The denominator changes from recorded to expected earnings分母从已记录盈利切换为预期盈利
Trailing P/E divides a dated price by EPS for a completed period, often the latest four quarters. Forward P/E uses forecast EPS for a named future window. A lower forward P/E means forecast EPS is higher, not that the stock is necessarily cheaper. Align earnings basis, fiscal window, diluted shares and estimate date, then test the forecast bridge.
滚动市盈率用注明日期的股价除以已完成历史期间的 EPS，常见口径是最近四个已报告季度；预期市盈率则用同一价格除以明确未来窗口的预测 EPS。因此，预期 P/E 低于滚动 P/E，只表示所选预测 EPS 高于历史 EPS，并不自动表示股票更便宜。必须对齐盈利口径、财政期间、稀释股数与估计日期，再解释预测桥并在后续检验其兑现。
Hold price still and label both earnings clocks固定价格，并标注两只盈利时钟
| Comparison field | Trailing P/E | Forward P/E | Required label |
|---|---|---|---|
| Core formula | Price on date d ÷ historical EPS | Price on the same date d ÷ forecast EPS | Price, currency, market and timestamp |
| Earnings window | Often TTM: latest four reported quarters | NTM, FY1, FY2 or another named period | Exact start/end dates and actual-versus-estimate mix |
| Evidence status | Reported accounting result, unless explicitly adjusted | Estimate from guidance, consensus or a research model | Source, version date and definitions |
| Share basis | Historical weighted-average basic or diluted shares used in EPS | Expected diluted shares for the forecast period | Treatment of awards, options and convertibles |
| Main limitation | Observable but possibly stale after a structural change | Potentially timely but dependent on assumptions | Reason the denominator could mislead |
| 比较字段 | 滚动 P/E | 预期 P/E | 必填标签 |
|---|---|---|---|
| 核心公式 | d 日股价 ÷ 历史 EPS | 同一 d 日股价 ÷ 预测 EPS | 价格、币种、市场与时间戳 |
| 盈利窗口 | 通常为 TTM：最近四个已报告季度 | NTM、FY1、FY2 或其他明确期间 | 准确起止日及实际数/估计数占比 |
| 证据状态 | 除非明确调整，否则为已报告会计结果 | 来自公司指引、一致预期或研究模型的估计 | 来源、版本日与定义 |
| 股数口径 | 历史 EPS 采用的加权平均基本或稀释股数 | 预测期预期稀释股数 | 股份奖励、期权与可转债处理 |
| 主要限制 | 可观察，但在结构变化后可能滞后 | 较及时，但依赖假设 | 分母可能误导的原因 |
TTM is a rolling window, not necessarily the latest fiscal year. On a date after one interim report, it may combine one quarter from the new fiscal year with three from the old one. NTM is the next twelve months from the estimate date. FY1 normally refers to the next or current fiscal-year estimate according to the data provider’s convention, and it can contain both actual and forecast quarters. Calendar-year EPS is another window. Never use those labels interchangeably.
TTM 是滚动窗口，不一定等于最近一个完整财政年度。在某份中期报告发布后，它可能把新财年的一个季度与旧财年的三个季度相加。NTM 是从估计日起向后十二个月；FY1 通常指数据来源约定的当前或下一个财政年度估计，可能混合实际季度和预测季度；自然年 EPS 又是另一窗口。这些标签不能互换。
If either denominator is zero or negative, the P/E is undefined or economically difficult to interpret. A negative multiple should not be ranked as if it were lower and therefore better. Explain the loss, the forecast path and the uncertainty instead of forcing a P/E comparison.
EPS 为零或负数时，常规比较停止若任一分母为零或负数，P/E 无定义或难以作经济解释。不能把负倍数像普通正数一样排序，并称其更低、更好。此时应解释亏损、预测转盈路径与不确定性，而不是强行比较市盈率。
Give every EPS denominator a passport给每个 EPS 分母配一张“护照”
A displayed P/E is incomplete without an EPS passport. “Forward 12×” can refer to a vendor consensus for the next fiscal year, an analyst’s NTM model, company guidance converted into EPS, or an adjusted figure that excludes selected costs. Two screens can be arithmetically correct and still disagree because they use different passports.
