market-cap-weighted proxy, KOSPI gained +49.8% and KOSDAQ +11.4%, leaving KOSDAQ -38.4 percentage points behind. But on an equal-weight basis, KOSPI was +14.5% and KOSDAQ +17.4%. This was not a market where all of KOSPI was strong. It was a market where mega-cap AI infrastructure names pulled the KOSPI index higher.

The KOSDAQ/KOSPI relative ratio is now 74.1, down -29.5% from its two-month high, with a 40-session standard score of -1.88σ. In relative-strength terms, it is already deeply oversold. The more important point is flow. Over the past five sessions, KOSDAQ foreigners plus institutions bought +KRW 707.9bn and program flows added +KRW 410.4bn. KOSPI saw -KRW 14.058tn of foreign-plus-institutional selling and -KRW 8.157tn of program selling over the same period. Price says KOSPI won, but flows are starting to look back at KOSDAQ.

Pearl Abyss is the representative watch name in this setup. It ranks 7th among the top 100 KOSDAQ names by drawdown, falling -41.3% from KRW 77,400 to KRW 45,450. That is roughly 19 percentage points deeper than the top-100 median drawdown of -22.0%. But it has 2026E OP of KRW 476.9bn, PER of 8.0x, and ROE of 36.9%. Recent five-session flow shows individuals -KRW 22.33bn, foreigners +KRW 20.77bn, and institutions +KRW 1.32bn. That is why this is not just a simple oversold bounce. It can be treated as an investable rebound candidate.

Still, the reversal is not confirmed. Twenty-session foreign-plus-institutional flow is still -KRW 7.38bn, and price has not yet confirmed a KOSDAQ relative-strength rebound. This is a phase where flows are starting to return, but prices have not confirmed yet. Pearl Abyss is less a stop-loss candidate than an oversold large-cap where rebound torque can rise if KRW 45,000 holds and foreign/private-fund buying continues.

This article is for research and commentary purposes only and is not investment advice. Market returns, relative ratios, flows, drawdown ranks, and consensus numbers are based on the Research OS local database. The data date is May 21, 2026. The figures are individual-stock-based proxy series, not official KRX index values. The KOSDAQ top 100 is constructed from a market-cap proxy using foreign-held shares and foreign ownership ratio, and may differ from actual market-cap rankings. Pearl Abyss 2026E OP of KRW 476.9bn, 2026E PER of 8.0x, and 2026E ROE of 36.9% are local DB consensus figures and may differ from actual results. [Unconfirmed] marks items that could not be reliably verified on the same basis in the comparison group. The analysis may be wrong. Data as of May 21, 2026 KST.

Pearl Abyss Flow Check — Not a Turn Yet. The ₩45,500 Test Between Foreign Bargain Buying and Institutional Selling

Mon, 18 May 2026 18:20:00 +0900

📚 Pearl Abyss × Crimson Desert Series 1Q Recognized Sales and the 2027 Cliff / Patch 1.07 Weekend Data / May 21 IR Checkpoint / What the DLC Comments Really Mean / Pearl Abyss Hub

Pearl Abyss materially reduced “2Q cliff” concerns with its 1Q26 results and 2Q guidance. Yet the stock kept falling after the earnings release. The reason is straightforward: the market has moved on from 2Q and is now pricing the 2027 earnings cliff, capital allocation, and flow dynamics. The tape over the past two weeks does not yet look like a reversal. Foreigners have partially returned since the May 13 forced selling, but institutions have not. This is a “flow-bottom test,” not a “flow turn.”

