Voltaire: “The Holy Roman Empire was neither holy, nor Roman, nor an empire.”

1) From EBIT to Free Cash Flow (FCF)

Why this bridge matters

Valuation rests on cash the business can distribute after it funds operations and growth. Accounting profit ≠ distributable cash.

Clean bridge (unlevered → firm FCF)

1. EBIT
2. – Cash taxes on EBIT → EBIT(1–T)
3. + Depreciation & amortization (non-cash)
4. – Capital expenditures (maintenance + growth)
5. – ΔNWC (increase in operating working capital: AR, inventory, AP)

= FCF to the firm (FCFF)

Notes you flagged correctly

• Treat “invested” working capital only (exclude excess cash and financing items).
• Repairs that simply keep assets productive are maintenance (economically similar to depreciation) and therefore a cash outflow in Capex.

4) Excess cash: don’t value it like operations

Enterprise Value (EV) values operating assets only. Equity value = EV + non-operating assets (excess cash, investments) – debt and other claims.

Workflow

1. DCF the operations to get EV.
2. Add excess cash & non-operating investments.
3. Subtract net debt and other non-equity claims.
4. Divide by diluted shares → equity value per share.

What is “excess”?

Cash beyond prudent operating needs + regulatory/strategic buffers. Back it with a simple policy (e.g., 2–3 months OpEx + near-term commitments).

5) Dividends vs. Buybacks (your implied question)

Are they “functionally the same”?

• In frictionless theory, yes: both return cash to owners.
• In practice, not identical due to:
• Price paid for buybacks (create value if below intrinsic; destroy if above).
• Tax treatment for some holders.
• Signaling/commitment: dividends are “sticky”; buybacks are flexible.
• Per-share metrics: buybacks shrink share count → boost EPS mechanically.

Simple buyback test

If intrinsic value per share ViV_iVi > repurchase price PPP, then buying back $1 of stock increases value per remaining share. If P>ViP > V_iP>Vi, you’re transferring value from remaining to selling shareholders.

6) Capital allocation hierarchy (how to use the cash)

1. High-IRR reinvestment in the core (organic growth).
2. Maintain fortress balance sheet (liquidity, terming out debt).
3. M&A only if risk-adjusted returns > organic alternatives.
4. Debt pay-down when expected after-tax return on cash < cost of debt and you value flexibility less.
5. Buybacks when price < intrinsic and business is durable.
6. Dividends when persistent surplus exists and credibility matters.

Decision memo you’d want on each dollar

• Expected IRR (and its drivers), risk, reversibility, and effect on optionality.

7) Debt, cash, and the “option value” of cash

• Cash is a real option: it preserves the right (not the obligation) to invest when opportunities appear or to withstand shocks.
• Debt lowers WACC up to a point (tax shield), but raises financial risk.
• Value the tax shield conservatively and stress test covenant headroom, refinancing windows, and rating triggers.

8) Worked micro-examples

A) Perpetuity value sensitivity

Assume FCF0=$100BFCF_0 = \$100BFCF0=$100B.

You see why your “~3% spread → ~$3T” napkin math works.

B) Rates and “other income”

• Cash & securities = $150B earning 2% → $3.0B pre-tax.
• Rates rise to 4% → $6.0B pre-tax.
• Δ = $3.0B pre-tax; at 20% tax → $2.4B after-tax.
• With 15.5B diluted shares, EPS tailwind ≈ $0.15.
• Helpful, but dwarfed by the perpetuity sensitivity to WACC−gWACC-gWACC−g.

C) Buyback math (value creation test)

• Intrinsic value per share Vi=$180V_i = \$180Vi=$180.
• Market price P=$150P = \$150P=$150.
• Company spends $15B → retires 100M shares.
• Value created for continuing holders ≈ (Vi−P)×shares bought(V_i - P) \times \text{shares bought}(Vi−P)×shares bought = $3B (ignoring second-order effects).
• If P=$210P = \$210P=$210 (> ViV_iVi), that destroys ≈ $3B of value.

