# ratio-analysis — Methodology Detail

Extracted from SKILL.md for progressive disclosure (US5).

## Retrieval Strategy

Follow the retrieval strategy decision tree in `contracts/retrieval.md`. This skill uses:
- Branch (a) for structured financial metrics via `search_xbrl_facts` with `list_xbrl_concepts` pre-condition for unfamiliar concepts.
- Branch (d) for simple lookups via `get_realtime_quote` / `search_companies` / `search_earnings_calendar`.

## Protocol

1. **Pre-retrieval**: call `get_company_fiscal_calendar/{ticker}` to resolve fiscal period format, then `get_ticker_coverage/{ticker}` .
2. **XBRL retrieval**: `search_xbrl_facts(ticker, concept=["Revenues","GrossProfit","OperatingIncomeLoss","NetIncomeLoss","Assets","Liabilities","Equity","OperatingCashFlow","InventoryNet","ReceivablesNet","CurrentAssets","CurrentLiabilities","InterestExpense","LongTermDebt"], fiscal_year=[latest, latest-1, latest-2, latest-3])` — batch all concepts × 4 years.
3. **Price data**: `get_realtime_quote(ticker)` for current stock price, market cap, PE (TTM).
4. **Peer identification**: `get_peer_comparison(ticker, metric)` (or `search_companies(search=<industry>)` — note there is no `sector` param) to identify peer tickers for cross-company comparison.
5. **Compute ratios** into 6 categories per the Ratio Definitions below.
6. **Cross-company comparison**: for each peer, fetch key ratios and present comparison table with mean/median/high/low.
7. **Output**: per file convention with YAML frontmatter .

## Ratio Definitions

#### Profitability
| Ratio | Formula | Interpretation |
|-------|---------|---------------|
| ROE | Net Income / Avg Total Equity | >15% = strong; measures return to shareholders |
| ROA | Net Income / Avg Total Assets | >5% = efficient; asset utilization |
| ROIC | (EBIT × (1 - Tax Rate) / (Total Debt + Equity - Cash) | > WACC = value-creating |
| Gross Margin | Gross Profit / Revenue | Industry-dependent; higher = pricing power |
| Operating Margin | Operating Income / Revenue | >15% = healthy operations |
| Net Margin | Net Income / Revenue | >10% = strong bottom-line efficiency |

#### Liquidity
| Ratio | Formula | Interpretation |
|-------|---------|---------------|
| Current Ratio | Current Assets / Current Liabilities | >1.5 = healthy; <1.0 = liquidity risk |
| Quick Ratio | (Cash + Receivables) / Current Liabilities | >1.0 = strong; acid test |
| Cash Ratio | Cash / Current Liabilities | Most conservative; >0.5 = adequate |
| Operating CF Ratio | Operating Cash Flow / Current Liabilities | >1.0 = can cover obligations from ops |

#### Leverage
| Ratio | Formula | Interpretation |
|-------|---------|---------------|
| Debt-to-Equity | Total Debt / Total Equity | <2.0 = conservative; >4.0 = aggressive |
| Interest Coverage | EBIT / Interest Expense | >3x = safe; <1.5x = distress risk |
| Debt-to-EBITDA | Total Debt / EBITDA | <3x = manageable; >5x = highly leveraged |

#### Efficiency
| Ratio | Formula | Interpretation |
|-------|---------|---------------|
| Asset Turnover | Revenue / Avg Total Assets | Higher = more efficient |
| Inventory Turnover | COGS / Avg Inventory | Higher = faster sales; watch for stockouts |
| Days Sales Outstanding | (Receivables / Revenue) × 365 | Lower = faster collections |
| Days Inventory Outstanding | (Inventory / COGS) × 365 | Lower = leaner operations |

