What fundamental analysis actually does基本面分析究竟在做什么
Fundamental analysis for beginners is the practice of translating a company’s business model into measurable drivers, checking those drivers against financial statements and notes, and comparing a range of justified outcomes with the assumptions embedded in a dated price. It evaluates a business and a valuation; it does not predict where the share price will move tomorrow.
对新手而言，基本面分析就是把商业模式翻译成可度量的经营驱动，用财务报表和附注核验这些驱动，再把一组有依据的结果与某一日期价格所隐含的假设比较。它评估的是企业与估值，不是预测股价明天往哪里走。
Write the company’s economic sentence先写出公司的经济句子
Begin without ratios. Write: “The company helps [customer] solve [costly problem], charges through [contract], must spend on [key resources], and keeps value when [retention or repeat-use condition] holds.” If you cannot complete that sentence from the business description, customer disclosures, and contract terms, an earnings multiple will not repair the missing understanding. The sentence also tells you which accounting lines matter and which popular metrics may be distractions.
先不要看比率，试着写完一句话：“公司帮助【哪类客户】解决【什么高成本问题】，通过【何种合同】收费，必须在【关键资源】上投入，并在【续约或重复使用条件】成立时保留价值。”如果读完业务说明、客户披露和合同条款后仍无法完成这句话，一个市盈率也无法弥补理解缺口。这句话还会告诉你哪些会计项目真正重要，哪些热门指标可能只是干扰。
VectorForge Systems sells monitoring software and maintenance support to industrial operators under multi-year contracts, plus implementation work. Its proposed economic engine is: reliable renewals fund product development; product improvements protect renewals and support expansion. All 20X1 and 20X2 amounts below are invented, stated in U.S. dollars, and refer to no real company or security.
贯穿全文的虚构公司VectorForge Systems 向工业运营商销售监测软件与维护支持，合同通常跨越多年，同时提供实施服务。它的待验证经济引擎是：稳定续约为产品开发提供资金，产品改进又保护续约并推动客户扩容。下文 20X1 和 20X2 的所有美元数据均为教学而虚构，不对应任何真实公司或证券。
Reduce unplanned equipment downtime and make maintenance decisions from operating data.
客户任务减少设备意外停机，并依据运行数据安排维护。
Recurring software and maintenance contracts, with implementation priced separately.
收费机制软件与维护合同形成经常性收费，实施服务单独定价。
Technical support, engineers, product development, computing, and implementation teams.
资源投入持续投入技术支持、工程师、产品开发、计算资源与实施团队。
The installed base must renew and expand enough to justify development and service costs.
经济检验已安装客户必须以足够比例续约和扩容，才能覆盖开发及服务成本。
Follow one contract through the accounting clock沿会计时间轴追踪一份合同
A signed contract, an invoice, a cash receipt, and recognized revenue can occur on different dates. Topic 606 centers revenue recognition on transferring promised goods or services, not simply on billing or collection. Read the revenue policy to identify performance obligations, allocation of transaction price, and whether each obligation is satisfied at a point in time or over time. Then read the contract-balance and remaining-performance-obligation disclosures. “Recurring contract value,” “bookings,” and “billings” may be management-defined and should never be silently substituted for reported revenue.