Key Summary

Two-week flow is not a turn yet. From May 4 to May 18, 2026, the stock fell from ₩58,500 to ₩45,800, a decline of roughly -21.7%. This reads less like a routine pullback and more like a post-1Q valuation reset.
Both foreigners and institutions were net sellers over the full two weeks. Per Naver Finance, cumulative net selling from May 4–18 was -502,638 shares for institutions and -546,424 shares for foreigners, translating to estimated net consideration of approximately -₩26.4B and -₩30.1B respectively.
That said, foreigners showed a partial reversal after the May 13 panic. Net buying came in at +56,653 shares on May 14, +206,722 on May 15, and +65,936 on May 18. Foreign re-entry at lower prices is a positive.
The institutional breakdown still looks weak. Per the Research OS local DB from May 4–15, institutional selling was led by securities firms (-₩17.3B) and private funds (-₩7.5B). Pension funds and insurers have not stepped in meaningfully; trust funds were only marginally positive at +₩0.9B.
Program trading and short selling have calmed, but loan balance risk remains. Program trading swung from -₩34.5B on May 13 to +₩3.0B on May 14, +₩9.7B on May 15, and +₩3.2B on May 18. Short-selling share fell from 19.26% to the 7% range. However, the stock loan balance continued to rise through May 15.
We are currently in stage 3 of a 4-stage process. Distribution and forced selling are behind us; foreign bargain buying and program normalization have appeared. But institutional re-entry and trend recovery have not.
The practical call is hold and observe; additional buying should wait. A genuine flow turn requires holding ₩45,500–₩46,000, reclaiming ₩47,200, and seeing trust funds, pension funds, and insurers add positions in the institutional breakdown.

Data covers May 4–18, 2026. Price, foreign, and institutional flow data are sourced from the Research OS local DB and Naver Finance Pearl Abyss foreign/institutional page.

1. The Price Action Signals a Post-Event Valuation Reset

Looking at the price alone, the correction is already significant.

Date	Close	Change	Volume
5/04	₩58,500	-0.68%	530K shares
5/13	₩51,900	-1.70%	2.23M shares
5/15	₩47,200	-6.72%	1.34M shares
5/18	₩45,800	-2.97%	740K shares

Over two weeks, the stock moved from ₩58,500 down to ₩45,800.

Decline = ₩45,800 ÷ ₩58,500 - 1
= -21.7%

A drawdown of this magnitude is hard to explain as simple post-earnings profit-taking. More precisely, even after 1Q26 results and 2Q guidance were confirmed, the market re-discounted 2027 earnings durability, DLC specificity, shareholder returns, and institutional flow.

The earnings debate around Pearl Abyss has already moved one chapter forward. The 1Q operating profit of ₩212.1B and the company’s 2Q guidance weakened the “immediate 2Q collapse” narrative. The fact that the stock still fell means the market has shifted to the next question:

2026 numbers look strong. But will earnings hold into 2027, and will those earnings return to shareholders?

This decline is therefore best understood not as a simple price correction, but as a period in which peak-earnings discounting and flow repositioning occurred simultaneously.

2. Cumulative Two-Week Flow Is Still Negative

The cumulative May 4–18 flow from Naver Finance is unambiguous.

Participant	Net Shares	Est. Net Consideration
Institutions	-502,638 shares	approx. -₩26.4B
Foreigners	-546,424 shares	approx. -₩30.1B

Foreign ownership also dropped from 6.86% on May 4 to 5.58% on May 18. Viewed across the full two weeks, both foreigners and institutions were net sellers. Overlooking this and saying only that “foreigners are coming back” would be an overly optimistic reading of the flow.

That said, the tone shifted somewhat over the last three trading sessions.

Date	Institutions	Foreigners	Interpretation
5/14	-67,686 shares	+56,653 shares	Foreign bargain buying begins
5/15	-3,902 shares	+206,722 shares	Strong foreign inflow
5/18	-18,417 shares	+65,936 shares	Foreign buying continues; institutions still selling

The key sentence is this:

Foreigners show signs of returning. Institutions do not yet.

Flow reversals rarely end with foreign buying alone. For a KOSDAQ gaming stock with significant earnings, event, and multiple debates, the reversal typically requires institutional model upgrades and position rebuilding as well. Right now we see the first traces of foreign bargain buying, but no institutional confirmation.

3. The Institutional Breakdown Points to Securities Firms and Private Funds

The Research OS local DB data from May 4–15 sharpens the picture of who is selling.

Sub-category	Net Consideration
Securities firms	-₩17.3B
Private funds	-₩7.5B
Pension funds etc.	-₩1.4B
Insurers	-₩0.3B
Trust funds	+₩0.9B
Institutions total	-₩25.7B
Retail	+₩59.4B

Securities firms and private funds drove institutional selling. Long-duration capital — pension funds and insurers — has not stepped in meaningfully. Trust funds were only marginally positive at +₩0.9B, which is insufficient to call a flow turn.

The structure looks equally poor when isolated to the sharp-decline period.