#### Valuation (requires current price from `get_realtime_quote`)
| Ratio | Formula | Interpretation |
|-------|---------|---------------|
| P/E (LTM) | Price / LTM EPS | Lower = cheaper; sector-dependent |
| P/E (NTM) | Price / NTM Consensus EPS | Forward-looking; from `search_earnings_calendar` |
| P/B | Price / Book Value Per Share | <1.0 = below book; financials focus |
| EV/EBITDA | Enterprise Value / EBITDA | Capital-structure neutral |
| P/S | Market Cap / Revenue | Growth check; <2.0 = reasonable |
| PEG | P/E / Earnings Growth Rate | <1.0 = undervalued per Peter Lynch |

#### Growth
| Ratio | Formula | Interpretation |
|-------|---------|---------------|
| Revenue CAGR (3yr) | (Revenue_t / Revenue_t-3)^(1/3) - 1 | Trend; >10% = strong growth |
| EPS CAGR (3yr) | (EPS_t / EPS_t-3)^(1/3) - 1 | >10% = strong earnings growth |
| Revenue CAGR (5yr) | (Revenue_t / Revenue_t-5)^(1/5) - 1 | Longer trend |
| EPS CAGR (5yr) | (EPS_t / EPS_t-5)^(1/5) - 1 | Longer earnings trend |

# ratio-analysis — Analyst Mode Definitions

Derived from the skill's own methodology structure (scripts/mode_backfill.py, spec 046 M1).

### Mode: preflight

**Objective**: Preflight analysis per the skill's methodology (see SKILL.md sections).

### Mode: triggers

**Objective**: Triggers analysis per the skill's methodology (see SKILL.md sections).

### Mode: defaults

**Objective**: Defaults analysis per the skill's methodology (see SKILL.md sections).

### Mode: methodology

**Objective**: Methodology analysis per the skill's methodology (see SKILL.md sections).

### Mode: retrieval-scope

**Objective**: Retrieval Scope analysis per the skill's methodology (see SKILL.md sections).

# ratio-analysis — Output Structure (full template)

Extracted from SKILL.md for progressive disclosure (US5). The skill body keeps a compact summary under `## Output Structure`.

1. **Executive Summary** — top 3-5 ratios with interpretation, overall financial health assessment
2. **Profitability Analysis** — ROE, ROA, ROIC, margins table with trailing 4-quarter trend and industry comparison
3. **Liquidity Analysis** — current, quick, cash, operating CF ratios with short-term risk assessment
4. **Leverage Analysis** — D/E, interest coverage, debt/EBITDA with solvency assessment
5. **Efficiency Analysis** — asset turnover, inventory turnover, DSO, DIO with operational assessment
6. **Valuation Snapshot** — P/E (LTM+NTM), P/B, EV/EBITDA, P/S, PEG with sector peer comparison
7. **Growth Trends** — revenue/EPS CAGR (3yr + 5yr) with trend commentary
8. **Cross-Company Comparison** — peer ratio comparison table with mean/median/high/low (optional: --peers flag)
9. **Coverage Gaps & Citations** — data not retrievable + citation index in `{ticker} {citation_id} page<N>` format

**Citations & memory**: follow `contracts/citation-and-memory.md` — ≥1 citation per 200 words; every material fact, table row, and metric is immediately followed by its inline clickable `https://agentii.ai/v/{ticker}/{citation_id}/{N}` link; a bottom **Citations** section provides a non-duplicative roll-up index; the closing TUI reply includes a compact **Key Citations** list (headline 5–10 facts) of clickable `/v/` URLs; and append the run to `agentii.md` per `contracts/agentii-md-schema.md`.

# Return on Invested Capital — Institutional Methodology

Methodology synthesized from institutional investment research; all text is an original paraphrase.

---

## The Core Principle: The One Dollar Test

A company creates value when the present value of cash flows from its investments exceeds the cost of those investments. One dollar invested must become worth more than one dollar in the market. This principle — not earnings growth, not revenue scale — is the fundamental test of corporate value creation.

Return on Invested Capital (ROIC) is the metric that operationalizes this test. It measures the efficiency with which a company converts invested capital into operating profit. Combined with the Weighted Average Cost of Capital (WACC), it answers the single most important question in fundamental analysis: is this company creating or destroying value with its investments?