签署合同、开具发票、收到现金和确认收入可能发生在不同日期。Topic 606 的核心是何时把承诺的商品或服务转移给客户，而不是是否已经开票或收款。阅读收入政策时，要识别履约义务、交易价格如何分摊，以及每项义务是在某个时点还是一段时间内履行；随后再看合同余额与剩余履约义务披露。“经常性合同价值”“订单额”“账单额”可能是管理层自定义指标，不能悄悄替代报表收入。
| 20X2 item | Illustrative reported fact | What a beginner may infer | What remains unproven |
|---|---|---|---|
| Subscription and maintenance revenue | $268m for the year | A large share of recognized activity is recurring in label | Actual renewal rate or future cash |
| Implementation and initial-license revenue | $72m for the year | Services and initial delivery affect mix | That implementation is equally profitable or recurring |
| Contract liabilities | $118m opening; $146m closing | Cash or billing preceded some revenue recognition | Backlog quality, cancellation rights, or collection timing |
| Remaining performance obligations | $410m; 62% expected within 12 months | Some contracted revenue is scheduled beyond year-end | Renewals not yet contracted or revenue unaffected by termination clauses |
| 20X2 项目 | 虚构披露事实 | 新手可以推断什么 | 仍不能证明什么 |
|---|---|---|---|
| 订阅与维护收入 | 全年 2.68 亿美元 | 按标签看，较大部分已确认业务具有经常性 | 实际续约率或未来现金 |
| 实施与初始许可收入 | 全年 7,200 万美元 | 服务和初始交付会影响收入结构 | 实施服务同样盈利或能够重复 |
| 合同负债 | 期初 1.18 亿；期末 1.46 亿美元 | 部分收款或开票早于收入确认 | 积压质量、取消权或回款时点 |
| 剩余履约义务 | 4.10 亿美元；预计 62% 在十二个月内确认 | 部分已签约收入将在期后履行 | 尚未签署的续约，或合同终止条款不会影响收入 |
One useful reconciliation is a billing-activity approximation: Reported revenue + Closing contract liabilities − Opening contract liabilities. For VectorForge, $340m + $146m − $118m = $368m. Label this a calculation, not a reported billing figure. It becomes unreliable when contract assets, acquisitions, foreign exchange, refunds, or classification changes are material. The $28m increase in deferred revenue is not extra revenue; it is a timing bridge that must be read with the notes.
可以做一项开票活动近似计算：已报告收入 + 期末合同负债 − 期初合同负债。VectorForge 的结果为 3.40 亿 + 1.46 亿 − 1.18 亿 = 3.68 亿美元。它是研究者计算，不是公司已披露的账单额；当合同资产、收购、汇率、退款或分类变化重大时，这个近似会失真。递延收入增加 2,800 万美元并不是额外收入，而是需要结合附注解释的时间桥。
Make the installed base explain the growth让存量客户解释增长来自哪里
The fictional contract-value bridge starts with $250m at the opening of 20X2. Contracts worth $25m are lost at renewal, price changes add $8m, existing customers expand by $17m, and new customers add $40m. Ending annualized recurring contract value is therefore $250m − $25m + $8m + $17m + $40m = $290m. This measure is illustrative and management-defined; its scope, currency treatment, acquisition effects, and treatment of temporary usage must be checked before any comparison.
虚构的合同价值桥从 20X2 年初 2.50 亿美元开始：续约时流失 2,500 万，提价增加 800 万，现有客户扩容增加 1,700 万，新客户增加 4,000 万，因此期末年化经常性合同价值为 2.50 亿 − 2,500 万 + 800 万 + 1,700 万 + 4,000 万 = 2.90 亿美元。该指标属于教学性的管理层定义值；比较前必须核对范围、汇率处理、收购影响及临时用量的处理方法。
| Calculation | Formula | 20X2 result | Interpretation |
|---|---|---|---|
| Gross dollar retention | (Opening recurring value − churn) ÷ opening value | ($250m − $25m) ÷ $250m = 90% | Ten percent of opening value was lost before expansion or price |
| Net dollar retention | (Opening value − churn + price + existing-customer expansion) ÷ opening value | $250m ÷ $250m = 100% | Price and expansion only offset lost value |
| Total recurring-value growth | (Ending value ÷ opening value) − 1 | ($290m ÷ $250m) − 1 = 16% | The increase depends on $40m from new customers |
| 计算 | 公式 | 20X2 结果 | 解释 |
|---|---|---|---|
| 总金额留存率 | （期初经常性价值 − 流失）÷ 期初价值 | （2.50 亿 − 2,500 万）÷ 2.50 亿 = 90% | 在提价和扩容前，期初价值流失了 10% |
| 净金额留存率 | （期初价值 − 流失 + 提价 + 存量扩容）÷ 期初价值 | 2.50 亿 ÷ 2.50 亿 = 100% | 提价与扩容只是抵消了流失 |
| 经常性价值总增长 | （期末价值 ÷ 期初价值）− 1 | （2.90 亿 ÷ 2.50 亿）− 1 = 16% | 增长依赖新客户贡献的 4,000 万美元 |
Reported fact: the fictional bridge components. Calculation: 90% gross retention, 100% net retention, and 16% total growth. Assumption: contract value represents comparable recurring scope. Judgment: new selling remains important because existing-customer price and expansion merely replace churn. Missing evidence: customer counts, cohort retention, discounting, contract length, and concentration.