没有 EPS 护照，页面上显示的 P/E 就是不完整信息。“预期 12×”可能指数据商的下一财年一致预期、分析者的 NTM 模型、由公司指引换算的 EPS，或剔除特定成本后的调整数字。两个页面即使算术都正确，也可能因护照不同而得出不同结果。
Write TTM ending date, forecast start and end, fiscal year-end, and which quarters are reported versus estimated. For a seasonal company, one shifted quarter can move most of the annual earnings.
期间身份写明 TTM 截止日、预测起止日、财政年末，以及哪些季度是实际数、哪些是估计数。对季节性公司而言，错移一个季度就可能移动全年大部分盈利。
Label reported GAAP or IFRS EPS separately from adjusted EPS. List each exclusion and its tax effect. Historical adjusted EPS and forecast adjusted EPS must use the same policy before their multiples are compared.
会计身份把已报告 GAAP 或 IFRS EPS 与调整后 EPS 分开。列明每个剔除项目及税务影响。比较倍数前，历史调整 EPS 与预测调整 EPS 必须采用同一政策。
State whether the source is a company range, a dated consensus snapshot or a self-built model. Preserve contributor count or dispersion when available. Consensus is an aggregation, not a disclosed company fact.
预测身份注明来源是公司区间、注明日期的一致预期快照，还是自建模型。在可获得时保留贡献者数量或分歧范围。一致预期是估计汇总，不是公司披露事实。
Historical diluted EPS uses a period-weighted denominator. Forward EPS requires a forecast share count that reflects expected issuance, repurchases and dilution. Current shares outstanding are not automatically either denominator.
股数身份历史稀释 EPS 使用期间加权分母；预期 EPS 则需要反映预计发行、回购和稀释的预测股数。当前发行在外股数不会自动等于任何一个分母。
Build an estimate version ladder, not one timeless number建立估计版本梯，而不是保存一个无日期数字
Forward EPS expires as information changes. Preserve versions so the research record can distinguish a genuinely improving forecast from a denominator that merely moved after bad news. The ladder below is not a reliability hierarchy: each rung has a different role, and company guidance may not supply enough items to calculate EPS without additional assumptions.
随着信息变化，预期 EPS 会过期。保留版本，才能区分预测真正改善与坏消息后分母被下调。下列梯子不是可靠性排名：每一级承担不同作用，而且公司指引未必提供足够项目直接计算 EPS，分析者仍可能需要补充假设。
| Rung | Record | What must remain visible |
|---|---|---|
| 0 — reported base | Latest quarterly EPS and TTM sum | Filing date, restatements, diluted denominator and accounting basis |
| 1 — company outlook | Revenue, margin or EPS range if actually provided | Issue date, covered period, stated basis and omitted variables |
| 2 — consensus snapshot | Aggregated estimate as of one timestamp | Provider, fiscal mapping, contributor context and reported/adjusted basis |
| 3 — research model | Driver bridge from volume and price to EPS | Every assumption, formula, share count and change from the previous version |
| 4 — outcome check | Actual result and driver delivery | Comparable scope, forecast error, explanation and effect on the next estimate |
| 层级 | 记录内容 | 必须保持可见的信息 |
|---|---|---|
| 0—报告基础 | 最近季度 EPS 与 TTM 合计 | 披露日、重述、稀释分母和会计口径 |
| 1—公司展望 | 公司实际提供的收入、利润率或 EPS 区间 | 发布日期、覆盖期间、既定口径和未提供变量 |
| 2—一致预期快照 | 某一时间戳的汇总估计 | 提供方、财政期间映射、贡献者背景和报告/调整口径 |
| 3—研究模型 | 从销量与价格到 EPS 的驱动桥 | 每项假设、公式、股数及相对上一版的变化 |
| 4—结果核查 | 实际结果与驱动因素兑现 | 可比范围、预测误差、解释及对下一版估计的影响 |
- Never overwrite an old estimate. Add a new dated row and explain the revision.
- Distinguish revisions caused by new operating information from mechanical changes in tax, interest or shares.
- When a provider rolls from FY1 to FY2, record the rollover; a sudden multiple change may be a period change, not new economics.
- Keep the price date fixed when isolating an EPS revision. Update price separately when answering a new valuation-date question.
