HYLink Digital: A Long-Term Owner's Investment Analysis

Who the customers are and how it charges. [Fact] In 2025 the top five customers contributed combined sales of 401 million yuan, 28.81% of total annual sales; the largest customer was GAC Aion at 8.36% of sales, followed by clients including Shanghai Huansheng Advertising, Chimelong Group, SAIC Volkswagen, and the Xiangjiang Group system. Purchases from the top five suppliers totaled 797 million yuan, 60.65% of total annual procurement, of which the Xiaohongshu system accounted for 30.61%, the Douyin system 14.87%, and the Tencent system 8.65%. [Inference] This shows the company is essentially a service provider connecting brand owners with traffic platforms: it earns integrated service revenue from clients for planning, creative, media buying, content production, and managed operations, then purchases traffic and ad inventory from media platforms. Customers are not overly concentrated in a single buyer, but the supply side relies heavily on a few large platforms, and bargaining power rests largely with those platforms.

Whether revenue is recurring, stable, and predictable. [Fact] Revenue fell from 13.214 billion yuan in 2021 to 1.391 billion yuan in 2025, a decline of about 89.5% over four years; in 2025 revenue fell a further 31.49% year over year. The 2023 annual report also explicitly stated that the revenue decline was primarily driven by cuts in client marketing budgets. [Inference] Demand for this kind of marketing-agency / integrated-marketing business persists over the long run, but the company's revenue recurrence, stability, and predictability are all weak, heavily affected by macro budgets, platform traffic structures, client media-buying preferences, and project-based orders.

Cost structure and dependencies. [Fact] In 2025 the top five suppliers accounted for 60.65% of procurement, mainly platforms such as Xiaohongshu, Douyin, Tencent, Baidu, and iQiyi; that same year, gross margin on the brand-marketing business was only 5.36%, while the city-operations business had a gross margin of 19.24%. In 2023 the brand-marketing gross margin was 11.62%, so by 2025 it had clearly fallen. [Inference] The company is strongly dependent on media channels and traffic platforms, and the traditional brand-marketing business is a low-margin, low-switching-cost business squeezed by platform rules; if city operations and cultural-tourism cannot keep scaling up, their share is still too small to change the overall earnings picture.

Is this a business I can understand. [Opinion] It can be understood, but it is not clean enough. Traditional integrated-marketing services are not complex in themselves, but the problem is that the company is no longer a pure brand-advertising agency; it has layered on multiple transformation lines including AI tools, content marketing, short dramas, e-commerce joint operations, digital cultural tourism, and city operations. The boundaries of the business model are wider than before, which makes it harder in the near term to judge which part can truly create shareholder value. Business understandability score: 3/5.

Would I be willing to hold this business if the stock market closed for five years. [Opinion] Not for now. The reason is not that "the advertising industry has no demand," but that this company has not yet proven it can convert demand into high-quality cash flow on a stable basis; it looks more like a transformation case that needs continuous monitoring than a business ownership stake that lets me sleep easy.

Industry and competitive landscape. [Fact] Research from iResearch in 2025 indicated that China's online advertising market was about 1,131.7 billion yuan in 2023, up 12.4% year over year; another industry report put China's combined internet advertising and marketing market at about 1,248.2 billion yuan in 2023, up about 11% year over year. The China Online Audiovisual Development Research Report (2025) published by Xinhuanet showed that, as of the end of 2024, the short-video user base reached 1.040 billion, micro-short-drama users reached 662 million, and live-streaming users reached 833 million. [Inference] Industry demand is not small, and is even still growing, but the profit pool does not necessarily belong to marketing-service providers; it more likely belongs to leading content platforms, traffic platforms, and brand owners with strong brand assets.

Could the industry be easily disrupted. [Fact] In its 2025 annual report, the company itself acknowledged that the industry is entering a new phase of rising concentration, AI-driven digital-intelligence transformation, scenario innovation, and tightening compliance and risk control. [Inference] This means the industry is not "in decline," but it is one where technology, rules, and traffic migration move very fast, and a service provider's old advantages are easily dissolved. For a company without platform-level control, disruption risk is high.