---

## Four ROIC Variants

ROIC is not a single number. The choice of variant answers different analytical questions. Using the wrong variant for the question produces errors of 60 percentage points or more.

### Variant 1: Excluding Goodwill + Acquired Intangibles, No Adjustment
**ROIC = NOPAT / (Invested Capital − Goodwill − Acquired Intangibles)**

Answers: What is the underlying organic return on the business, excluding acquisition effects?

Use when: Assessing the standalone economics of the operating business, comparing companies with different acquisition histories.

Typical value for Microsoft (FY2022): **94%**

### Variant 2: Including Goodwill + Acquired Intangibles, No Adjustment
**ROIC = NOPAT / Invested Capital (as reported)**

Answers: What is the total return including acquisition costs, as reported in financial statements?

Use when: GAAP-based comparison, regulatory or compliance context.

Typical value for Microsoft (FY2022): **49%**

### Variant 3: Excluding Goodwill, With Intangibles Capitalization Adjustment
**ROIC = Adjusted NOPAT / (Invested Capital − Goodwill + Capitalized Intangibles)**

Answers: What is the organic return including internally-generated intangible assets?

Use when: Comparing companies that build vs. buy their intangible assets (R&D-heavy vs. acquisition-heavy).

Typical value for Microsoft (FY2022): **48%**

### Variant 4: All Included, With Intangibles Capitalization Adjustment
**ROIC = Adjusted NOPAT / (Invested Capital + Capitalized Intangibles)**

Answers: What is the total economic return, properly accounting for all intangible investments regardless of accounting treatment?

Use when: Holistic assessment of value creation across both organic and acquired intangibles.

Typical value for Microsoft (FY2022): **34%**

### The 60pp Spread

The range from 34% to 94% for the same company demonstrates why ROIC comparisons are meaningless without understanding the construction. Each variant answers a different analytical question. The analyst must:

1. Select the variant appropriate to the question
2. Apply the SAME variant across all companies being compared
3. Document which variant was used and why

---

## NOPAT: Numerator Calculation

### Base Formula
```
NOPAT = EBITA − Cash Taxes
```

Where:
- **EBITA** = EBIT + Amortization of Acquired Intangibles + Embedded Interest on Operating Leases
- **Cash Taxes** = Tax Provision + Change in Deferred Taxes (ΔDTL − ΔDTA) + Tax Shield on Net Interest (Net Interest Expense × Marginal Tax Rate)

### Key Adjustments

**Amortization of Acquired Intangibles**: Added back to EBITA because ongoing maintenance investment in these assets is already expensed on the income statement. Excluding amortization would double-count the maintenance cost. This differs from depreciation, where ongoing maintenance (capex) is capitalized, not expensed.

**Operating Lease Embedded Interest**: For US GAAP companies, operating lease expenses include an embedded interest component reported within operating expenses. This must be added back to arrive at true operating profit. IFRS companies already separate lease depreciation and interest, requiring no adjustment.

**Cash Tax Rate Estimation**: For growing businesses, the cash tax rate is typically approximately 95% of the reported tax rate. Use the effective tax rate from the income statement as the starting point.

**TCJA Section 174 (Effective 2022)**: US companies must amortize R&D costs over 5 years for domestic research (15 years for foreign) rather than expensing immediately. This affects approximately 25% of Russell 3000 companies with positive EBIT. When calculating NOPAT for affected companies, adjust for the timing difference between R&D expense recognition and R&D cash outlay.