Sub-category	Net Consideration (5/13–5/15)
Securities firms	-₩4.0B
Private funds	-₩7.0B
Pension funds etc.	-₩1.0B
Institutions total	-₩12.0B
Retail	+₩33.1B

During the May 13–15 plunge, retail absorbed the selling while private funds and securities firms were the sellers. This is not a constructive accumulation structure. A healthy pattern would have retail distributing during a sharp decline while foreigners, institutions — specifically trust funds, pension funds, and insurers — step in. Pearl Abyss is not there yet.

This point matters for the investment thesis. Pearl Abyss’s 2026 earnings are already strong. But a re-rating on 2027 earnings defense and DLC/shareholder return/DokeV roadmap requires institutional models to move first. Without institutional re-entry, even strong 1Q earnings can leave the stock stuck in a “cheap but peak” range.

4. Program and Short Selling Have Normalized; Loan Balance Risk Remains

There are positive elements in the flow picture. Program trading and short-selling intensity have clearly peaked.

Item	Recent Trend	Interpretation
Program trading	5/13 -₩34.5B → 5/14 +₩3.0B → 5/15 +₩9.7B → 5/18 +₩3.2B	Normalized after forced selling
Short-selling share	5/13 19.26% → 5/15 6.90% → 5/18 7.20%	Short pressure has peaked
Stock loan balance	5/12 6.10M shares → 5/15 6.57M shares	Short position burden remains
5/18 loan balance	0 shares in local DB	Treated as data anomaly; not used

May 13 was a forced-selling episode with simultaneous program selling and elevated short activity. Since then, program trading has flipped to net buying and short-selling share has fallen from 19.26% to the 7% range. That is clearly positive.

However, the stock loan balance continued to rise through May 15, which means the short position has not been fully unwound. The more precise interpretation is:

Attack intensity has decreased. But the short position burden remains.

The distinction matters: downside pressure has not disappeared — the pace of downside attacks has simply slowed.

5. Current Stage: “Bottom Test”

The cleanest framework for Pearl Abyss flow is a four-stage model.

Stage	Description	Status
1	Distribution begins	Past
2	Forced/panic selling	Occurred 5/13–5/15
3	Foreign bargain buying; program normalization	Current
4	Institutional re-entry; trend recovery	Not yet

We are at stage 3. The May 13–15 panic is behind us, and foreign bargain buying with program normalization has appeared. A bottom-formation possibility therefore exists.

But stage 4 has not arrived. The institutional breakdown is still weak, and at ₩45,800 the stock has not reclaimed its May 15 close of ₩47,200. Calling it a flow turn requires at minimum two things.

First, price must reclaim ₩47,200. Recovering the May 15 panic close is the signal that “the forced selling failed.”

Second, trust funds, pension funds, or insurers must post net buying in the institutional breakdown. Foreign buying alone can mark a bottom test; the return of long-duration institutional capital is a trend-change signal.

6. Price Levels: ₩45,500 and ₩47,200

The levels to watch at the current price are straightforward.

Level	Significance
₩45,500–₩46,000	Near-term support
₩47,200	May 15 close; first recovery target
₩50,500–₩52,000	Bottom of the broken trading range; structural recovery
₩53,000+	Zone confirming joint foreign + institutional return
Below ₩45,500	Risk of further decline and renewed short attack

₩45,500–₩46,000 is the near-term support. Holding this range while foreign buying continues keeps the bottom test intact.

₩47,200 is the first recovery target. As the May 15 panic close, reclaiming this level signals that the forced-selling episode has been fully reabsorbed.

₩50,500–₩52,000 carries greater importance. This is the underside of the broken trading range. Recovering it marks not merely a bounce but a structural normalization of price and flow.

Above ₩53,000, joint foreign and institutional re-entry is required. If price rises while institutions continue selling, the move is likely short covering or a technical rebound rather than a genuine reversal.

Conversely, a break below ₩45,500 would reopen the risk of further decline and a renewed short attack. With stock loan balance still elevated, a downside break could re-stimulate short positioning.

7. Practical Call: Hold and Observe; Wait on Adding

Three things are currently positive for Pearl Abyss.

First, foreigners have started returning since May 14.

Second, program selling has calmed since May 13.

Third, short-selling share has peaked.

Three things are currently negative.