- 不要覆盖旧估计；新增注明日期的一行，并解释修订原因。
- 区分由新经营信息引起的修订，与税率、利息或股数机械变化引起的修订。
- 当数据源从 FY1 滚动到 FY2 时，记录滚动过程；倍数突然变化可能只是期间切换，而不是经济性变化。
- 单独分析 EPS 修订时保持价格日期不变；只有回答新的估值日问题时，才另行更新价格。
FrostArc: why the forward multiple fallsFrostArc：预期倍数为何下降
FrostArc Equipment is a completely fictional manufacturer of snow-maintenance machines with a March 31 fiscal year-end. All prices, results and forecasts are illustrative; they do not describe a security or current consensus. The comparison date is August 20, 20X5, after the June quarter was reported. Hold the hypothetical price at CU36.00 throughout.
FrostArc Equipment 是一家完全虚构的除雪维护设备制造商，财政年度截止日为 3 月 31 日。全部价格、业绩和预测只用于演示，不对应任何证券或当前一致预期。比较日为 20X5 年 8 月 20 日，当时 6 月季度已经披露；全文固定使用假设价格 CU36.00。
| Window | Quarterly EPS components | EPS | P/E at CU36 |
|---|---|---|---|
| TTM ended June 20X5 — reported diluted | Sep 0.20 + Dec 1.10 + Mar 0.85 + Jun 0.25 | CU2.40 | 15.00× |
| TTM — analyst-adjusted warranty view | CU2.40 + CU0.45 after-tax warranty adjustment | CU2.85 | 12.63× |
| FY20X6 — blended | Jun actual 0.25 + Sep estimate 0.30 + Dec 1.35 + Mar 1.15 | CU3.05 | 11.80× |
| NTM ended June 20X6 — forecast diluted | Sep 0.30 + Dec 1.35 + Mar 1.15 + Jun 0.50 | CU3.30 | 10.91× |
| 窗口 | 季度 EPS 组成 | EPS | CU36 对应 P/E |
|---|---|---|---|
| 截至 20X5 年 6 月 TTM—已报告稀释口径 | 9 月 0.20 + 12 月 1.10 + 3 月 0.85 + 6 月 0.25 | CU2.40 | 15.00× |
| TTM—分析者调整保修费用口径 | CU2.40 + 税后保修调整 CU0.45 | CU2.85 | 12.63× |
| FY20X6—实际与预测混合 | 6 月实际 0.25 + 9 月估计 0.30 + 12 月 1.35 + 3 月 1.15 | CU3.05 | 11.80× |
| 截至 20X6 年 6 月 NTM—预测稀释口径 | 9 月 0.30 + 12 月 1.35 + 3 月 1.15 + 6 月 0.50 | CU3.30 | 10.91× |
Calculation: trailing P/E = CU36.00 ÷ CU2.40 = 15.00×. Forward NTM P/E = CU36.00 ÷ CU3.30 = 10.91×. Because price is unchanged, the entire multiple decline comes from forecast EPS being 37.5% above reported TTM EPS: CU3.30 ÷ CU2.40 − 1. After adjusting the fictional one-time warranty charge, the implied improvement is only CU3.30 ÷ CU2.85 − 1 = 15.8%. The comparison has not proved which normalization is appropriate; it has located the question.
计算：滚动 P/E = CU36.00 ÷ CU2.40 = 15.00×；预期 NTM P/E = CU36.00 ÷ CU3.30 = 10.91×。由于价格不变，倍数下降全部来自预测 EPS 比已报告 TTM EPS 高 37.5%，即 CU3.30 ÷ CU2.40 − 1。若调整虚构的一次性保修费用，隐含改善仅为 CU3.30 ÷ CU2.85 − 1 = 15.8%。这项比较尚未证明哪种正常化正确，只是定位了需要回答的问题。
Here FY20X6 contains one reported and three forecast quarters, while NTM contains four forecast quarters and ends three months later. Calling both “forward EPS” conceals CU0.25 of denominator difference and mixes forecast coverage. Show both labels instead of selecting the lower multiple.