---

## Invested Capital: Denominator Calculation

### Operating Approach (Preferred)

```
Invested Capital = Net Working Capital + Net PP&E + Operating Lease ROU Assets
                 + Goodwill + Acquired Intangibles + Other Long-Term Operating Assets
```

Where:
- **Net Working Capital** = Current Assets (excluding excess cash) − Non-Interest-Bearing Current Liabilities
- **Cash** included in NWC = 2% of revenue for steady-state firms, up to 5% for high-growth or unpredictable firms (operational cash needs)
- **Excess Cash** = Total Cash − Operational Cash (2-5% of revenue). Stripped from invested capital and addressed separately in capital allocation analysis.
- **Operating Lease ROU Assets**: Included for US GAAP companies (already on balance sheet per ASC 842)

### Explicit Exclusions

- Excess cash (addressed in capital allocation, not operating analysis)
- Equity investments and non-consolidated subsidiaries
- Finance subsidiaries
- Overfunded pension assets
- Tax loss carryforwards

### Financing Approach (Alternative, for Reconciliation)

```
Invested Capital = Total Debt + Total Leases + Total Equity − Excess Cash
```

Both approaches should produce identical results when the classifications are consistent. The operating approach is preferred because it maps more directly to business operations.

---

## ROIIC: Return on Incremental Invested Capital

### Formula
```
ROIIC = (NOPAT_t − NOPAT_{t-1}) / (Invested Capital_{t-1} − Invested Capital_{t-2})
```

### Application

Use rolling 3-year or 5-year periods to dampen year-to-year noise. The 1-year ROIIC is too volatile for meaningful analysis.

**Critical Warning**: ROIIC overstates economic returns when above the cost of capital, and understates when below. This is a mathematical property of the ratio, not a flaw in the calculation. A company with ROIIC of 25% likely has true marginal returns of approximately 18-22%.

**Interpretation**:
- ROIIC consistently above WACC over 3-5 years: management is allocating capital to value-creating opportunities
- ROIIC consistently below WACC: capital is being deployed into value-destroying projects
- ROIIC negative with positive base ROIC: current year investments have not yet matured; do not conclude value destruction from a single year

---

## DuPont Decomposition

### Formula
```
ROIC = (NOPAT / Sales) × (Sales / Invested Capital)
     = NOPAT Margin × Invested Capital Turnover
```

### Strategy Mapping

The decomposition maps directly to competitive strategy:

- **High Margin × Low Turnover** → Differentiation Strategy: the company earns high margins on each unit sold but requires significant invested capital per unit of revenue. Examples: luxury goods, specialty pharmaceuticals, enterprise software.

- **Low Margin × High Turnover** → Cost Leadership Strategy: the company earns thin margins but generates high revenue relative to invested capital. Examples: discount retail, commodity manufacturing, grocery.

- **High Margin × High Turnover** → Exceptional (and rare): indicates both pricing power AND capital efficiency. Often signals a genuine moat. Examples: network-effect platforms, asset-light subscription businesses.

- **Low Margin × Low Turnover** → Commodity Trap: the business is neither differentiated nor efficient. Likely destroying value.

### Empirical Evidence (1990-2022, Russell 3000)

Companies sustaining 10+ years in the top ROIC quintile achieved this primarily through NOPAT margin dominance (2.7× the universe average) rather than capital turnover (1.5× the universe average). Differentiation is the more durable path to sustained high returns.

---

## Economic Profit

### Formula
```
Economic Profit = (ROIC − WACC) × Invested Capital
```

Economic profit translates the ROIC-WACC spread from percentages to absolute dollars. A company with ROIC of 15%, WACC of 8%, and Invested Capital of $10 billion generates $700 million in economic profit annually.

### Sector-Level Distributions (2018-2022, Russell 3000)

- Top decile: approximately $890 billion/year in combined economic profit
- Bottom decile: approximately $270 billion/year in combined economic destruction
- The distribution is highly skewed: a small number of companies generate the vast majority of economic profit

---

## Intangible Capitalization

### The Problem

Accounting standards treat investments in intangible assets (R&D, brand building, customer acquisition) as period expenses on the income statement. This systematically:
- Understates earnings for intangible-intensive growth companies
- Overstates earnings for companies harvesting past intangible investments
- Understates invested capital, inflating ROIC for intangible-intensive companies
- Destroys the information content of book value

