model-update

​

What is a Financial Model Update?

​

Why It Matters

• Price target integrity: Your investment recommendation (Buy/Hold/Sell) rests on your price target, which rests on your model. A stale model means a stale recommendation
• Client credibility: When a PM calls asking “what is your updated EPS estimate?”, having a current answer vs. “I have not updated yet” makes or breaks the relationship
• Estimate revision signal: Changes in analyst estimates are one of the strongest predictors of stock price movement. Upward revisions drive stocks higher; downward revisions drive them lower
• Learning: Understanding why your estimates were wrong improves forecasting accuracy over time

​

Key Concepts

​

Worked Example: Updating the Starbucks (SBUX) Model After Q1 FY2025

​

Step 1: Identify What Changed

​

Step 2: Analyze Why Actuals Differed from Estimates

Traffic:  -6% (worse than -4% estimate)
Ticket:   +2% (in line with estimate)
Analysis: Traffic decline accelerated despite new product launches.
          Management cited: (1) value perception gap vs. competitors,
          (2) mobile order wait times deterring repeat visits,
          (3) reduced afternoon daypart traffic.
Question: Is this cyclical (temporary) or structural (permanent)?
Assessment: Partially structural -- value perception requires pricing
            action, which management committed to in Q2. Traffic likely
            stabilizes but does not fully recover in FY2025.

China comps:  +6% (vs. +2% estimate) -- recovery stronger than expected
EMEA comps:   +4% (vs. +3% estimate) -- steady
Japan comps:  +5% (vs. +4% estimate) -- tourism boost
Analysis: China recovery is the key driver. Store count growth in China
          accelerated. Management noted "acceleration in digital
          engagement" and partnership momentum with local platforms.
Question: Is China recovery sustainable?
Assessment: Likely yes for 2-3 quarters as comps cycle easy comparisons.
            Sustainable growth rate in China is probably +3-5% long-term.

Gross margin: +60bps vs. estimate
  - Lower coffee prices: +30bps (commodity tailwind)
  - Product mix shift (cold beverages higher margin): +20bps
  - Supply chain efficiencies: +10bps

Operating margin: +70bps vs. estimate
  - Gross margin beat: +60bps (flows through)
  - SG&A discipline: +10bps (lower marketing spend)

Analysis: Commodity tailwind may be temporary (coffee futures are volatile).
          Mix shift and supply chain efficiencies are sustainable.
Assumption change: Raise FY2025 gross margin by 30bps (50% of beat
          flows through, as commodity benefit may not persist all year).

​

Step 3: Revise Forward Estimates (Bridge Analysis)

ASSUMPTION LOG -- SBUX Model Update, January 29, 2025

1. US Comp Sales: Lowered Q2-Q4 estimates by 200-250bps each.
   Reason: Q1 traffic decline of -6% was worse than expected.
   Management guidance lowered to "-2% to flat" for the year.
   I assume gradual improvement from -4% in Q1 to +0.5% in Q4
   as value menu launches and mobile order improvements take effect.

2. International Comp Sales: Raised Q2-Q4 estimates by 100-150bps.
   Reason: China recovery stronger than modeled. Q1 at +6% vs. +2%
   estimate. Raising to reflect easier comparisons and digital
   engagement improvements. Assume +5% in Q2, +4.5% in Q3, +4% in Q4.

3. Gross Margin: Raised by 30bps for the full year.
   Reason: Q1 beat of 60bps. Applying 50% flow-through because
   coffee commodity tailwind may not persist. Mix shift toward cold
   beverages is sustainable.

4. Share Count: Lowered by 10M shares.
   Reason: Company accelerated buyback pace in Q1. Assuming this
   pace continues given strong FCF and management commitment.

5. Store Count: Lowered net openings from 1,600 to 1,500.
   Reason: Management lowered guidance, citing "optimization of
   real estate strategy" and focus on profitability over growth.