FY1 不自动等于 NTM本例 FY20X6 包含一个已报告季度和三个预测季度，而 NTM 包含四个预测季度，截止日还晚三个月。把二者都称为“预期 EPS”会掩盖 CU0.25 的分母差异，并混淆预测覆盖度。应同时展示标签，而不是挑选较低倍数。
Trace the forecast bridge and score delivery追踪预测桥，并评价兑现情况
FrostArc’s NTM bridge starts with CU24.0m TTM net income. It adds CU4.5m for warranty normalization, CU5.0m for backlog, CU2.0m for price and mix, and CU2.5m for ramp savings; additional depreciation removes CU1.0m and financing and tax remove CU2.35m. Forecast net income is CU34.65m; divided by 10.5m expected diluted shares, EPS is CU3.30. Using 10.0m historical shares would misleadingly produce CU3.465 and 10.39×.
FrostArc 的虚构 NTM 模型从已报告 TTM 净利润 CU24.0m 出发：税后保修项目加回 CU4.5m，积压订单转化增加 CU5.0m，价格与组合增加 CU2.0m，产能爬坡节省 CU2.5m；新增折旧减少 CU1.0m，融资及税务影响减少 CU2.35m。预测净利润为 CU34.65m；除以预期稀释股数 10.5m，得到 EPS CU3.30。若错误沿用历史 10.0m 股本，EPS 会变成 CU3.465，P/E 会被压低至 10.39×。
The charge is a fictional fact; non-recurrence is an assumption. Check product cohorts, claim frequency, reserve movement and new exposure. “One-time” is a forecast claim.
保修正常化虚构文件中的事实是发生了费用；假设则是该水平不会重复。需要检查产品批次、索赔率、准备金变动以及当前销售是否形成新风险。“一次性”本身是一项预测主张。
Backlog is not earnings. Check cancellation, delivery dates, pricing, inputs and expected margin. A rising balance can coexist with delayed revenue or weak profit.
积压订单转化积压订单不是盈利。应检查取消权、交付日期、价格调整、投入品可得性与预期利润率。订单余额上升仍可能伴随收入延迟或利润薄弱。
Savings depend on throughput, scrap, labor learning and uptime; depreciation may rise first. Show both the expected benefit and commissioning cost.
产能爬坡节省取决于产量、废品率、劳动学习和开工率，而折旧可能早于产出上升。预测桥应同时展示预期收益与产能投运成本。
| Prior estimate for a completed fictional quarter | Estimate | Actual | Delivery status | Next question |
|---|---|---|---|---|
| Backlog converted | CU14.0m | CU12.6m | Partial: 90.0% | Timing delay or cancellation? |
| New-line uptime | 85% | 78% | Missed by 7 percentage points | Commissioning issue or lasting bottleneck? |
| Net price change | +3.0% | +3.2% | Met | Volume or mix offset? |
| Diluted shares | 10.2m | 10.3m | Missed: 1.0% higher | Awards, issuance or averaging? |
| Diluted EPS | CU0.30 | CU0.25 | Absolute error 16.7% | Which driver explains the miss? |
| 虚构已完成季度的先前估计 | 估计 | 实际 | 兑现状态 | 下一个问题 |
|---|---|---|---|---|
| 积压订单转化 | CU14.0m | CU12.6m | 部分兑现：90.0% | 是时点延迟还是取消？ |
| 新产线开工率 | 85% | 78% | 低 7 个百分点 | 投运问题还是持续瓶颈？ |
| 净价格变化 | +3.0% | +3.2% | 兑现 | 销量或组合是否抵消？ |
| 稀释股数 | 10.2m | 10.3m | 未兑现：高 1.0% | 股份奖励、发行还是加权平均？ |
| 稀释 EPS | CU0.30 | CU0.25 | 绝对误差 16.7% | 哪项驱动造成差额？ |
Absolute forecast error is |CU0.25 − CU0.30| ÷ |CU0.30| = 16.7%. One seasonal quarter is not a permanent score. Keep driver definitions across versions, record whether misses reverse or compound, and separate timing from lost orders. The scorecard tests forward-EPS assumptions; it does not certify forecast skill.