### Industry-Specific Capitalization Rates

Use the framework from academic research (Iqbal, Rajgopal, Srivastava, Zhao 2022) covering 42 Fama-French industries. Key parameters by expense category:

**R&D Capitalization**:
- Investment portion: ranges 7% to 98% by industry (average 76%)
- Asset life: ranges by industry (average 4.4 years)

**Non-R&D SG&A Capitalization**:
- Investment portion: ranges 0% to 80% by industry (average 54%)
- Asset life: ranges by industry (average 3.3 years)

**Important**: These are industry-specific, NOT the simplified Peters-Taylor approach (100% R&D + 30% non-R&D SG&A). Peters-Taylor fails to recognize maintenance R&D and applies uniform rates across industries.

### Perpetual Inventory Method

```
Net Capitalized Intangibles_t = Capitalized R&D Stock_t + Capitalized SG&A Stock_t

Capitalized Stock = [Expense_{t-1} × Capitalization Rate] / (Growth Rate + Amortization Rate)
```

Growth rate assumption: 7% (historical Russell 3000 rate, 1990-2021).

### Impact on ROIC

The adjustment adds to BOTH NOPAT (through net intangible investment) and Invested Capital (through capitalized stock). The net effect depends on growth:

- **High-growth, high-ROIC companies** (Microsoft): NOPAT increases modestly, Invested Capital increases substantially → ROIC DECLINES (49% → 34%)
- **Loss-making, high-growth companies** (Snowflake): NOPAT increases substantially, Invested Capital increases modestly → ROIC IMPROVES (−416% → +3%)
- **Steady-state companies**: NOPAT increase ≈ Amortization of past intangibles → ROIC UNCHANGED

---

## Cash Flow Statement Reclassification (Amazon 2020 Case)

Standard CFS categories misclassify economically meaningful distinctions. Four reclassifications restore relevance:

### 1. Stock-Based Compensation: Operating → Financing
SBC = Selling shares to employees. The operating cash flow adjustment is analytically wrong. Results: Operating CF reduced by SBC amount.

Amazon 2020: $9.2B reclassified. Technology sector median SBC = 25% of Operating CF.

### 2. Operating Leases: Financing → Investing
Buy vs. lease is an investment decision. Both should appear in investing activities.

Amazon 2020: $10.7B reclassified.

### 3. Intangible Investments: Operating → Investing
SG&A split into maintenance (operating expense) and investment (capitalize).

Amazon 2020 methodology: Fulfillment costs 0% investment (already capitalized as PP&E). Technology & Content 75% investment (5-year life). Marketing 50% investment (3-year life). G&A 20% investment (3-year life).

Result: $44.4B intangible investment identified. Amortization of past intangibles = $25.4B. Operating CF INCREASES $35.1B (to $101.2B). EBITDA DOUBLES. EV/EBITDA multiple collapses from 35.0× to 18.2×.

### 4. Marketable Securities: Investing → Cash Equivalents
When marketable securities are functionally equivalent to cash (short-term, highly liquid, non-operating).

Amazon 2020: $22.2B net purchases reclassified.

### Combined Impact on Amazon 2020

| | Operating | Investing | Financing |
|---|---|---|---|
| Reported | $66.1B | −$59.6B | −$1.1B |
| Adjusted | $101.2B | −$114.7B | $18.8B |

**Free cash flow is UNCHANGED.** Properly defined, FCF = Operating CF − Capex (both reported and adjusted). The adjustment changes the PORTRAYAL of business economics, not the underlying cash generation.

---

## Free Cash Flow Invariance Proof

Intangible capitalization does NOT change free cash flow. This is the critical consistency check:

```
FCF = NOPAT − ΔInvested Capital (both reported and adjusted)
```

When intangibles are capitalized:
- NOPAT increases (expenses moved to balance sheet)
- ΔInvested Capital increases by the same amount (net intangible investment)
- The two effects cancel exactly

The adjustment improves UNDERSTANDING (what portion of spending is investment vs. maintenance) without changing VALUATION (the discounted value of free cash flows). This is not a valuation trick — it is an analytical improvement.