Main competitors and industry standing. [Fact] Comparable listed companies include at least BlueFocus and Sanrenxing. In May 2026, Google Finance showed BlueFocus with a market value of about 65.38 billion yuan and Sanrenxing about 13.02 billion yuan, while HYLink, based on 8.43 yuan and 253.34 million shares, had a market value of about 2.136 billion yuan. [Inference] Whether by scale, depth of client relationships, or capital-market pricing, HYLink is not in the industry's top tier; it looks more like a small-to-mid-sized service provider trying to find its footing again after a severe contraction.

Industry attractiveness score: 2/5. [Opinion] It is not a "bad industry," but for a marketing-service provider it is one that inherently lacks a strong moat, faces profit pressure from platforms, and changes quickly in both technology and regulation. HYLink at present cannot be called "a good company in a good industry"; it is closer to "a weak service provider in an industry where demand exists but the competitive structure is unfriendly."

Moat and Management

Moat assessment. [Opinion] My assessment of HYLink's moat is on the low side, for the following reasons.

On brand advantage, [Fact] the company has built up some history in the industry, but in 2025 the top five customers accounted for only 28.81% combined and the largest customer 8.36%, showing that customers move around and the brand has not formed strong lock-in; at the same time, the 2025 brand-marketing gross margin was only 5.36%. [Inference] This looks more like "a service provider with some name recognition" than a brand that clients cannot do without.

On cost advantage, [Fact] the top five suppliers accounted for 60.65% of procurement, with core purchases coming from large traffic platforms. [Inference] When upstream platforms are concentrated and powerful, an intermediary service provider struggles to build a real cost advantage. Gross margin falling from about 10.8% in 2023 to about 5.4% in 2025 is exactly the sign of an inability to resist upstream pressure.

On scale advantage, [Fact] revenue fell from 13.214 billion yuan in 2021 to 1.391 billion yuan in 2025; the company's scale has already shrunk significantly. [Inference] In media bargaining, client coverage, and talent attraction, a smaller scale is usually not a source of advantage but the result of advantage being lost.

On network effects and switching costs, [Inference] the company is not a platform and does not control scarce distribution channels; it has neither a content platform's two-sided network effect nor the high switching costs of SaaS-type software. Customers can move between agencies, and platforms will not change rules for the company. Supporting evidence is the low customer concentration alongside high supplier concentration.

On channel and data advantages, [Fact] the company discloses that it uses AI, big data, and intelligent tools to advance its business and is developing new scenarios such as digital cultural tourism; but R&D expense fell from 179 million yuan in 2023 to 74 million yuan in 2024 and then to 32 million yuan in 2025. [Inference] This means the company has made tooling attempts but has not yet proven a software-like, platform-like data moat; judging by investment intensity and ultimate profit performance, it looks more like "efficiency tools" than "exclusive capability."

On corporate culture, operating capability, and capital allocation, the "old management" and the "new management" need to be assessed separately. [Fact] The Beijing bureau of the CSRC determined that the company had failed to disclose, as required, non-operating fund occupation of 181.53 million yuan by its controlling shareholder and actual controller, and had under-provisioned bad-debt allowances, resulting in false statements in the 2021 and 2022 annual reports; in September 2025, the Beijing CSRC bureau imposed administrative penalties on the company, Su Tong, and Guo Jianjun. In February 2025, the company completed a change of control, with the controlling shareholder changing from Su Tong to Hunan Xiangjiang New Area Development Group and the actual controller becoming the Changsha State-owned Assets Supervision and Administration Commission. [Opinion] On the old management, my conclusion is "clearly not worthy of trust"; on the new management, my conclusion is "there is room for improvement, but the track record is too short to award a high score."

Is the moat widening, stable, or narrowing. [Opinion] Over the past few years it has clearly been narrowing. The sharp revenue contraction, margin compression, collapse in profitability, and near-depletion of net assets all indicate that the old moat was not effectively in place; new businesses and state-capital resources might bring an opportunity for rebuilding, but for now this remains a "potential possibility," not a moat that has already been realized.

How long and how much capital a competitor would need to replicate it. [Inference] Replicating traditional integrated-marketing capabilities does not require extremely high capital; what matters more are client relationships, teams, and platform resources, and those things are not firmly held in this industry. What is genuinely hard to replicate is the platforms' own traffic and data, not the agency. For HYLink, this barrier is not high.

