The short answer: look for connected signals简要答案：寻找能够相互连接的信号
When facing an unfamiliar stock, first look for evidence that demand is useful and repeatable; then ask how gross profit is created, how much capital must be reinvested, whether accounting results become cash, whether the balance sheet can absorb disappointment, whether value reaches each share, and what future performance the price already requires. No signal should stand alone. A stronger reading explains both the confirming connection and the contradictory combination that could overturn it.
面对一只陌生股票，先寻找需求是否有用且可重复的证据；再判断毛利如何产生、维持业务需要再投入多少资本、会计结果能否兑现为现金、资产负债表能否承受失望、价值是否真正落到每股，以及当前价格已经要求公司实现怎样的未来表现。任何信号都不能单独成立；更可靠的解读必须同时说明哪些连接相互确认，以及哪些矛盾组合可能推翻结论。
Read a stock as a stack, not a scorecard把股票读成信号栈，而不是评分卡
A single metric can be accurate yet misleading. Faster orders can require uneconomic discounts; wider gross margin can come from mix rather than durable pricing; profit can rise while owners receive less per share. Each layer therefore asks what must be true for the previous layer to matter.
单个指标可以完全准确，却仍然具有误导性。订单加快可能依靠不经济的折扣；毛利率扩大可能来自产品组合，而不是持久定价；利润上升时，每位股东获得的结果仍可能下降。信号栈解决的是阅读顺序问题：每一层都追问上一层若要真正有意义，还必须满足什么条件；只有相邻披露能够相互印证，结论才暂时成立。
| Layer | Signal worth finding | Neighbor that must agree |
|---|---|---|
| Demand quality | Repeat need, retention, backlog conversion, stable customer behavior. | Price, returns, cancellations, concentration. |
| Gross-profit mechanism | Price, volume, mix, sourcing, utilization, or service contribution. | Operating costs and competitive response. |
| Reinvestment return | Additional operating result relative to additional capital employed. | Maintenance needs and accounting definitions. |
| Cash realization | Timing and durability of cash after required reinvestment. | Receivables, inventory, contracts, taxes, provisions. |
| Capital and per share | Liquidity, maturities, dilution, buybacks, option and conversion effects. | Downside capacity and owner outcome. |
| Valuation requirement | Growth, margin, capital intensity, and duration implied by price. | Evidence from every layer above. |
| 层级 | 值得寻找的信号 | 必须与之吻合的相邻层 |
|---|---|---|
| 需求质量 | 重复需求、留存、积压订单兑现、稳定客户行为。 | 价格、退货、取消、集中度。 |
| 毛利机制 | 价格、销量、组合、采购、产能利用或服务贡献。 | 经营费用与竞争反应。 |
| 再投资回报 | 新增经营结果相对于新增投入资本。 | 维持性投入与会计定义。 |
| 现金兑现 | 满足必要再投资后，现金出现的时点与持续性。 | 应收款、存货、合同、税项、预计负债。 |
| 资本与每股 | 流动性、到期债务、稀释、回购、期权与转换影响。 | 下行承受力与股东结果。 |
| 估值要求 | 价格隐含的增长、利润率、资本强度和持续期。 | 上方每一层的证据。 |
Begin with demand that survives a price question先看能经受价格追问的需求
Demand quality is stronger when the buyer’s need, purchasing trigger, renewal or replacement cycle, and switching alternative are visible. Reported growth is only the starting observation. Separate units from price and mix; identify acquisitions, currency, backlog, cancellations, returns, renewal cohorts, and customer concentration where the business model makes them relevant. A demand signal becomes more credible when it persists without worsening payment terms or exceptional incentives.
当买方需求、采购触发点、续约或替换周期以及可切换方案都清楚时，需求质量才更强。已披露增长只是起点。应把销量、价格和组合分开，并根据商业模式识别收购、汇率、订单积压、取消、退货、续约群组和客户集中度。若需求能够在付款条件没有恶化、也不依赖异常激励的情况下延续，可信度才会上升。
A maintenance distributor reports repeat order frequency together with price, product mix, cancellations, and customer concentration.
有上下文的信号维护零部件分销商同时披露重复下单频率、价格、产品组合、取消和客户集中度。
Bookings rise, but the filing does not define the measure, distinguish acquisition effects, or show whether orders can be cancelled.