First, the full two-week cumulative flow shows net selling by both foreigners and institutions.

Second, securities firms and private funds drove institutional selling; long-duration capital is absent.

Third, stock loan balance risk remains.

The conclusion therefore is:

This is a “flow-bottom test,” not a “flow turn.”

Existing holders should observe while watching whether ₩45,500–₩46,000 holds. Adding to the position is better deferred until ₩47,200 is reclaimed, foreign buying continues, and the institutional breakdown improves.

The practical decision framework is as follows.

Condition	Judgment
₩45,500–₩46,000 holds + foreigners sustain net buying	Maintain position
₩47,200 reclaimed + program stays net buy	First stabilization
₩50,500–₩52,000 reclaimed + institutional breakdown improves	Flow turn probability rising
Trust funds / pension funds / insurers turn net buyers	Genuine reversal signal
Break below ₩45,500 + foreigners turn net sellers	Reduce exposure or reassess

8. How This Connects to the Existing Pearl Abyss Thesis

This is a flow note, not an earnings or product note — but it cannot be read in isolation from the broader thesis.

As covered in 1Q Recognized Sales and the 2027 Cliff, the 2Q guidance weakens the “immediate cliff” framing. The stock fell anyway because the market is looking at the 2027 earnings cliff.

Patch 1.07 Weekend Data defends the product and sales-rank side of the thesis. Patch cadence and review improvement are positives. But strong product data alone will not drive an immediate price response if flow remains poor.

May 21 IR Checkpoint is the next potential catalyst for institutional flow. If shareholder return policy, the DLC roadmap, and DokeV development stage are clarified, institutional models could start moving again.

What the DLC Comments Really Mean is the core argument for removing the 2027 cliff discount. For a flow turn to materialize, this argument needs to be translated into hard numbers or an official roadmap.

In short, what is needed now is not a single piece of good news. Price, flow, product data, and company communication all need to align.

9. Final Judgment

For Pearl Abyss, the return of foreign buyers since May 14 and the normalization of program trading are positives. The May 13 panic is not simply rolling forward. Short-selling share has come down to the 7% range, easing attack intensity.

But the institutional breakdown is still weak. Securities firms and private funds have been selling; long-duration buying from pension funds, insurers, and trust funds remains insufficient. The fact that retail primarily absorbed the sell-off during the sharp-decline period is also not a constructive accumulation structure.

Therefore, a bottom cannot be ruled out, but concluding that a flow turn has arrived would be premature.

The key levels are ₩45,500–₩46,000 and ₩47,200. Holding ₩45,500–₩46,000 while foreign buying continues, then reclaiming ₩47,200, constitutes first stabilization. Recovering ₩50,500–₩52,000 with institutional breakdown improvement is required before a full flow turn can be discussed.

The practical conclusion is simple.

Hold and observe. Additional buying after confirming ₩47,200 recovery and institutional breakdown improvement.

Now is not the time to aggressively add — it is the time to watch who is absorbing the post-panic supply. The next genuine reversal for Pearl Abyss becomes more credible when sustained foreign buying + return of long-duration institutional capital + answers at the May 21 IR that break the 2027 cliff narrative all arrive together.

This post is for research and commentary purposes only and does not constitute investment advice. Price, volume, foreign, and institutional flow data are sourced from the Research OS local DB and the Naver Finance Pearl Abyss foreign/institutional page. Institutional sub-category, program trading, short selling, and stock loan balance data are from the local DB and may differ from exchange or brokerage source data in timing. The May 18 stock loan balance reading of 0 shares is treated as a data anomaly and was excluded from the analysis. Net consideration figures are estimates based on share counts and prices during the relevant period and may differ from actual executed amounts. Price reference levels (₩45,500–₩46,000, ₩47,200, ₩50,500–₩52,000, ₩53,000+) represent the analyst’s technical and flow-based judgment and carry no guarantee. Pearl Abyss’s share price may change rapidly depending on Crimson Desert sales data, the May 21 IR, DLC/shareholder return/DokeV roadmap developments, global macro conditions, KOSDAQ flow, and short and loan positions. This analysis may be wrong. Data as of May 18, 2026 KST.