---

## Edge Cases

### Zero or Negative Invested Capital
Companies with negative net working capital (supplier financing exceeding operational assets) or significant accumulated losses may report negative invested capital. ROIC is undefined. Flag and use alternative metrics: ROE, Gross Profit / Total Assets, or Gross Profit / Enterprise Value.

### Loss-Making Companies
When NOPAT is negative, ROIC is negative but the magnitude is not economically meaningful (a company losing $1M on $100M of capital and one losing $10M on $100M both show negative ROIC, but the economic difference is large). For loss-making companies, focus on: (a) path to positive NOPAT, (b) revenue growth trajectory, (c) cash burn rate vs. remaining capital.

### IFRS vs. GAAP
IFRS companies report operating leases differently (depreciation + interest already separated). The NOPAT adjustment for embedded lease interest is unnecessary for IFRS filers. Always check the accounting standard before applying adjustments.

### Negative Goodwill
Bargain purchases create negative goodwill. Exclude from invested capital and note that reported ROIC is inflated.

# Source Attribution

This file records source material used in this skill's methodology, per FR-025.

Format: `{source_dir}/{filename}.{ext}::Report Title, Author, Date`

## Sources

- `MorganStanley/financial-analysis/MS-Counterpoint-Global-2022-return-on-invested-capital-how-to-calculate-roic-and-handle-common-issues.json::Return on Invested Capital: How to Calculate ROIC and Handle Common Issues, Mauboussin/Callahan, October 2022`
- `MorganStanley/financial-analysis/ROIC-and-the-Investment-Process-Mauboussin-2023.json::ROIC and the Investment Process, Mauboussin/Callahan, June 2023`
- `MorganStanley/financial-analysis/ROIC-and-Intangible-Assets-Mauboussin-2022.json::ROIC and Intangible Assets, Mauboussin/Callahan, November 2022`
- `MorganStanley/financial-analysis/MS-Counterpoint-Global-2022-intangibles-and-earnings-improving-the-usefulness-of-financial-statements.json::Intangibles and Earnings: Improving the Usefulness of Financial Statements, Mauboussin/Callahan, April 2022`
- `MorganStanley/financial-analysis/MS-Counterpoint-Global-2021-categorizing-for-clarity-cash-flow-statement-adjsutments-to-improve-insight.json::Categorizing for Clarity: Cash Flow Statement Adjustments to Improve Insight, Mauboussin/Callahan, October 2021`
- `MorganStanley/financial-analysis/MS-Counterpoint-Global-2021-the-impact-of-intangibles-on-base-rates.json::The Impact of Intangibles on Base Rates, Mauboussin/Callahan, June 2021`

# ratio-analysis — Tool Fallbacks

Extracted from SKILL.md for progressive disclosure (US5).

| Tool | Failure Mode | Fallback Action | Coverage Annotation |
|------|-------------|-----------------|---------------------|
| `get_realtime_quote` | Rate limit / unavailable | Use `search_earnings_calendar` for EPS estimates; flag valuation ratios as "price data unavailable" | "Real-time price unavailable; valuation ratios omitted" |
| `search_xbrl_facts` | Concept not found | Try alternative concept names via `list_xbrl_concepts` | "XBRL concept unavailable; used alternative" |
| `search_companies` | Sector undefined | Use SIC code from `get_company_profile` | "Peer identification via SIC code" |

# EBITDA Adjustment Methodology -- Institutional Standards

## Protocol

### EBITDA Calculation Architecture
EBITDA is a non-GAAP financial metric constructed through three distinct pathways, each serving a different analytical purpose and data availability profile.

The top-down approach starts with operating income (EBIT) from the income statement and adds back depreciation and amortization from the cash flow statement. This is the standard institutional method when a clean operating income line is available, as D&A is not typically disclosed with line-item granularity on the face of the income statement. The supporting mechanics: since GAAP income statements embed D&A within cost of goods sold, R&D, and SG&A without separate disclosure, analysts must retrieve D&A from the cash flow statement's operating activities section (where it appears as a reconciling item from net income to cash from operations) and add it back to EBIT.