结论应比标题数字更窄已披露事实：虚构桥接中的各组成项。计算：总金额留存率 90%、净金额留存率 100%、总增长 16%。假设：合同价值采用可比的经常性范围。判断：存量客户提价与扩容只抵消流失，因此新客户销售仍很重要。缺失证据：客户数量、批次留存、折扣、合同期限和客户集中度。
Treat R&D as a claim about future usefulness把研发视为对未来用途的主张
VectorForge reports $54m of research and development expense in 20X2 and $46m in 20X1, while total revenue rises from $305m to $340m. R&D grew 17.4%: ($54m ÷ $46m) − 1. R&D intensity rose from 15.1% ($46m ÷ $305m) to 15.9% ($54m ÷ $340m). Those are facts and calculations, not proof that innovation improved. Read the accounting policy for which software costs are expensed or capitalized, and keep differently treated development spending visible.
VectorForge 在 20X2 报告研发费用 5,400 万美元，20X1 为 4,600 万美元；同期总收入从 3.05 亿增至 3.40 亿美元。研发费用增长 17.4%，即 5,400 ÷ 4,600 − 1；研发强度从 15.1%（4,600 ÷ 30,500）升至 15.9%（5,400 ÷ 34,000）。这些只是事实和计算，不能证明创新已经改善。还要阅读会计政策，确认哪些软件成本被费用化或资本化，并让处理方式不同的开发支出保持可见。
- Look for product releases and adoption that solve the customer job described in the economic sentence, not a long feature list.
- Ask whether renewal, expansion, support burden, and gross margin improve after the spending, allowing for a realistic lag.
- Separate maintenance engineering, new-product work, acquired technology, and capitalized software when disclosure permits.
- Check share-based compensation and diluted shares; engineering investment paid with equity still has an owner cost.
- 寻找真正解决经济句子中客户任务的产品发布与采用证据，而不是只数功能清单。
- 在考虑合理滞后后，观察研发投入之后的续约、扩容、支持负担和毛利率是否改善。
- 披露允许时，区分维护性工程、新产品、收购技术与资本化软件。
- 核对股份支付与稀释股数；用股权支付工程师仍然会给所有者带来成本。
A beginner should resist both shortcuts: “high R&D means a moat” and “high R&D means waste.” The supportable statement is conditional. Spending is more credible when it precedes measurable customer value, renewals, or product economics; it is less credible when costs rise while retention, release delivery, and competitive position deteriorate. The filing rarely resolves this in one period, so note the lag you expect and the observation date that would test it.
新手应同时抵制两种捷径：“研发高就有护城河”和“研发高就是浪费”。有证据的表述必须带条件：如果投入之后出现可度量的客户价值、续约或产品经济改善，研发主张更可信；如果成本上升，而留存、交付和竞争位置持续恶化，可信度就下降。单一期间通常无法解决这个问题，因此要写明预期滞后，以及未来用来检验它的观察日期。
Translate the price into an operating hurdle把价格翻译成经营门槛
Fundamental analysis reaches price by asking what must become true, not by forecasting the next quote. Use clearly dated market inputs. In a fictional snapshot dated December 31, 20X2, VectorForge has an equity value of $1.80bn, debt of $0.20bn, and cash of $0.30bn. Enterprise value = Equity value + Debt − Cash = $1.70bn. These values are assumptions for the lesson, not current market data.
基本面分析通过询问“什么必须成真”连接到价格，而不是预测下一次报价。市场输入必须注明日期。在虚构的 20X2 年 12 月 31 日快照中，VectorForge 股权价值 18.0 亿美元、债务 2.0 亿、现金 3.0 亿，因此企业价值 = 股权价值 + 债务 − 现金 = 17.0 亿美元。这些都是教学假设，不是当前市场数据。
- Assume a required annual return of 10% for five years. The year-five enterprise-value hurdle is $1.70bn × (1.10)^5 = $2.738bn.
- Assume the business could be valued at 20 times year-five free cash flow. Required year-five free cash flow is $2.738bn ÷ 20 = $136.9m.
- Against illustrative normalized current free cash flow of $32m, the required five-year CAGR is ($136.9m ÷ $32m)^(1/5) − 1 = 33.7%.
- At an assumed 18% year-five free-cash-flow margin, required revenue is $136.9m ÷ 18% = $760.5m, implying 17.5% annual growth from $340m.