Pearl Abyss Short Sale Data: Shorts Hit, But Couldn't Push

Tue, 21 Apr 2026 16:30:00 +0900

📚 Series 6/6: Pearl Abyss × Crimson Desert Thesis — series hub →

Sixth entry in the series. Prior: 5M milestone and re-rating, sell-side consensus gap, initial thesis, BDO second-order catalyst, and 1Q26 earnings preview. This one is the intraday-microstructure read: what did the April 21 tape actually say about who’s in control?

TL;DR

Shorts dumped 64,362 shares and still couldn’t hold the close below the day’s short-average price. The average short price printed KRW 54,245; Pearl Abyss closed at KRW 54,400. Every new short opened today is underwater at the close.

Foreigners (+28,741) and institutions (+7,960) bought the tape — net +36,701 shares absorbed against the short flow. The day ran low-to-close 53,300 → 54,400, a +2.06% recovery off the low. In one sentence: they sold it, and it didn’t break.

The flow story is more important than the one day. Short volume peaked on April 16 at 298,379 shares and has collapsed since: 100,764 → 86,097 → 64,362. Down -78.4% from the peak in four sessions. Shorts keep showing up — but the harder they press, the less it moves.

1. Today’s numbers

1.1 Price and flow

Close: KRW 54,400
Day change: +0.74%
Open / High / Low: KRW 53,700 / 55,100 / 53,300
Foreign net buy: +28,741 shares
Institutional net buy: +7,960 shares
Foreign + institutional net: +36,701 shares

1.2 Short-sale data

Short volume: 64,362 shares
Short-sale turnover: KRW 3,491,300,350
Uptick-rule applied: 55,369 shares
Uptick-rule exempt: 8,993 shares
Uptick-exempt ratio: 14.0%
Average short price: KRW 54,245

1.3 The arithmetic

Average short price = short turnover ÷ short volume:

KRW 3,491,300,350 ÷ 64,362 = KRW 54,244.7 → 54,245

Close vs. average short:

54,400 − 54,245 = +155 KRW

Every share shorted today is KRW 155 underwater at the close.

2. Reading the tape

2.1 This wasn’t a “shorts-were-strong” day — this was a “shorts showed up weakly” day

64,362 shares is not zero — but set against the 100k–300k-range short-pressure days Pearl Abyss has seen recently, today’s intensity is one notch softer. What matters more than the size is the outcome:

Average short: KRW 54,245
Close: KRW 54,400
Result: close above the average short

Translation: shorts got their fills, but couldn’t drag the close below their own average. Today’s shorts read as secondary pressure during a pullback, not as the primary driver of direction.

2.2 Price action was louder than the short tape

Intraday path:

Open 53,700 → Low 53,300 → Close 54,400

Low-to-close recovery: +KRW 1,100, i.e., +2.06%:

(54,400 − 53,300) ÷ 53,300 = 2.06%

Plus foreigners + institutions combined net +36,701 shares. The message is clean:

Absorption exceeded selling + shorting.

Today’s uptick-exempt share of 14.0% is low by recent high-pressure-day standards. The character reads as trading/hedging-driven, not as a coordinated push.

3. The last 10 sessions

Date	Short vol	Short turnover	Avg short	Uptick exempt	Net short position
4/21	64,362	3.491B	54,245	14.0%	pending
4/20	86,097	4.668B	54,224	24.6%	pending
4/17	100,764	5.423B	53,812	19.1%	1,876,700
4/16	298,379	16.799B	56,302	29.2%	1,912,458
4/15	112,944	6.324B	55,989	44.7%	1,775,454
4/14	115,298	6.678B	57,931	10.9%	1,748,500
4/13	266,406	15.040B	56,454	26.0%	1,738,365
4/10	121,743	6.731B	55,296	11.0%	1,643,197
4/9	232,474	13.011B	55,968	25.2%	1,598,112
4/8	98,973	5.656B	57,145	2.3%	1,579,992

4. Structural reads in the 10-day window

4.1 Short intensity collapsed after the April 16 peak

Short volume path:

4/8 98,973
4/9 232,474
4/10 121,743
4/13 266,406
4/14 115,298
4/15 112,944
4/16 298,379 ← 10-day peak
4/17 100,764
4/20 86,097
4/21 64,362

The key is that 4/16 was the short peak. The sequence since:

298,379 → 100,764 → 86,097 → 64,362

Drop from peak:

Δ = 298,379 − 64,362 = 234,017 shares
% = 234,017 ÷ 298,379 = 78.4%

Reading the recent tape as “shorts re-accelerating” is wrong. The accurate read is “short intensity is cooling fast after the April 16 push.”

