Model Update

Model Update

Workflow

Step 1: Identify What Changed

Determine the update trigger:

• Earnings release: New quarterly actuals to plug in
• Guidance change: Company updated forward outlook
• Estimate revision: Analyst changing assumptions based on new data
• Macro update: Interest rates, FX, commodity prices changed
• Event-driven: M&A, restructuring, new product, management change

Step 2: Plug New Data

After Earnings

Update the model with reported actuals:

Segment Detail (if applicable):

• Update each segment's revenue and margin
• Note any segment mix shifts

Balance Sheet / Cash Flow Updates:

• Cash and debt balances
• Share count (buybacks, dilution)
• Capex actual vs. estimate
• Working capital changes

Step 3: Revise Forward Estimates

Based on the new data, adjust forward estimates:

Key Assumption Changes:

• What assumptions are you changing and why?
• Revenue growth rate: old → new (reason)
• Margin assumption: old → new (reason)
• Any new items (restructuring charges, one-time gains, etc.)

Step 4: Valuation Impact

Recalculate valuation with updated estimates:

Step 5: Summary & Action

Estimate Change Summary:

• One paragraph: what changed, why, and what it means for the stock
• Is this a thesis-changing event or noise?

Rating / Price Target (draft opinion, pending analyst sign-off — never an issued view):

• Maintain or change rating? — stance is [推断]
• New price target (if changed) with methodology — target arithmetic is [测算]
• Upside/downside to current price

Step 6: Output

• Updated Excel model (if user provides the existing model) — built through xlsx-author as a Class A model workbook: every revised blue input carries its Source: cell comment, and the Inputs and Checks tabs travel with it
• Estimate change summary (Markdown in-session by default for short-form; state the choice). If the user wants it as a document, report-render builds the PDF — never hand-roll one, because a hand-rolled PDF does not emit [n] as link annotations 用户要 Word 时同样走 report-render（DocxReport，与 Report 同一套调用）——要出 Word 就先载入 report-render 技能，再动手建；手搓 python-docx / docx-js 会丢掉标签配色、[n] 跳转与封面页，机制与实测记录在那个技能里，不在这里重述。
• Updated price target derivation

Step 7: Cite the Inputs in the Workbook

A model's citation vehicle is cell comments, not a Sources section. Every hardcoded input cell carries a comment in exactly this form:

Source: <System or Document>, <Date>, <Reference>, <URL>

e.g. Source: FY2025 annual report, 2026-03-28, consolidated income statement "Revenue", https://… or Source: 同花顺 iFinD 行情数据, retrieved 2026-07-25, [证券代码] 最新成交价 (headline financials cross-checked: Source: 万得 上市公司财务数据, retrieved …, 营收 reconciled).

• Every hardcoded number gets one — a reported actual you plugged in, a guidance midpoint, a consensus figure, an FX or rate assumption. No exceptions, including the ones you are sure are obvious.
• Name the provider for any consensus input; an unattributed consensus number does not exist.
• A calculated cell carries a formula instead, and no source comment. The formula is its own provenance. A source comment on a calculated cell means a hardcode is hiding inside it — go find it.
• Colour reinforces the same distinction: blue #0000FF for a hardcoded input, black for a formula, green #008000 for a link to another sheet, purple #800080 for a same-sheet reference with no calculation.
• Judgement assumptions (a growth rate you chose, a margin path) are [测算]: the comment states the basis and the cell is listed in the assumptions block, so the reader can see it was decided rather than retrieved.

In the accompanying estimate-change summary, tag each figure once: [披露] for the company's reported actuals, [测算] for your revised estimates and derived valuation, [预期] for the Street comparison with the provider named, [推断] for an attribution you are asserting without a record, [媒体] for anything resting on an uncorroborated report. One tag style per document.

Step 8: Coverage and Limitations

Close the summary with a coverage block. It is written even when every input was retrieved, because a reader cannot otherwise tell an input we checked and could not find from one we never looked for:

## 覆盖范围与局限
Retrieved: <YYYY-MM-DD HH:MM TZ>  ·  Scope: <Company> (<ticker>) model, <what triggered the update>

| 检查项 | 结论 | 源 | 检索于 |
|---|---|---|---|
| Reported quarterly actuals | 有记录 | <filing / release> | <date> |
| Segment detail | 检索范围内未发现 | <filing> (not broken out) | <date> |
| Forward guidance | 有记录 | <transcript / release> | <date> |
| Street consensus, post-print | 源不可用 | <provider> (out of quota) | <date> |

本次未能覆盖: <what failed, and which model inputs are therefore still stale>
Inputs left at prior values: <cells not refreshed, and why>

The three states are 有记录 / 检索范围内未发现 / 源不可用. "检索范围内未发现" is a statement about our retrieval, not about the company — never write it as "the company does not disclose it" unless a disclosure explicitly says so.

Important Notes

• Always reconcile your estimates to the company's reported figures before projecting forward
• Note any non-recurring items and whether your estimates are GAAP or adjusted
• Track your estimate revision history — it shows your analytical progression
• If the quarter was noisy, separate signal from noise in your estimate changes
• Check consensus after updating — how do your revised estimates compare to the Street?
• Share count matters — dilution from stock comp, converts, or buybacks can materially affect EPS
• A stale cell is a silent error: if a source was unavailable and an input kept its prior value, that belongs in the coverage block, not in your head