Can it raise prices and ride out downturns. [Opinion] Neither looks strong at present. The 2025 brand-marketing gross margin was only 5.36%, while 2022 to 2025 were consecutive loss years; this is not the profile of a company that can easily raise prices in an inflationary environment or hold its earnings steady in an economic downturn. Past margins look more like the joint result of cycle, scale, and working capital than a structural advantage.

Moat strength score: 1/5.

Management and capital-allocation score: 2/5. [Reasoning] The new controlling shareholder brings credit backing and financing support, which is a positive; but the old governance footprint is too heavy, the new team's tenure is too short, and in 2025, with negative operating cash flow and a very thin net-asset base, the company still pushed ahead with cultural-tourism and city-operations expansion and took on substantial debt growth. Whether capital allocation is truly centered on per-share intrinsic value remains unproven.

Financial Quality

The raw data in the table below comes mainly from the 2023, 2024, and 2025 annual reports and the Q1 2026 report; the gross margin, free cash flow, leverage ratio, and working-capital days in the table are figures I calculated from the raw statement data.

Revenue growth and margin trends. [Fact] Revenue declined sharply for four straight years; 2025 revenue was only about 10.5% of the 2021 level. Net margin fell from 1.73% in 2021 to -46.94% in 2025. In 2025 revenue fell 31.49% year over year while operating costs fell only 26.74%, and the brand-marketing gross margin dropped to 5.36%. [Inference] This shows the company's problem is not just "a cyclical downturn in revenue," but that the business model failed to defend its profit structure as the balance sheet shrank. The revenue decline did not buy a healthier margin; instead, a more severe reversal of operating leverage emerged.

Operating cash flow and free cash flow. [Fact] Operating cash flow was 701 million yuan in 2023, but turned to -414 million yuan in 2024 and further to -666 million yuan in 2025; capital expenditure over the same period was very low, only 356,500 yuan in 2024 and 2.0931 million yuan in 2025, so free cash flow in 2024 and 2025 was essentially the same large negative figure as operating cash flow. More tellingly, in Q1 2026 revenue grew 112.74% year over year, yet operating cash flow was still -184 million yuan. [Inference] The company is not an "asset-light, strong-cash-flow" model but a service provider whose cash performance is dominated by working capital; a business recovery does not automatically translate into cash collection.

Are the profits real profits or accounting profits. [Fact] In the supplementary cash-flow statement for 2023, the "decrease in operating receivables" was as high as 2.548 billion yuan, and the "increase/decrease in operating payables" also had an enormous impact on cash flow; in 2025, by contrast, the "decrease in operating receivables" was -24 million yuan and the "increase/decrease in operating payables" was -283 million yuan, directly dragging down operating cash flow. [Inference] The company's cash flow is mainly affected by changes in receivables, payables, and prepayments; the strong cash flow of 2023 looks more like collections and working-capital release during a contraction and should not be misread as high-quality operating cash flow. Conversely, the negative cash flows of 2024-2025 reveal how fragile cash generation becomes once the business base shrinks.

Return on capital. [Fact] The company's weighted-average ROE was 11.79% in 2021, -31.94% in 2022, -56.60% in 2023, and -78.27% in 2024. For 2025, with end-of-period net assets attributable to the parent down to only 56 million yuan, continuing to measure with ROE becomes close to distorted; using the average net assets attributable to the parent for 2024 and 2025, 2025 ROE works out to roughly -270%. [Opinion] This is not "a high-return business in a temporary trough" but a case where the return on capital has already spun severely out of control.

Is the balance sheet sound. [Fact] Total liabilities at the end of 2023, 2024, and 2025 were 2.309 billion yuan, 1.981 billion yuan, and 2.409 billion yuan respectively; total assets were 3.096 billion yuan, 2.182 billion yuan, and 2.179 billion yuan respectively, implying debt-to-asset ratios of about 74.6%, 90.8%, and 110.6%. At the end of 2025, short-term borrowings were 1.459 billion yuan, long-term borrowings 246 million yuan, and cash 375 million yuan, for net debt of about 1.331 billion yuan. At year-end 2025, total owners' equity on a consolidated basis was already negative 231 million yuan; only net assets attributable to the parent remained marginally positive at 56 million yuan; by the end of Q1 2026, net assets attributable to the parent fell further to 31 million yuan. [Inference] From a long-term owner's perspective, this is no longer a sound balance sheet but an obviously fragile balance.