- 假设未来五年要求年回报率为 10%，第五年企业价值门槛为 17.0 亿 ×（1.10）^5 = 27.38 亿美元。
- 假设第五年企业价值可以按当年自由现金流 20 倍计量，则所需第五年自由现金流为 27.38 亿 ÷ 20 = 1.369 亿美元。
- 相对于虚构的当前标准化自由现金流 3,200 万美元，所需五年复合增速为（1.369 亿 ÷ 3,200 万）^(1/5) − 1 = 33.7%。
- 若假设第五年自由现金流率为 18%，所需收入为 1.369 亿 ÷ 18% = 7.605 亿美元，相比当前 3.40 亿意味着每年增长 17.5%。
The calculation assumes a 10% return, a 20× future multiple, an 18% future cash margin, unchanged net debt, and no value from interim distributions or retained cash. Change any assumption and the hurdle moves. Its purpose is to ask whether 90% gross retention, 100% net retention, new-customer dependence, and the R&D evidence can plausibly support the required growth and margin together.
这是反向门槛，不是内在价值计算假设了 10% 要求回报、第五年 20 倍自由现金流、18% 自由现金流率、净债务不变，并忽略期间分配或留存现金的价值。任何假设变化都会移动门槛。它的用途，是检验 90% 总金额留存、100% 净金额留存、对新客户的依赖和研发证据，能否共同支持所要求的增长与利润率。
End with a four-layer beginner memo以四层新手备忘录收尾
| Layer | VectorForge entry | Required language |
|---|---|---|
| Reported | Revenue mix, contract balances, renewal bridge, R&D expense | “The company reports…” with period and units |
| Calculated | Billing proxy, retention, R&D intensity, reverse price hurdle | Show formula and inputs |
| Assumed | Comparable contract scope, return, future multiple and margin | “The analysis assumes…” |
| Judged | Growth depends heavily on new sales; price hurdle appears demanding | State competing explanation and confidence |
| Missing | Cohort retention, customer concentration, development outcomes | Name the next disclosure or date |
| 层级 | VectorForge 记录 | 应使用的措辞 |
|---|---|---|
| 已披露 | 收入结构、合同余额、续约桥、研发费用 | 使用“公司报告……”并注明期间与单位 |
| 已计算 | 开票近似、留存、研发强度、反向价格门槛 | 展示公式与输入 |
| 假设 | 合同范围可比、要求回报、未来倍数与利润率 | 使用“本分析假设……” |
| 判断 | 增长高度依赖新销售；价格门槛看起来较高 | 写出竞争性解释与确信程度 |
| 缺失 | 客户批次留存、集中度、开发成果 | 指定下一份披露或日期 |
The provisional conclusion is not “up” or “down.” It is: VectorForge has a visible recurring contract base, but the example’s existing-customer economics only offset churn, while the price screen requires strong cash-flow growth and margin improvement. Evidence would strengthen if cohort retention rises without heavier discounting and R&D outputs reduce support costs or expand customer use. It would weaken if deferred revenue growth comes from longer billing terms while gross retention falls. Revisit the memo when those observations change.
暂时结论不是“涨”或“跌”，而是：VectorForge 拥有可观察的经常性合同基础，但案例中的存量客户经济性只是抵消流失，而价格初筛要求现金流快速增长并提高利润率。如果客户批次留存在没有加大折扣的情况下上升，且研发产出降低支持成本或扩大客户使用，证据会加强；如果递延收入增长只是来自更长开票期限，同时总金额留存下降，证据会减弱。上述观察变化时再更新备忘录。
Use uploads to reconcile the business story用上传文件核对商业叙述
For this analysis, prepare the complete annual filing, the latest interim filing, earnings release, revenue-policy note, contract-balance and remaining-performance-obligation tables, R&D accounting policy, and the dated market inputs used for valuation. Include both 20X1 and 20X2 equivalents when definitions are comparable. State the question as a test, such as: “Do renewal economics and product investment support the growth and margin required by the price?”
Ask Stock Explained to separate reported figures, derived ratios, assumptions, interpretations, and missing items. Afterward, verify each material output against the uploaded page: contract-value scope, opening and closing balances, period, units, retention denominator, R&D classification, debt, cash, and diluted share count. Recalculate the renewal bridge and reverse hurdle independently. If the tool cannot point to the underlying table or note, keep that statement out of the conclusion.