亮眼但不完整订单额上升，但披露没有定义指标、区分收购影响，也未说明订单能否取消。
Who pays, why now, what can they substitute, and what evidence would show that the demand was pulled forward rather than repeated?
追问需求机制谁付款、为何现在购买、可以替代什么，以及什么证据会表明需求只是提前发生而不是能够重复？
Ask what created the margin—and what it consumed追问毛利由什么创造，又消耗了什么
Gross margin is an economic clue, not a quality certificate. Reconcile the change to price, volume, mix, supplier terms, freight, input costs, utilization, warranty, service content, and acquisition accounting. Then continue below gross profit. A favorable mix can raise the percentage while selling expense, engineering support, or customer-acquisition cost absorbs the benefit. Compare segment definitions and note whether management changed the denominator.
毛利率是经济机制的线索，不是质量证书。应把变化调节到价格、销量、组合、供应商条款、运输、投入成本、产能利用、质保、服务含量和收购会计，再继续观察毛利以下的支出。产品组合改善可以提高毛利率，但销售费用、工程支持或获客成本可能吞掉收益。还要比较分部定义，并确认管理层是否改变了分母。
Next, examine the return on reinvestment. The useful question is what additional operating result followed from additional inventory, equipment, acquired intangibles, working assets, or capitalized development. A rough incremental return can be framed as change in normalized after-tax operating profit divided by change in operating capital, but both terms require a consistent scope. Acquisition years, asset write-downs, leases, and early capacity builds can make a short-period result incomparable.
随后观察再投资回报：新增存货、设备、收购无形资产、经营性资产或资本化开发投入，带来了多少新增经营结果。可以用“标准化税后经营利润变化 ÷ 经营资本变化”形成粗略增量回报，但分子和分母必须保持一致范围。收购年份、资产减记、租赁以及提前建设产能，都可能让短期结果失去可比性。
- Stronger combination: stable demand, explained margin improvement, and returns that remain attractive after the necessary capital is included.
- Contradiction: margin expands while inventory ages, service obligations rise, or new capacity remains underused.
- Missing evidence: management names “mix” but does not disclose the products, segments, or costs that changed.
- 更强组合：需求稳定、毛利改善原因清楚，并且纳入必要资本后回报仍有吸引力。
- 矛盾组合：毛利扩大，但存货老化、服务义务上升或新增产能长期闲置。
- 缺失证据：管理层用“组合”解释变化，却没有披露哪些产品、分部或成本发生改变。
Read cash realization beside capital fragility把现金兑现与资本脆弱性放在一起
Cash quality is about timing, repeatability, and claims on the cash—not simply whether one annual figure is positive. Trace reported profit through noncash charges, receivables, inventory, contract assets and liabilities, supplier timing, taxes, provisions, interest, and required capital spending. Distinguish a deliberate build for secured demand from inventory that lacks a dated conversion path. Distinguish temporary tax timing from a rate that cannot recur.
现金质量关注的是时点、可重复性以及谁对现金拥有请求权，而不仅是某个年度数字是否为正。把已报告利润一路追踪到非现金项目、应收款、存货、合同资产与负债、供应商结算时点、税项、预计负债、利息和必要资本支出。要区分服务确定需求的主动备货与没有明确变现路径的存货，也要区分暂时性税务时点与不可持续税率。
Then ask whether the balance sheet grants time for the thesis to work. Map cash restrictions, debt maturities, floating rates, covenants, collateral, leases, guarantees, preferred claims, purchase commitments, pensions, litigation, and refinancing dependence. Net debt can summarize a position, but it can hide near-term maturity, trapped cash, covenant pressure, or obligations excluded from the chosen definition. Fragility matters most when the operating thesis requires more time or more capital.