Receivables, payables, inventory, and working capital. [Fact] Although accounts receivable at year-end 2025 fell from 912 million yuan in 2024 to 705 million yuan, because revenue kept shrinking the receivables turnover days actually rose from about 164 days to about 185 days; accounts payable fell from 704 million yuan to 344 million yuan, and payable days shortened from about 143 days to about 95 days. [Inference] On one side, collection difficulty has not clearly improved; on the other, the company is paying down supplier balances, squeezing cash flow from both directions. Once a marketing-service provider loses upstream credit and downstream collection efficiency, cash flow deteriorates rapidly.

R&D, share count, and shareholder returns. [Fact] R&D expense fell from 179 million yuan in 2023 to 74 million yuan in 2024 and then to 32 million yuan in 2025. The share count was 253.34 million shares at year-end 2024, year-end 2025, and the end of Q1 2026, with no significant dilution via large-scale issuance and no substantive buybacks. The company made no annual profit distribution in 2023, 2024, or 2025. [Inference] This is not a shareholder-return company with "steady dividends and steady buybacks," nor a company doubling down on long-term R&D in a trough; it looks more like one preserving liquidity and survival.

Accounting and governance risk. [Fact] The Beijing CSRC bureau's 2025 penalty decision clearly stated that the company had historically failed to disclose non-operating fund occupation of 181.53 million yuan by its controlling shareholder, and had under-provisioned receivables bad-debt allowances, resulting in overstated total profit of 17.3296 million yuan in the 2021 annual report and 69.3931 million yuan in the 2022 annual report. The company also made retrospective adjustments for prior-period accounting errors in 2025. [Opinion] Even if these problems have been exposed and remediated, for a long-term owner, once the trust score is lost, rebuilding it takes many years.

Owner Earnings and Intrinsic Value

Owner Earnings on a conservative basis. Buffett-style "owner earnings" can generally be understood as: net profit attributable to the parent + major non-cash expenses - maintenance capital expenditure - the increment in working capital needed to sustain operations. For HYLink, I adopt a more conservative basis: I do not treat bad-debt and credit impairment as "good non-cash expenses" that can be easily added back, because such losses have recurred in the company's history and have been confirmed by regulatory penalties.

[Fact] Net profit attributable to the parent in 2025 was -653 million yuan; depreciation-and-amortization items (fixed-asset depreciation, right-of-use-asset amortization, intangible-asset amortization, and long-term deferred-expense amortization) totaled about 46 million yuan; capital expenditure was about 2 million yuan; operating cash flow was -666 million yuan and free cash flow about -668 million yuan. Free cash flow in 2024 was also about -414 million yuan. [Opinion] Measured by the cash a long-term owner could actually take home, conservative Owner Earnings for 2025 is approximately -670 million yuan, well below zero.

Why 2023 should not be misjudged as high Owner Earnings. [Fact] Operating cash flow in 2023 was 701 million yuan and free cash flow about 699 million yuan, which looks strong; but the supplementary cash-flow data shows that the "decrease in operating receivables" that year contributed 2.548 billion yuan, while the "decrease in operating payables" was a drag of 1.572 billion yuan. [Inference] This is a textbook case of working-capital release during a contraction, not a stable compounding machine. Treating 2023 as the norm would severely overstate intrinsic value.

How many times Owner Earnings the current valuation implies. [Opinion] Because conservative Owner Earnings is negative, the Owner Earnings multiple implied by the current price is meaningless. A more useful question is: how much sustainable owner earnings does the current price imply the company must eventually earn for it to be reasonable.

[Fact] Based on the May 21, 2026 share price of about 8.43 yuan and 253.34 million shares, equity market value was about 2.136 billion yuan. [Assumption] If investors require a 12% return on equity and a long-term perpetual growth rate of 2%, the stable Owner Earnings needed to support the current market value is about 210 million yuan. [Inference] That is nearly the level of net profit attributable to the parent the company achieved in 2021, when revenue was 13.214 billion yuan, whereas 2025 revenue was only about 10.5% of the 2021 level. Re-earning distributable cash flow close to the historical peak on a vastly smaller business base is extremely difficult.