9) Practical modeling habits (answers to the “how should I do this?” you implied)

• Separate operating DCF (EV) from non-operating assets/liabilities.
• Use nominal cash flows with a nominal WACC; be consistent.
• Build a WACC–g data table; chart value vs. the spread.
• Make maintenance vs. growth Capex explicit; tie growth Capex to incremental ROIC.
• Tie ΔNWC to drivers (DSO, DPO, DIO) so growth that requires inventory/receivables is funded in the model.
• Reconcile CFO – Capex = FCF to the EBIT(1–T) bridge each forecast year (two ways to the same number).
• Check ROIC vs. WACC: long-run ggg should be feasible given reinvestment rate =gIncremental ROIC= \frac{g}{\text{Incremental ROIC}}=Incremental ROICg.

10) Quick FAQ based on your notes

Q: Why is the Gordon piece “dangerous” near small spreads?

Because a 50 bps change in WACC or g can swing billions when WACC−gWACC-gWACC−g ≈ 3%. Always show ranges.

Q: Should I include interest income from “excess cash” in the operating DCF?

No. Keep operations clean; add excess cash (valued at face or mark-to-market) after the EV DCF.

Q: When should I favor dividends over buybacks?

• If shares look fully valued and you still have surplus cash every year, a stable dividend transfers cash without price-risk.
• If shares are undervalued and cash flows are durable, buybacks usually dominate.

Q: How do I decide what’s “excess” cash?

Quantify: operating buffer (e.g., 2–3 months OpEx + working capital needs) + committed outlays (debt maturities, Capex) + a conservative shock buffer. Everything beyond that is excess for capital allocation.

11) A tiny checklist you can reuse

2) “Supplier credit as a form of capital”

When suppliers let you pay later (Accounts Payable), they’re financing part of your operations. During the delay:

• You hold the goods/cash.
• They effectively fund that working capital.

In cash-flow terms, rising A/P is a source of cash (negative working-capital investment). Businesses with strong bargaining power (or subscription/prepayment models) can finance growth with customer float and supplier credit instead of external debt/equity. That’s why ΔNWC is in the FCF bridge.

3) What is a perpetuity—and what it means to treat Apple like one

A perpetuity is a stream of cash flows that continues forever. A growing perpetuity grows at constant ggg.

In DCFs we:

1. Forecast explicit years (5–10) while the business “converges” to steady state.
2. Replace all later years with a single terminal value using the growing perpetuity formula.

Saying “treat Apple’s future cash flows as a perpetuity” means: after your explicit period, you assume cash flows grow forever at a stable, conservative rate and discount them at WACC.

4) Relationship between WACC and ggg

• WACC === r is the opportunity cost of capital for the whole firm (blend of debt/equity, after tax).
• ggg is the long-run nominal growth rate of free cash flow in steady state.

Valuation rides on the spread r−gr-gr−g.

• If rates or risk premia push rrr up, value falls (denominator widens).
• If durable reinvestment at good returns pushes ggg up, value rises (denominator tightens).
• Small changes near a small spread (e.g., 3%) swing value massively—handle with sensitivity tables.

7) Why Apple says it holds so much cash—and how to model “excess cash” cleanly

Why hold big cash/securities? (common rationales)

• Global tax/withholding frictions (historically more binding; still relevant).
• Supply-chain & capex optionality (supplier prepayments, tooling, strategic buys).
• Shareholder return program smoothing (dividends/buybacks across cycles).
• M&A dry powder.
• Ratings/credit-market access (large buffers support cheap debt windows).

How to treat in a model

1. Operate the DCF on core operations only → Enterprise Value.
2. Add non-operating assets (cash & marketable securities marked at fair value).
3. Subtract net debt & other claims → Equity value.
4. If you assume Apple pares “excess” cash:
• Choose a policy buffer (e.g., 2–3 months OpEx + near-term commitments).
• Liquidate surplus (no operating impact), then use of cash:
• Debt paydown (reduce interest expense; adjust WACC if rating changes).
• Buybacks (reduce share count; model average repurchase price and timing).
• Special dividend (no share count change; simple transfer to owners).