​

Step 4: Recalculate Valuation

Blended price target:
  DCF value:       $98  (weight: 33%)  = $32.34
  NTM P/E value:   $98  (weight: 34%)  = $33.32
  EV/EBITDA value:  $95  (weight: 33%)  = $31.35
  Blended:         $97  (rounded)

Current stock price: $88
Upside to price target: $97 / $88 = 10.2%
Prior upside: $92 / $88 = 4.5%

Multiple check: $97 PT / $3.78 NTM EPS = 25.7x P/E
Historical range: 22-30x. Current: within range. Reasonable.

​

Step 5: Rating Decision and Summary

STARBUCKS (SBUX) -- MODEL UPDATE SUMMARY
Date: January 29, 2025

HEADLINE: Mixed quarter -- US traffic worse, margins better.
EPS raised on margin beat; revenue lowered on US comp miss.
Net: raising PT to $97 from $92. MAINTAIN HOLD.

ESTIMATE CHANGES:
| | Old | New | Change |
|--|-----|-----|--------|
| FY25 Revenue | $37.5B | $37.2B | -0.8% |
| FY25 EPS | $3.60 | $3.78 | +5.0% |
| FY26 EPS | $4.00 | $4.15 | +3.8% |
| Price Target | $92 | $97 | +5.4% |
| Rating | Hold | Hold | No change |

WHY NOT UPGRADE TO BUY?
Despite the EPS raise, we maintain Hold because:
1. US traffic decline is accelerating (-6% in Q1), and the
   turnaround timeline is uncertain
2. The margin beat was partly commodity-driven (temporary)
3. At $88, the stock offers 10% upside to PT -- below our
   15% threshold for a Buy rating
4. Management is executing well on margins but the revenue
   growth story remains unproven under the new CEO

WHAT WOULD MAKE US UPGRADE?
- US comp sales turn positive for 2 consecutive quarters
- Evidence that menu innovation is driving incremental traffic
- Stock declines to $80 or below (15%+ upside to $97 PT)

NEXT CATALYST: Q2 FY2025 earnings (late April). Key focus:
US comp trajectory and impact of value menu launch in February.

​

Full Skill Workflow (From SKILL.md)

​

Phase 1: Identify What Changed

​

Phase 2: Plug New Data

• Update each segment’s revenue and margin
• Note any segment mix shifts
• Identify which segments drove the beat/miss

• Cash and debt balances (for net debt and EV calculation)
• Share count (buybacks, dilution from stock comp)
• Capex actual vs. estimate
• Working capital changes (inventory, receivables, payables)
• Any new debt issuance or repayment

• Restructuring charges
• Asset impairments
• Legal settlements
• Acquisition-related costs
• One-time tax benefits or charges
• For each: quantify, determine if it should flow through to forward estimates

​

Phase 3: Revise Forward Estimates

For each changed assumption, record:
- What: Revenue growth rate changed from X% to Y%
- Why: Q[X] beat/miss + management commentary
- Duration: Temporary (1-2 quarters) or permanent
- Confidence: High/Medium/Low

​

Phase 4: Recalculate Valuation

Price Target / NTM EPS = Implied P/E multiple
  - Is this within the historical range?
  - Is this consistent with the company's growth and quality?

Price Target / Current Price = Upside/Downside
  - Does the upside/downside support the rating?
  - Buy: typically requires 15%+ upside to PT
  - Hold: typically -10% to +15%
  - Sell: typically >10% downside to PT

​

Phase 5: Summarize and Recommend

1. What changed: 2-3 sentences on the key developments
2. Estimate changes: Table showing old vs. new for key metrics
3. Thesis impact: Is this a thesis-changing event or noise?
4. Rating/PT: Maintain or change? State new price target with methodology
5. Upside/downside: Calculate and state explicitly
6. Next catalyst: What is the next event that will test the thesis?