Discounted Owner Earnings valuation. The table below shows my three scenarios, with amounts in millions of yuan, all as estimates of Owner Earnings attributable to shareholders. It is so conservative because the company is still in a period of governance repair and business restructuring. The scenarios are not forecasts but a reverse test of "what would need to happen for the current price to hold."

[Opinion] Even using the optimistic scenario, which is relatively friendly to the transformation, the resulting per-share value is only around 5 yuan, below the current price of 8.43 yuan. Under this framework, the valuation ranges I give are: Conservative intrinsic value range: 0.2-1.5 yuan per share; Fair intrinsic value range: 1.5-3.5 yuan per share; Optimistic intrinsic value range: 3.5-5.5 yuan per share. The current price is still about 53% above the optimistic upper bound and about 141% above the fair upper bound.

Relative valuation. [Fact] As of May 2026, HYLink's share price was about 8.43 yuan; based on the official share count, market value was about 2.136 billion yuan. On 2025 revenue of 1.391 billion yuan, the price-to-sales ratio is about 1.53x; on year-end 2025 net assets attributable to the parent of 56 million yuan, the static price-to-book is about 38x; on end-of-Q1-2026 net assets attributable to the parent of 31 million yuan, the rolling near-term price-to-book is about 69x. At year-end 2025, interest-bearing debt (short-term plus long-term borrowings) was about 1.705 billion yuan and cash 375 million yuan, for a rough EV of about 3.466 billion yuan and a corresponding EV/Sales of about 2.49x. Because the company's net profit, free cash flow, and EBITDA are all negative, PE, P/FCF, and EV/EBITDA carry no normal comparative meaning.

[Fact] Over the same period, Google Finance showed BlueFocus with a market value of about 65.38 billion yuan and a PE of about 243.8x, and Sanrenxing with a market value of about 13.02 billion yuan and a PE of about 100.9x. [Inference] Peer valuations are not cheap in themselves, and several of these companies also carry transformation, sentiment, and thematic premiums. [Opinion] This is exactly the reminder: we cannot treat HYLink as cheap just because its peers are all expensive. For a company with negative Owner Earnings and nearly depleted net assets, high peer valuations do not constitute a margin of safety.

Asset or liquidation value. [Fact] At year-end 2025, the company held cash of 375 million yuan, trading financial assets of 4 million yuan, and other equity-instrument investments of 108 million yuan; but it also had accounts receivable of 705 million yuan, prepayments of 215 million yuan, deferred tax assets of 206 million yuan, intangible assets of 35 million yuan, and goodwill of 18 million yuan; total liabilities were 2.409 billion yuan, and total owners' equity on a consolidated basis was negative 231 million yuan. Net assets attributable to the parent were 56 million yuan, equivalent to book net assets of only about 0.22 yuan per share; by the end of Q1 2026 this fell to about 0.12 yuan per share. [Inference] On a more conservative liquidation view, stripping out deferred tax assets, goodwill, and intangibles and discounting prepayments and receivables, the residual value left to shareholders is very likely low and could even be negative.

Buy range and hold range. [Opinion] Buying with a margin-of-safety requirement of at least 30%-40%: Ideal buy price range: 1.5-2.5 yuan per share; Acceptable hold price range: 2.5-3.5 yuan per share; Clearly overvalued price range: above 5.5-6 yuan per share. At about 8.43 yuan on May 21, 2026, I believe it is already in the clearly overvalued range.

Margin of Safety and the Bear Case

Is the current price cheap enough. [Opinion] Not cheap enough, and not "a little not cheap" but clearly not cheap under a long-term owner's framework. There are three layers of reasons: First, conservative owner earnings are negative, meaning what you are buying is not distributable cash flow but a transformation narrative; Second, book net assets are extremely thin, and the balance sheet provides insufficient downside protection; Third, the past governance problems are already proven, so a trust discount should exist rather than disappear.