再判断资产负债表是否给观点留出足够时间。梳理受限现金、债务到期、浮动利率、契约、抵押、租赁、担保、优先请求权、采购承诺、养老金、诉讼和再融资依赖。净债务可以概括状况，却可能隐藏短期到期、无法调回的现金、契约压力或被定义排除的义务。当经营观点需要更多时间或资金时，资本脆弱性尤其重要。
| Promising signal | Contradictory neighbor | What to open next |
|---|---|---|
| Backlog grows. | Inventory and cancellation language grow faster. | Backlog definition, inventory aging, cancellation rights. |
| Profit improves. | Receivables age and factoring increases. | Credit policy, allowance, financing and cash notes. |
| Liquidity appears ample. | A maturity or covenant test arrives before the operating milestone. | Debt schedule, covenant calculation, subsequent events. |
| 积极信号 | 矛盾的相邻信号 | 下一步打开什么 |
|---|---|---|
| 订单积压增长。 | 存货与取消条款增长更快。 | 积压订单定义、存货账龄、取消权。 |
| 利润改善。 | 应收账龄拉长且保理增加。 | 信用政策、减值准备、融资与现金附注。 |
| 流动性看似充足。 | 债务到期或契约测试早于经营里程碑。 | 债务到期表、契约计算、期后事项。 |
Translate valuation into required future evidence把估值翻译成未来必须出现的证据
Valuation should come last because price converts the operating story into a required result. Record the security, price, timestamp, share-count definition, debt and cash scope, and the source period behind each input. Then reverse the question: what demand duration, margin, reinvestment rate, cash realization, and dilution outcome must occur for the price to be reasonable under the chosen method? This is more informative than declaring a multiple high or low in isolation.
估值应放在最后，因为价格会把经营故事转换为必须实现的结果。记录证券类别、价格、时间戳、股数定义、债务与现金范围，以及每项输入对应的来源期间。随后反向提问：在所选方法下，需求要持续多久、利润率达到什么水平、再投资率和现金兑现怎样、稀释结果如何，当前价格才算合理？这比孤立判断某个倍数高低更有信息量。
Separate reported facts from model assumptions. A filed share count is a fact for its date; a future margin is an assumption; discounting is a calculation; “too demanding” is a judgment. Test combinations rather than one variable at a time. Lower margins may also reduce reinvestment capacity, and slower demand can increase inventory and refinancing pressure. A valuation case that survives only when every favorable assumption occurs together is fragile.
必须把已披露事实与模型假设分开。某日披露的股数是事实；未来利润率是假设；折现过程是计算；“要求过高”是判断。敏感性分析应测试变量组合，而不是一次只动一个值：利润率下降也可能削弱再投资能力，需求放缓还可能增加存货与再融资压力。只有所有有利假设同时实现才成立的估值情景，本身就很脆弱。
Give contradictions more weight than isolated highlights矛盾组合应比孤立亮点获得更高权重
The most useful signals often appear as disagreements between documents or layers. Do not average them into a neutral score. Name the conflict, identify whether timing or definitions could reconcile it, and specify the next disclosure that would decide. Silence is not proof of deterioration, but a previously useful metric disappearing is a research signal that deserves a source-level explanation.
最有价值的信号经常表现为文件之间或层级之间的不一致。不要把它们平均成一个中性分数。应明确写出冲突，判断时点或定义能否调节，并指定能够解决问题的下一项披露。沉默本身不能证明恶化，但过去有用的指标突然消失，确实值得在来源层面追问原因。
| Combination | Possible reading | Evidence still needed |
|---|---|---|
| Higher prices + stable units + lower retention | Near-term pricing may be offsetting customer loss. | Cohorts, cancellations, contract terms, competitor response. |
| Wider margin + older inventory + supplier concentration | Scarce or premium stock may help margin while increasing obsolescence and dependency risk. | Aging, write-downs, purchase commitments, supplier agreement. |
| Higher total profit + lower per-share result | Financing or compensation dilution may absorb operating progress. | Share bridge, awards, conversions, issuance purpose. |
| Low multiple + maturity before recovery milestone | Apparent cheapness may reflect financing risk rather than neglect. | Covenant headroom, refinancing terms, downside liquidity. |
| 组合 | 可能解读 | 仍需证据 |
|---|---|---|
| 价格提高＋销量稳定＋留存下降 | 短期提价可能正在抵消客户流失。 | 客户群组、取消、合同条款、竞争反应。 |
| 毛利扩大＋存货变老＋供应商集中 | 稀缺或高价存货可能提高毛利，却增加过时和依赖风险。 | 账龄、减记、采购承诺、供应协议。 |
| 总利润提高＋每股结果下降 | 融资或薪酬稀释可能吸收经营进步。 | 股份桥接、奖励、转换、发行用途。 |
| 低倍数＋债务到期早于复苏里程碑 | 表面便宜可能反映融资风险，而非市场忽视。 | 契约余量、再融资条款、下行情景流动性。 |
Signal-stack example: the attractive distributor信号栈示例：看似亮眼的专业分销商
The following company and figures are entirely hypothetical. AlloyBridge Components distributes precision bearings and seals to industrial maintenance contractors. Its illustrative filing reports sales of $140 million, gross profit of $42 million, year-end inventory of $58 million, cost of goods sold of $98 million, one supplier representing 44% of purchases, current net income of $19 million, and 23 million diluted weighted-average shares. The prior period showed $18 million of net income and 20 million diluted shares.