​

Phase 6: Output

• Updated Excel model (if the user provides the existing model file)
• Estimate change summary (markdown or Word document)
• Updated price target derivation (showing old vs. new)

​

Common Mistakes (and How to Avoid Them)

​

Daily Workflow Scenarios

​

Scenario 1: Post-Earnings Model Update (Same Day)

1. 4:15-4:45 PM: Download press release. Pull key actuals into the model. Compare to your estimates and consensus. Quantify beat/miss on each line.
2. 4:45-5:30 PM: Listen to the earnings call (or read the transcript as it is published). Note guidance changes and key assumption shifts.
3. 5:30-7:00 PM: Update the model:
   • Plug Q3 actuals
   • Revise Q4 and FY2025 estimates based on Q3 trends and updated guidance
   • Update balance sheet (cash, debt, share count)
   • Recalculate DCF and price target
4. 7:00-8:00 PM: Write the estimate change summary. Include: old vs. new table, key assumption changes, rating/PT action.
5. 8:00-9:00 PM: Draft the earnings update report (pages 1-3, most critical sections first).
6. Next morning 7:00 AM: Publish the updated model and report.

​

Scenario 2: Model Update After Guidance Change (No Earnings)

1. Same day (1 hour): Read the guidance update. Why are they raising? Is it demand-driven, pricing-driven, or acquisition-related?
2. Same day (1 hour): Update the model: raise revenue estimates for the current quarter and full year. Flow through to EBITDA and EPS at existing margin assumptions.
3. Same day (30 minutes): Recalculate price target. If the upside changes materially, consider a rating change.
4. Same day (30 minutes): Publish a brief note: “SBUX guides Q2 revenue above consensus. Raising FY2025 EPS to $X from$X. Maintain Buy, PT $XX from$XX.”

​

Scenario 3: Multi-Model Update During Earnings Season

1. Pre-plan: Prioritize by market cap and client interest. The largest name gets the most thorough update.
2. For each company:
   • Day of earnings: Plug actuals, note initial beat/miss (30 minutes per company)
   • Day after: Full model update with revised estimates and price target (2-3 hours for priority names, 1 hour for less important names)
   • Day after: Publish note
3. End of week: Review all updates for consistency. Are there sector-level themes (pricing power, demand weakness) that should be reflected across multiple models?

​

Scenario 4: Macro-Driven Multi-Model Update

1. Day 1 (2 hours): Update the DCF WACC assumption across all models. Reduce by 25-50bps depending on how much of the rate cut you expect to persist.
2. Day 1 (1 hour): For highly leveraged companies, update interest expense projections. Calculate the EPS impact of lower rates.
3. Day 2 (2 hours): Recalculate all price targets. Publish a sector note: “Rate cut impact on [sector] coverage: Raising PTs across the board. Biggest beneficiaries: [names with most leverage or most DCF sensitivity].”

​

Practice Exercise

​

How to Add to Your Local Context

# Install the plugin
claude plugin install equity-research@financial-services-plugins

open ~/.claude/skills/equity-research/model-update.md

{
  "mcpServers": {
    "file-access": {
      "command": "file-server-mcp",
      "args": ["--root", "/path/to/models"]
    }
  }
}

{
  "mcpServers": {
    "consensus": {
      "command": "factset-estimates-mcp",
      "args": ["--api-key", "YOUR_KEY"]
    }
  }
}

​

Best Practices

• GAAP vs. Adjusted consistency: Note whether your estimates are GAAP or adjusted, and handle non-recurring items consistently across all periods. Mixing the two is a common error
• Track your revision history: Maintain a log of every estimate change with the date and reason. This shows your analytical progression and helps identify systematic biases
• Separate signal from noise: If the quarter was noisy (one-time charges, accounting changes, FX impacts), separate these from the underlying business trends in your estimate changes
• Check consensus after updating: How do your revised estimates compare to the Street? Being materially above or below consensus is worth noting and explaining
• Share count matters: Dilution from stock-based compensation, convertible debt, or buybacks can materially affect EPS even when net income is unchanged
• Model the balance sheet: Cash generation, debt paydowns, and capital returns matter for equity value. Do not just model the income statement
• Flag key sensitivities: After updating, note which assumptions your price target is most sensitive to. This helps PMs understand the risk/reward
• Speed matters during earnings season: A model update published 3 days after earnings is worth less than one published same-day. Prioritize speed for important names

​

Dependencies

• Excel or equivalent spreadsheet tool for the financial model
• Access to earnings press release and transcript

• Bloomberg/FactSet for consensus comparison
• SEC EDGAR for 10-Q filing details
• Python for automated data extraction and model population