The most fragile assumption in the valuation. The most fragile is not "whether revenue can grow" but whether revenue growth can convert into stable positive cash flow and positive owner earnings. Q1 2026 already provides a warning: revenue grew 112.74% year over year, but operating cash flow was still -184 million yuan. [Inference] If growth continues to be accompanied by working-capital consumption, concessions to platforms, and added debt, then growth itself will not create value.

If growth comes in below expectations, does the investment still hold. [Opinion] Most likely not. Because the company's current book cushion is too thin, there is almost no buffer where "even if growth disappoints, shareholders can still get their money back through assets and cash flow."

If margins decline, does the investment still hold. [Opinion] It would also be hard to hold. The 2025 brand-marketing gross margin was only 5.36% and the overall gross margin about 5.41%, already close to "not much room for error." A further small decline and expenses, interest, and credit impairment would swallow all operating results.

If the valuation multiple contracts, would it cause permanent loss. [Opinion] Yes, and it may be one of the main sources of permanent loss. Because the current price already clearly embeds the option of "a successful transformation after the state-capital takeover." If the facts later show the transformation is slower than the market imagines, the share price need not wait until operations deteriorate to fall; the valuation could contract first the moment the imagined upside narrows.

Is there a "good company at a bad price" situation here. [Opinion] This is closer to "a highly uncertain company + a not-cheap price" than to "a good company at a bad price." If, over the next two to three years, real operations improve, the statements are rebuilt, and ST pressure is removed, then it will not be too late to discuss "whether the price is appropriate."

The strongest bull case. The bulls' strongest logic is not hard to summarize: [Fact] In 2025 the company's controlling shareholder changed to Xiangjiang Group and the actual controller to the Changsha State-owned Assets Supervision and Administration Commission; Xiangjiang Group also provided guarantee support for several of the company's financings. The company is advancing new directions including city operations, digital cultural tourism, AI tools, and content-driven marketing, and Q1 2026 revenue has already recovered clearly year over year. [Inference] If state-capital credit support solves the financing problem, the bad legacy is gradually cleared, and city-operations and cultural-tourism projects bring higher-margin revenue, then the company's income statement and cash-flow statement could both improve significantly in 2026-2027.

Why this investment could still be wrong. [Opinion] I think the right way is to flip it around: if I buy today, where am I most likely to be wrong? The answer is that I would be mistaking "a state-capital background + a transformation narrative" for "a business that has already built high-quality cash flow." In reality, many turnarounds can only repair the financials, not necessarily the business model; many revenue gains only repair scale, not necessarily shareholder returns.

What facts would overturn my current negative judgment. If the following facts emerge in the future, I would admit that today's "Avoid" judgment may have been wrong: First, the company achieves positive operating cash flow for at least four consecutive quarters, and not by relying on a one-off collection or asset disposal; Second, short-term borrowings fall markedly and net assets attributable to the parent expand again to above 500 million yuan, restoring a balance-sheet cushion; Third, the share of new-business revenue rises significantly and the company's overall gross margin stabilizes back above 12%-15%; Fourth, there are no new accounting-error corrections, fund occupations, or major disclosure problems. These are not slogans but facts that must be verified in the statements.

The worst permanent capital-loss scenario. [Opinion] The worst case is not "short-term share-price volatility" but: the transformation fails to take effect for a long time, cash flow stays negative, financing relies increasingly on external guarantees, and the company ultimately survives through asset sales, debt restructuring, low-priced issuance, or other equity reorganization, with old shareholders severely diluted in the process. Because net assets attributable to the parent were only 31 million yuan at the end of Q1 2026, this risk cannot be treated as a low-probability event.

Investment Checklist and Final Judgment

Comparison with other opportunities. [Opinion] Comparing HYLink with a broad-based index, I do not think it is clearly superior. A broad-based index has the advantages of diversification, automatic elimination of the weak, and long-term reliance on the overall profits of Chinese companies rather than the governance repair of a single firm; HYLink's current return depends heavily on whether one company can smoothly complete its transformation, financing, and trust repair. For balanced-to-conservative 10-year capital, this risk compensation is not enough. Compared with risk-free yield-type assets, HYLink currently pays no dividend, has negative conservative owner earnings, and cannot offer a clear, realizable cash-return anchor.