以下公司与数字完全虚构。AlloyBridge Components 向工业维护承包商分销精密轴承与密封件。虚构披露报告销售额 1.40 亿美元、毛利 4,200 万美元、期末存货 5,800 万美元、销售成本 9,800 万美元；一家供应商占采购额 44%；本期净利润 1,900 万美元，稀释加权平均股数 2,300 万股。上期净利润为 1,800 万美元，稀释股数为 2,000 万股。
| Evidence class | Work shown | What it permits |
|---|---|---|
| Reported facts | The seven figures above, each tied to the hypothetical filing period and note. | They establish inputs, not quality or attractiveness. |
| Calculations | $42m ÷ $140m = 30% gross margin. $58m ÷ $98m × 365 = about 216 days using year-end inventory. $19m ÷ 23m = $0.83 per share; $18m ÷ 20m = $0.90. | The simplified per-share result fell about 8.2% despite higher total net income. |
| Assumptions | Year-end inventory approximates the period level; supplier purchase share signals operational dependence; no special conversion terms alter diluted shares. | These are testable choices, not reported conclusions. |
| Judgment | The 30% gross margin is not enough to call the stock attractive because inventory duration, supplier dependence, and per-share dilution contradict the highlight. | Priority moves to aging, purchase commitments, share issuance, and the supplier agreement. |
| 证据类别 | 展示过程 | 允许得出的内容 |
|---|---|---|
| 已披露事实 | 以上七项数字分别连接到虚构报告期间和附注。 | 它们只确立输入，不能直接证明质量或吸引力。 |
| 研究者计算 | 4,200 万 ÷ 1.40 亿 = 30% 毛利率。5,800 万 ÷ 9,800 万 × 365 ≈ 216 天，分子使用期末存货。1,900 万 ÷ 2,300 万 = 每股 0.83 美元；1,800 万 ÷ 2,000 万 = 每股 0.90 美元。 | 虽然总净利润上升，简化每股结果约下降 8.2%。 |
| 研究假设 | 期末存货近似代表期间水平；供应商采购占比代表经营依赖；没有特殊转换条款改变稀释股数。 | 这些是可检验选择，不是已披露结论。 |
| 研究判断 | 仅凭 30% 毛利率不能称股票有吸引力，因为库存持续时间、供应商依赖和每股稀释形成反证。 | 优先级应转向账龄、采购承诺、股份发行和供应协议。 |
Average inventory, aging buckets, write-down policy, purchase commitments, reasons for share growth, and termination rights in the supplier arrangement are missing. Until they are found, the example supports a research priority—not a verdict.
缺失证据会改变结论平均存货、账龄分组、减记政策、采购承诺、股数增长原因以及供应安排中的终止权仍然缺失。在找到这些信息之前，示例只支持新的研究优先级，而不支持最终结论。
Compress the stack without losing its tensions压缩信号栈，但不要丢掉其中的张力
End with three sentences. First, state the strongest connected signal and cite its sources. Second, state the contradiction with the greatest ability to reverse the reading. Third, state the next observable filing item, period, or threshold. This format prevents a long review from collapsing into a list of attractive ratios and makes later updates comparable with the original evidence.
最后用三句话收束。第一句写最强的相互连接信号并注明来源；第二句写最可能推翻当前解读的矛盾；第三句写下一项可观察的披露项目、期间或阈值。这样可以避免长篇研究最终坍缩成一串漂亮比率，也让后续更新能够与原始证据比较。
Compare signals across an uploaded filing set在上传披露组合中对照信号
To map the signal stack, upload the latest annual and interim filings, notes, proxy material, event filings, and an earlier comparison period. InfiniSynapse Stock Explained can help organize the packet, locate passages about each layer, and place apparently conflicting statements side by side for review.
The output is a starting map, not verified evidence. Confirm issuer, form, reporting period, units, definitions, table headers, page locations, and every material number in the original uploaded files. When two passages use different scopes or dates, preserve the disagreement rather than asking the summary to choose one. Provide a separately dated price input if valuation is part of the question.