本例绝对预测误差 = |CU0.25 − CU0.30| ÷ |CU0.30| = 16.7%。不能用一个季节性季度给分析者永久打分。应在多个估计版本中保持驱动定义一致，记录偏差是反转还是累积，并区分时点偏差与订单流失。兑现评分卡检验的是预期 EPS 背后的假设，不是对管理层或预测能力的认证。
Use the gap as a question, not a verdict把倍数差距当成问题，而不是结论
| Observed pattern at one price | Possible explanation | Evidence that separates explanations |
|---|---|---|
| Forward P/E below trailing P/E | Recovery, normalization, seasonality shift, optimistic forecast or denominator mismatch | Quarter bridge, adjustment policy, estimate revisions and delivery history |
| Forward P/E near trailing P/E | Stable earnings—or offsetting improvements and dilution | Net-income bridge, share schedule and driver sensitivity |
| Forward P/E above trailing P/E | Expected decline, unusually strong trailing period or conservative forecast | Backlog, pricing, cost, guidance and cycle evidence |
| Peers show different gaps | Different fiscal calendars, forecast bases, cycles or accounting adjustments | Rebuild each denominator on one window and policy |
| 同一价格下的观察模式 | 可能解释 | 区分解释所需证据 |
|---|---|---|
| 预期 P/E 低于滚动 P/E | 复苏、正常化、季节错位、预测乐观或分母不一致 | 季度桥、调整政策、估计修订与兑现历史 |
| 预期 P/E 接近滚动 P/E | 盈利稳定，或改善被稀释抵消 | 净利润桥、股数表与驱动敏感性 |
| 预期 P/E 高于滚动 P/E | 预计下滑、滚动期异常强或预测保守 | 积压订单、定价、成本、指引与周期证据 |
| 同业差距不同 | 财政日历、预测口径、周期或会计调整不同 | 按同一窗口和政策重建各分母 |
Trailing P/E can anchor the historical record; forward P/E can express a dated operating thesis. Neither shows cash conversion, balance-sheet resilience, required reinvestment or the probability distribution around EPS. A disciplined conclusion names which denominator is more decision-relevant, states why, presents the other as a control, and records the next result that can disprove the forecast.
滚动 P/E 可以锚定历史记录；预期 P/E 可以表达注明日期的经营假说。二者都不会展示现金转换、资产负债表韧性、必要再投资或 EPS 的概率分布。严谨结论应说明哪个分母与当前问题更相关及其原因，把另一个作为对照，并记录下一项能够证伪预测的结果。
At a fixed price, lower forward P/E is simply the reciprocal effect of higher forecast EPS. Whether the price offers value depends on the durability, risk and capital needs of that earnings path, plus evidence outside this two-number comparison.
不要把更低的预期倍数直接称为“便宜”在固定价格下，更低的预期 P/E 只是更高预测 EPS 所产生的倒数效应。价格是否具有价值，取决于盈利路径的持久性、风险与资本需求，也取决于这两个数字之外的证据。
Audit both denominators against uploaded evidence用上传证据审计两个分母
InfiniSynapse is this website’s analysis tool. Upload annual and interim filings, earnings releases, company outlook, share notes and a dated estimate sheet. Ask it to place four reported TTM quarters beside four forecast NTM quarters and tag source, accounting basis and actual-versus-estimate status. Do not treat it as a source of unverified live prices or current consensus.
Request bridges for reported-to-adjusted EPS, TTM-to-forward net income and historical-to-forecast diluted shares. Preserve every estimate version. Check material output against uploaded filings, guidance and assumptions, including fiscal dates, tax, units and denominators. A plausible EPS can still combine incompatible quarters or definitions.
InfiniSynapse 是本网站提供的分析工具。上传年度与中期报告、业绩公告、实际采用的公司展望原文、股份支付与资本结构附注，以及注明日期的估计表。要求工具把四个已报告 TTM 季度与四个预测 NTM 季度并列，并为每个单元格标注来源、会计口径及实际/估计状态。不要把工具当作未经核验的实时价格或当前一致预期来源。
要求分别建立“已报告到调整 EPS”“TTM 到预期净利润”以及“历史到预测稀释股数”桥。保留每个估计版本，不要覆盖旧数字。对照上传披露、指引原文和假设表核查全部重大输出，包括财政日期、税务处理、单位和股数分母。引用复核不可省略，因为看似合理的 EPS 仍可能拼接了不兼容的季度或定义。
Provide complete filings, outlook text, share notes and each estimate vintage; require source locations for every quarter and adjustment.