Investment Checklist. The table below is my overall judgment, which is [Inference/Opinion], based on the operating, financial, governance, and valuation facts already cited above.

Final rating. Avoid.

One-sentence investment thesis. This is not a good business already proven to generate real, distributable cash flow over the long run, but a company with impaired governance and a fragile balance sheet whose valuation already reflects a transformation option in advance.

Core bull case.

• The controlling shareholder has changed to Xiangjiang Group, and the state-capital background does improve financing and survival capability.

• The company has some resource-integration room in city operations, digital cultural tourism, and AI tools.

• Q1 2026 revenue recovered year over year, indicating the business has not completely lost demand.

Core bear case.

• Consecutive losses from 2022 to 2025, consecutive large negative operating cash flows in 2024-2025, and still no improvement to positive in Q1 2026.

• Net assets attributable to the parent are extremely thin, only 56 million yuan at year-end 2025 and only 31 million yuan at the end of Q1 2026; the balance sheet is fragile.

• A history of fund occupation, under-provisioned bad debt, and false disclosure confirmed by regulators, making the trust cost very high.

• Suppliers are highly concentrated in traffic platforms, the traditional brand-marketing gross margin is low, and there is no pricing power or moat.

• At the current price of 8.43 yuan, there is no clear discount even to my intrinsic value under the optimistic scenario.

Reasons not to buy. First, I do not see a long-term moat that is already proven. Second, I do not see a balance sheet reassuring enough. Third, I do not see positive owner earnings on a conservative basis. Fourth, I do not see the current price offering a margin of safety that pays for these risks. Of these four, I usually will not buy if even two are unmet; here, almost all four are unmet.

Key assumptions. If someone insists on investing, this investment requires at least: the company turns operating cash flow positive in 2026-2027; short-term borrowings and total leverage fall year by year; new businesses contribute verifiable high-margin revenue; no new financial retrospective corrections or major regulatory problems occur.

Fair buy price. The long-term-value buy range I give is 1.5-2.5 yuan per share, built on the premise of "fair intrinsic value of 1.5-3.5 yuan per share + a 30%-40% margin of safety." If there is no price pullback, there is no need to force a purchase.

Target holding period. From a value-investing perspective, a holding period of more than 5 years is only worth discussing after the statements have substantively repaired; at the current stage, I do not recommend entering on a "buy first, watch later" basis.

Expected annualized return. Taking a purchase at the current price of 8.43 yuan and a 10-year hold as an example, based on my valuation ranges above, the expected annualized return is roughly: Conservative scenario -16% to -25%; Neutral scenario -8% to -12%; Optimistic scenario -4% to +1%. This is not because I am forecasting the share price, but because the current buy price is already above the intrinsic value range I can give.

Maximum loss risk. In the worst case, if the transformation fails, cash flow stays negative, and refinancing comes at the cost of diluting shareholders, a permanent capital loss of 70%-100% for shareholders is not far-fetched. For a company with net assets of only a few tens of millions of yuan, this risk must be faced squarely.

Tracking metrics. If I keep tracking it in the future, I will watch only these: operating cash flow, free cash flow, short-term borrowing balance, net assets attributable to the parent, accounts-receivable turnover days, brand-marketing gross margin, the share of city-operations/cultural-tourism revenue, whether there are new accounting-error corrections, whether related-party fund occupation or abnormal guarantees recur, and whether the ST status changes.

Signals that would trigger a reassessment. If there are four consecutive quarters of positive operating cash flow, net debt falls significantly, gross margin recovers to double digits, and there are no new governance flaws, I would reassess and possibly raise the rating; conversely, if net assets keep falling, cash flow keeps deteriorating, or disclosure problems recur, I would firmly maintain Avoid.

Final recommendation. Coolly put, HYLink is not a "cheap long-term compounding asset" right now but a "high-uncertainty rebuilding case." For a 10-year-plus, balanced-to-conservative capital account, I would rather reserve capital for businesses that can be understood simply, can consistently generate free cash flow, have a cleaner governance record, and are genuinely trading at a discount today. HYLink can be tracked, but it is not worth buying now.

This report is based on public information and does not constitute investment advice. Markets carry risk; invest with caution.