绘制信号栈时，上传最新年报和中期报告，并加入附注、委托投票材料、重大事件披露以及比较所需的早期期间。InfiniSynapse“一眼看懂这只股票”可以帮助整理资料包，定位与需求、毛利、再投资、现金、融资、股数和风险有关的段落，并把看似矛盾的说法并排呈现以供复核。
生成结果只是起始地图，不是已核验证据。必须在原始上传文件中确认发行人、表格、报告期、单位、定义、表头、页码和每个重要数字。若两段内容采用不同范围或日期，应保留分歧，而不是要求摘要替你选择。若涉及估值，还要单独提供带日期的价格输入。
Prepare comparable filings and one narrow question, then ask for source locations behind each confirming or contradictory signal. Verify the proposed comparison in the originals.
准备可比披露和一个边界清楚的问题，要求返回每项确认或矛盾信号的来源位置，再回到原文核验建议对照。
Open Stock Explained打开“一眼看懂这只股票”Questions that follow from the signal stack读完信号栈后常见的问题
Start with where demand sits in the cycle and what created it, then read margins, inventory, capacity, and financing against that position. A peak margin or backlog figure can be weakest precisely when replacement demand has been pulled forward.
研究周期性公司时，哪类信号应该放在最前面？先判断需求位于周期哪个位置以及它由什么创造，再把利润率、存货、产能和融资放在同一阶段下解读。需求被提前释放时，峰值毛利或订单积压反而可能最脆弱。
No. It can reflect pricing power, valuable service, scarce supply, favorable mix, capitalization choices, or temporary cost movement. Trace the mechanism and confirm that operating expense, reinvestment, returns, or customer behavior do not absorb the apparent advantage.
高毛利率是否总是强信号？不是。它可能来自定价能力、有价值的服务、稀缺供应、有利组合、资本化选择或暂时成本变化。必须追踪机制，并确认经营费用、再投资、退货或客户行为没有吞掉表面优势。
Treat the mismatch as a question, not an automatic accusation. Examine billing milestones, payment terms, receivable aging, inventory commitments, contract assets, supplier timing, taxes, and seasonality. Decide what timing pattern would reconcile the two and when it should appear.
需求信号很强但现金兑现较弱，应如何解释？把不匹配视为问题，而不是自动指控。检查开票里程碑、付款条件、应收账龄、存货承诺、合同资产、供应商时点、税项和季节性，并写出什么时间路径能够调节二者，以及它应在何时出现。
Only after measuring the owner outcome. Reconcile diluted and period-end shares, awards, conversions, repurchases, and issuance purpose. Aggregate profit can rise while profit per share falls, as the hypothetical distributor demonstrates.
总利润增长能否抵消股数上升？只有衡量股东实际结果后才能回答。需要调节稀释股数与期末股数、奖励、转换、回购和发行用途。正如虚构分销商示例，总利润上升时每股利润仍可能下降。
Record the disappearance, compare definitions and prior filings, and look for a replacement measure or explanation. Absence alone does not prove deterioration, but it reduces observability and should lower confidence until another source answers the same operating question.
如果公司停止披露一个有用经营指标怎么办？记录该指标消失，比较定义与历史披露，并寻找替代指标或解释。缺失本身不能证明恶化，但会降低可观察性；在其他来源回答同一经营问题前，应降低确信度。
Official sources and interpretation boundary官方来源与解释边界
These official U.S. sources describe common filing locations, financial statements, and stock-evaluation concepts. For a specific issuer, use the applicable home-market regulator, exchange, accounting framework, security terms, and current company filings. The relevance and definition of each signal vary by business model and reporting regime.
This signal stack prioritizes evidence; it is not individualized investment advice, a fixed score, or a prediction. Missing disclosure can limit confidence without proving misconduct, and even a coherent stack cannot eliminate operating, financing, valuation, or market risk.
以上美国官方资料说明了常见披露位置、财务报表和股票评价概念。研究具体发行人时，应使用适用的本土监管机构、交易所、会计框架、证券条款和公司当前披露。每类信号的重要性与定义都会随商业模式和报告制度变化。
这套信号栈用于教育性地安排证据优先级，不是个性化投资建议、固定评分模型或预测。披露缺失会限制确信度，但不能证明不当行为；矛盾可以确定研究优先级，却不会自动决定交易；即使信号栈相互一致，也不能消除经营、融资、估值或市场风险。