4.2 Heavy short days and winning short days aren’t the same thing

4/13: 266,406 shorts, close above the day’s average short.
4/16: 298,379 shorts, close below the day’s average short.
4/21: 64,362 shorts, close above the day’s average short.

Of the last 10 sessions, the one short sellers actually won was 4/16. Today is the opposite — they participated, but the result was weak. “Short flow present” and “short advantage” are not synonyms.

4.3 Net short position is still high, but not in lock-step re-accumulation

4/8 1,579,992
4/9 1,598,112
4/10 1,643,197
4/13 1,738,365
4/14 1,748,500
4/15 1,775,454
4/16 1,912,458
4/17 1,876,700

Position ran up 4/8 → 4/16. But 4/17 dropped 35,758 shares:

1,912,458 − 1,876,700 = 35,758 shares decline

The level is still elevated. But there isn’t enough data to declare “structural short re-accumulation confirmed.” Today is a high-position contested zone — neither a clean downtrend resolution nor a clean short-driven takeover.

5. Composite read

One-line verdict

Today wasn’t a shorts-setting-direction day. It was a day when foreign + institutional buying and a low-to-close recovery drained short efficacy.

Why

Today’s short volume (64,362) is low relative to recent peaks.
Close KRW 54,400 is above the average short KRW 54,245 — new shorts are underwater.
Flow support: foreign +28,741, institutional +7,960.
Intraday low KRW 53,300 → close KRW 54,400 recovery.

What kind of zone is this

Upside not yet unlocked, but downside control is slipping — a range-redefinition zone.

“Strong long advantage” is too early. “Short advantage” is the wrong read. Neutral with a slight positive tilt is the honest description.

6. Practical checkpoints

6.1 Confirm an upside turn

Break the KRW 55,100 intraday high.
Foreign + institutional 2–3 consecutive net-buy days.
Short volume staying suppressed below 100k.

6.2 Reverse the read back to bearish if all three:

Short volume re-accelerates.
Net short position breaks back above 1.91M.
Close below the day’s average short for consecutive sessions.

7. Final note

Pearl Abyss is still a name with shorts on the book. That remains true. But the recent data shows short efficacy decaying, not short pressure strengthening. Today specifically: short volume fell + close above average short + foreign and institutional co-buying + low-to-close recovery — all four together.

Cleanest summary:

Shorts are still in the stock, but price-setting power is slipping. Today wasn’t short advantage — today was buy absorption of short pressure.

Appendix: Arithmetic checks

Today’s average short price

3,491,300,350 ÷ 64,362 = 54,244.7 → 54,245

Today’s uptick-exempt ratio

8,993 ÷ 64,362 = 13.97% → 14.0%

Drop from 4/16 peak to today

(298,379 − 64,362) ÷ 298,379 = 78.4%

4/16 → 4/17 net short position decline

1,912,458 − 1,876,700 = 35,758 shares

Disclaimer: For research and information purposes only. Not investment advice. Names cited are for analytical illustration; readers should perform their own due diligence and consult licensed advisors before any investment decision.

Pearl Abyss Daily: KRW 55,000 Support and 5M Copy Watch

Thu, 09 Apr 2026 18:00:00 +0900

Date: 2026-04-09 Price: KRW 56,500 (close) | Intraday Low: KRW 55,000 Category: Daily Monitoring | Crimson Desert Global Sales Rank: #1 Maintained

1. Price Action

Early selling pressure drove the stock to KRW 55,000, but buying demand pushed the close back to KRW 56,500. While the surface appears flat, internally it was an intensely contested day. Both the failure to break below KRW 55,000 and the failure to breach KRW 56,600 were confirmed simultaneously, indicating a range-bound phase within a bearish trend.

Key Levels:

Level	Price
Support	KRW 55,000
Resistance	KRW 56,600
Bullish Reversal Confirmation	Sustained above KRW 57,000

2. Short Selling Analysis

2.1 Transactional Short Selling

Short selling volume hit 232,474 shares (+134.9% day-over-day). However, the average execution price (KRW 55,968) was below the closing price (KRW 56,500), meaning new shorts closed at a loss on the day. The uptick rule exemption share of 25.2% suggests a significant mix of hedging and market-making activity alongside directional shorts.