Concepts to know

• Non-operating vs. operating cash
• Capital returns (dividends vs. buybacks; price vs. intrinsic value)
• Tax effects (gains on sale of securities are already in accounting flows; your valuation layers add market value of those securities, not re-forecast their coupons)
• Optionality value (buffers can be worth more than their yield)

Depreciation vs. Capex (how they relate)

• Capex is cash out today for long-lived assets.
• Depreciation is the non-cash expense that spreads that cost over useful life in accounting.

In economic steady state, maintenance Capex ≈ Depreciation (keeping the asset base level).

Growth requires Capex > Depreciation. In your FCF bridge, you add back Depreciation (non-cash) and subtract Capex (cash).

Apple’s value drivers & scenario sensitivities

Now: iPhone is the anchor (high share of industry profit despite minority unit share). Strategy = premium positioning, high ASP, ecosystem lock-in.

Key drivers to model:

1. iPhone: Units, ASP, and gross margin. Sensitivities to premiumization vs. share erosion; China mix.
2. Services: Installed base × ARPU × take-rate; high margin; regulatory risk (app store fees, anti-steering rules).
3. Wearables/Other: Attach rates to the base; margin mix.
4. Geography: Currency, China exposure, Europe regulation.
5. Cost structure: Silicon integration; supply chain concentration; logistics.
6. Capital returns: Buyback cadence and pricing; sustainable dividend growth.

Challenges to stress:

• Regulatory (app store economics; antitrust).
• Platform transition risk (AI interface shift; if value creation moves to LLM-native platforms).
• Supply-chain geopolitics (China/Taiwan risk).
• Replacement cycles lengthening.

Build a “bear/base/bull” with explicit levers: (i) iPhone units ±10–15%, ASP ±5–10%, (ii) Services ARPU growth ±300–500 bps, (iii) WACC ±100 bps, ggg ±50 bps. Watch the terminal value dominate when r−gr-gr−g gets tight.

Your POS thesis (phone → computer → Personal Operating System)

There’s real logic here:

• Aggregation: If the handset + watch + earbuds + home devices + car integration + cloud + identity become the persistent personal OS, the platform’s switching costs and network effects deepen (developers target your APIs; users stay for identity, data, payments, messages).
• Economic levers: Higher lifetime value per user (hardware margin + services ARPU + payments/identity rails).
• Moats: Trusted identity, private on-device AI, frictionless payments, superior device-to-device UX.

Risks: regulation (self-preferencing, app store economics), open AI agents that intermediate OS control, and cross-platform standards that blunt lock-in. As a long-term value narrative, POS can justify a sturdier ggg—but only if you can trace drivers to cash: installed base growth, ARPU expansion, retention, and take-rates

Economist

• Trade-offs everywhere — Thomas Sowell: “There are no solutions; only trade-offs.”

Apply: Ask “what do we give up?” anytime someone says “free.”

• Unseen effects — Frédéric Bastiat, What Is Seen and What Is Not Seen.

Apply: List second-order effects before deciding.

• Incentives rule — (Often paraphrased from Steven Landsburg: “People respond to incentives; the rest is commentary.”)

Apply: Change incentives → expect behavior to change.

• Transaction costs — Ronald Coase.

Apply: Ask “does this arrangement minimize the cost of coordinating/exchanging?”

• Knowledge problem — F. A. Hayek: markets aggregate dispersed information.

Apply: Prefer systems that harness local knowledge over top-down guesses.

• Comparative advantage — David Ricardo.

Apply: Specialize in the thing you do relatively best; outsource the rest.

• The Nirvana fallacy — Harold Demsetz: compare to realistic alternatives, not utopia.

Apply: “Compared to what?” should be on your whiteboard.

• Nudges & frictions — Richard Thaler: small design choices change outcomes.

Apply: Make the good choice easy, the bad choice a bit harder.

• Institutions matter — Douglass North: rules shape incentives over time.

Apply: When performance lags, check the rules of the game, not just the players.

• Measurement humility — Angus Deaton: data are noisy; averages hide distributions.

Apply: Look at medians, tails, and definitions before acting on a metric.