为本页分析准备完整年度披露、最新中期披露、业绩新闻稿、收入政策附注、合同余额与剩余履约义务表、研发会计政策，以及估值所用的带日期市场输入。在定义可比时，同时加入 20X1 与 20X2 对应资料。把问题写成检验句，例如：“续约经济与产品投入是否支持价格所要求的增长和利润率？”
要求 Stock Explained 分开呈现已披露数字、推导比率、假设、解释和缺失项。生成后回到上传页面核验每项重要输出：合同价值范围、期初期末余额、期间、单位、留存率分母、研发分类、债务、现金和稀释股数。独立重算续约桥与反向门槛。若无法定位到原始表格或附注，就不要把该表述写入结论。
Upload complete, comparable source pages and ask one explicit question; keep every original open while you verify the resulting evidence chain.
上传完整且可比的来源页面，并提出一个明确问题；核验证据链时始终打开原始文件。
Open Stock Explained打开“一眼看懂这只股票”Beginner questions after the first model完成第一个模型后的新手问题
No. It estimates business economics and tests whether a dated price requires plausible outcomes. Short-term prices can move for reasons the operating model does not forecast, while even a correct business view can be paired with an unattractive purchase price.
基本面分析等同于预测股价吗？不等同。它评估企业经济性，并检验某一日期价格要求的结果是否合理。短期价格可能因经营模型无法预测的原因波动；即使对企业判断正确，买入价格也可能缺乏吸引力。
No. It may reflect growth, advance billing, longer contract terms, acquisitions, currency, or classification changes. Reconcile opening and closing contract balances with revenue, contract assets, remaining obligations, billing terms, and cancellations before interpreting the movement.
递延收入上升一定是好信号吗？不一定。它可能来自增长、提前开票、更长合同期限、收购、汇率或分类变化。解释前，应把期初期末合同余额与收入、合同资产、剩余履约义务、开票条款及取消情况一起核对。
Use both when available. Gross retention shows what remains before price and expansion; net retention includes price and existing-customer expansion. Confirm whether the denominator uses customers, revenue, or annualized contract value and whether acquisitions or currency are excluded.
新手应该使用哪一种留存率？资料允许时两者都看。总金额留存率显示提价和扩容前还剩多少，净金额留存率则包含提价与存量客户扩容。必须确认分母采用客户数、收入还是年化合同价值，以及是否排除收购或汇率。
You cannot tell from the expense alone. Define an expected lag, then look for relevant product adoption, retention, expansion, support efficiency, gross margin, and competitive evidence. Check capitalized development and stock compensation so the spending boundary stays consistent.
怎样判断更高研发投入是否有效？单凭费用无法判断。先定义合理滞后，再观察相关产品采用、留存、扩容、支持效率、毛利率和竞争证据；同时核对资本化开发支出与股份支付，保持投入边界一致。
Not initially. A transparent reverse hurdle can reveal which growth, margin, return, and future-multiple assumptions matter. Build a fuller model only when you can explain the business drivers, avoid false precision, and test several internally consistent scenarios.
新手一开始就需要完整现金流折现模型吗？不需要。透明的反向门槛已经能够揭示增长、利润率、要求回报和未来倍数中哪些假设最重要。只有在能解释经营驱动、避免虚假精确并检验多组内部一致情景时，再建立更完整模型。
Accounting and filing sources会计与披露来源
The sources explain U.S. public-company filing access, financial-statement structure, revenue recognition principles, and cautions around company-defined measures. They do not validate VectorForge, whose facts are entirely illustrative. For a real issuer, use its latest original filings, auditor report, amendments, and home-market accounting framework.
This material is educational, not individualized investment advice or a return forecast. It omits personal objectives, portfolio exposure, tax, liquidity, and risk capacity. Accounting evidence can narrow uncertainty, but it cannot eliminate business failure, model error, valuation risk, or market loss.
以上来源说明美国公众公司披露的查找方式、财务报表结构、收入确认原则以及公司自定义指标的注意事项。它们并不验证 VectorForge；该公司及数据完全是教学示例。分析真实发行人时，应使用其最新原始披露、审计报告、修订文件和所属市场会计框架。
本文仅供学习，不构成个性化投资建议或收益预测，也没有纳入个人目标、组合暴露、税务、流动性与风险承受力。会计证据可以缩小不确定性，却无法消除经营失败、模型错误、估值风险或市场亏损。