提供完整披露、展望原文、股本附注和每个估计版本，并要求为每个季度与调整项标出来源位置。
Open Stock Explained打开“一眼看懂这只股票”Questions readers ask next读者接下来常问的问题
No. A TTM denominator normally sums the latest four reported quarters, which may cross fiscal years. Some services display the latest annual EPS instead. Read the period label and reconstruct the quarters before comparing the number.
滚动 P/E 是否总是基于最近财政年度？不是。TTM 分母通常合计最近四个已报告季度，可能跨越两个财政年度；某些服务则展示最近年度 EPS。比较前要阅读期间标签，并重建季度组成。
Not necessarily. FY1 follows a fiscal-year convention and may mix actual and estimated quarters. NTM covers the next twelve months from the estimate date. They match only when their start and end dates and estimate coverage match.
FY1 P/E 与 NTM P/E 是否相同？不一定。FY1 遵循财政年度约定，可能混合实际与估计季度；NTM 则覆盖估计日起的未来十二个月。只有起止日期与估计覆盖完全一致时，两者才相同。
Use the source that matches the research question, but label it. Company guidance may be a range or omit interest, tax and shares; consensus aggregates external estimates; a self-built EPS model exposes assumptions. None becomes fact merely by being widely displayed.
预期 P/E 应使用公司指引还是一致预期 EPS？选择与研究问题匹配的来源，并清楚标注。公司指引可能是区间，也可能未提供利息、税与股数；一致预期汇总外部估计；自建 EPS 模型能公开假设。任何数字都不会因广泛展示而自动成为事实。
The arithmetic is possible, but the comparison mixes policies. Rebuild either reported-to-reported or adjusted-to-adjusted EPS, listing exclusions and tax effects. Show the mixed result only to explain why a vendor display differs.
已报告滚动 EPS 能否与调整后预期 EPS 比较？算术上可以，但会混合政策。应重建“报告对报告”或“调整对调整”的 EPS，并列出剔除项目及税务影响。混合结果只适合用于解释数据页面为何不同。
Update when price, forecast information, fiscal mapping or expected diluted shares materially change. Keep the old version and date the new one. Revision history is needed to judge whether lower forward P/E came from price movement, forecast change or period rollover.
预期 P/E 应多久更新一次？当价格、预测信息、财政期间映射或预期稀释股数发生重大变化时更新。保留旧版并为新版注明日期。只有修订历史才能判断倍数下降来自价格、预测变化还是期间滚动。
It could reflect a lower price, a higher EPS estimate, a roll into a later fiscal period or a definition change. Freeze the price first, compare estimate versions and rebuild the denominator before assigning an operating interpretation.
业绩发布后预期 P/E 下降意味着什么？可能是价格下降、EPS 估计上升、滚动至更晚财政期间，或定义改变。先固定价格，再比较估计版本并重建分母，然后才能作经营解释。
Sources and limits资料来源与边界
SEC EDGAR provides filings, SEC C&DIs address non-GAAP measures, IAS 33 covers EPS, and CFA Institute readings discuss multiples. None validates FrostArc, its estimates, consensus or price.
FrostArc and every date, amount, outlook and result are fictional. Actual comparison needs current documents, provider definitions and the applicable framework. This denominator-control method is not a forecast, assurance conclusion, price target or individualized recommendation.
SEC EDGAR 提供发行人披露，SEC C&DIs 讨论非 GAAP 指标呈现，IAS 33 涵盖基本与稀释 EPS，CFA Institute 资料讨论估值倍数。这些来源不会提供或验证 FrostArc、其估计、任何当前一致预期或任何报价。
FrostArc 以及本页全部日期、金额、展望、估计与兑现结果均为虚构。现实比较需要当前原始文件、数据提供方定义与适用报告框架。本页是用于控制分母的教育方法，不是预测、鉴证结论、目标价或个性化投资建议。
- U.S. Securities and Exchange Commission — EDGAR Search and Access
- U.S. Securities and Exchange Commission — Non-GAAP Financial Measures C&DIs
- IFRS Foundation — IAS 33 Earnings per Share
- CFA Institute — Market-Based Valuation: Price and Enterprise Value Multiples
- CFA Institute — Equity Valuation: Concepts and Basic Tools