2.2 Positional Short Interest

While transactional short selling surged, outstanding short positions remain –40.8% below their peak. Today’s short selling spike is interpreted as options expiry-driven transactional and hedging activity rather than structural short accumulation.

3. Flow Analysis

3.1 Key Broker Flow

The upper range (KRW 56,600–57,500) was dominated by selling from Shinhan and foreign brokers, while the lower bound (KRW 55,000) was defended by Kiwoom-led contra buying (+87,003 shares). Sellers did not “seize control” of the market today.

3.2 Foreign Ownership Trend

Foreigners have executed large-scale cumulative buying (+2.53 million shares) over three weeks, forming the structural base for the current price level. However, as of 4/8, daily flow turned to –85,926 shares, indicating fading buying momentum. This overlaps with cumulative institutional net selling (–366,478 shares from 4/2–4/8), suggesting an adjustment phase.

4. Crimson Desert — Product Tracker

4.1 Key KPIs

Metric	Value	Notes
Global Sales Rank	#1 Maintained	Steam Global
Review Positive Rate	83.5%	Very Positive maintained
CCU Trough	75,650	Down from 79,017 prior day
Average Retention Rate	67.0%	Long tail resilience intact

4.2 Sales Milestone Assessment

Sales rank anchor and review quality remain solid, but the consecutive decline in CCU troughs and 67% average retention clearly signal weakening long-tail endurance. 5 million copies is virtually certain, but the path beyond 8 million has visibility but remains uncertain.

5. Catalyst & Event Calendar

Timing	Event	Implication
Within 1 week	Possible official 5M announcement	Already being priced in, “Sell the News” risk
Early-Mid May	Q1 earnings release	Key catalyst — consensus reset

The market’s core question is not “will 5 million happen?” but rather “how much upside remains after 5 million?” Short positions are aware of this, and today’s aggressive selling is interpreted not as a denial of 5M but as pre-pricing the limited price utility after the announcement.

6. Valuation Framework

Premium vs Discount Factors (2027E EV/EBIT basis)

Premium Factors:

Proprietary game engine
Black Desert IP longevity
Crimson Desert-based follow-on revenue streams
Dokkebi option value

Discount Factors:

Hit-driven dependency
Long-tail uncertainty
Absence of shareholder return policy
Limited next-title visibility

What the market watches is not absolute sales volume, but the extent to which that sales volume opens up 2–3 year cash flows and multiple re-rating. 5M vs 6M is meaningful, but 5M vs 5.2M is relatively less significant.

7. Shareholder Return — A Structural Discount

Pearl Abyss’s current price defense logic relies entirely on Crimson Desert’s commercial success, sales volume, and earnings expectations — not shareholder returns. Conviction in cash flow attribution to shareholders remains weak, meaning no policy backstop exists to defend the valuation floor should event risks materialize.

8. Why Today Was So Contested

Today’s intense battle is explained by the simultaneous collision of three factors:

Options Expiry Flow: Hedging, program trading, rebalancing, and rollover volumes reinforced the short selling and short-term selling execution environment.
Diminishing Information Value of the 5M Announcement: Most market participants recognize its imminence → the debate is not about the announcement itself but the magnitude of post-announcement upside → “Sell the News” logic forming.
Flow Crossover: Fading foreign buying + sustained institutional selling vs. retail/Kiwoom contra-buying collision.

9. Conclusion & Monitoring Checklist

Final Assessment

Pearl Abyss is currently in a phase where the stock price isn’t collapsing, but game vitals are holding. Short selling is strong but lacks the power to dominate the downside, while sales event expectations still prop up the floor.

Bullish Reversal Confirmation

KRW 56,600 breakout and sustained (preferably KRW 57,000)
Foreign holdings resume increase
CCU trough recovery, average retention rate back to 70%

Bearish Re-acceleration Confirmation

KRW 55,000 break with volume
Short interest re-accumulation
Sustained institutional net selling, foreign selling continuity
Declining new review count and positive rate

Event Monitoring

Timing of official 5M announcement
Q1 earnings release date disclosure and conference call tone
Company messaging on the 6M+ trajectory