The bottom-up approach starts with net income and reconstructs upward: net income + taxes + interest expense (net of interest income) + depreciation + amortization. This method is used when a clean operating income figure is unavailable or when the analyst needs to verify consistency between the top-down and bottom-up figures. Both approaches must produce identical EBITDA values; divergence indicates either a data extraction error or a classification difference that must be reconciled before proceeding to adjustments.

The direct approach (Revenue - COGS_excl_D&A - R&D_excl_D&A - SG&A_excl_D&A) is theoretically ideal but practically infeasible because public companies do not disclose the D&A component embedded within each expense category. This approach exists conceptually to illustrate why the cash flow statement is the necessary source for D&A in EBITDA calculations.

### The GAAP to Non-GAAP Adjustment Bridge
The progression from GAAP net income to adjusted EBITDA follows a structured reconciliation ladder. The first bridge step is EBITDA itself (net income + interest + taxes + D&A), representing unlevered operating profitability before non-cash charges. The second bridge step adds back items that are embedded within GAAP operating expenses but are not reflective of ongoing operating performance: stock-based compensation expense, acquisition-related intangible amortization, inventory step-up amortization from purchase price allocations, and depreciation (if moving toward an EBITDAC or EBITDAX variant).

The third bridge step addresses non-recurring, unusual, or infrequent items recognized within operating income under GAAP. These include restructuring charges, severance costs, one-time asset write-offs, gains or losses on asset sales, litigation settlements, and impairment charges. Under GAAP, these items must be reported within operating income rather than below the line; FASB does not permit their exclusion from earnings. The analyst reclassifies them to arrive at a normalized operating profit measure.

### Recurring vs. Non-Recurring Classification Framework
Classification of items as recurring or non-recurring is the central analytical judgment in EBITDA adjustment. The institutional test is three-part: frequency (has the charge appeared in two or more of the last eight quarters?), nature (is the charge related to ongoing business operations or a discrete identifiable event?), and magnitude (is the charge material enough to distort period-over-period comparisons?). Items that fail any single prong warrant scrutiny but do not automatically qualify as non-recurring. Items failing two prongs (particularly frequency and nature) are presumptively recurring and should remain in adjusted EBITDA. A company reporting "restructuring charges" for three consecutive years is incurring recurring operational restructuring, not non-recurring events, regardless of management's characterization.

## Practitioner Standards

### Management Adjustments vs. Credit Adjustments
Two distinct frameworks govern EBITDA adjustment philosophy. Management adjustments reflect the company's own characterization of items as non-recurring or non-operational. These adjustments appear in earnings press releases and investor presentations, and their primary objective is to present the most favorable normalized earnings profile. Investment banks typically accept management's adjustments at face value when preparing pitchbook valuation summaries and fairness opinions; sell-side equity research analysts are moderately more skeptical, accepting the disclosed EBITDA while arguing for a valuation multiple discount to reflect lower earnings quality.

Credit adjustments reflect a lender's perspective and are substantially more conservative. Creditors disallow add-backs for items that represent genuine economic costs to the business: stock-based compensation (dilutive to equity holders even if non-cash, and reflective of true employee compensation cost), cash restructuring charges (real cash outflows), and litigation settlements (cash payments that reduce funds available for debt service). Credit-adjusted EBITDA is systematically lower than management-adjusted EBITDA, and the spread between the two measures quantifies the degree of earnings management in the company's disclosure.

### Specific Add-Back Categories and Analytical Treatment

**Stock-based compensation**: GAAP requires expensing SBC at grant-date fair value, producing a non-cash charge embedded within COGS, R&D, and SG&A. Most equity analysts exclude SBC from adjusted EBITDA, treating it as a non-cash item analogous to depreciation. This practice is controversial: while SBC is non-cash, it represents real economic dilution to existing shareholders and is a genuine compensation expense. For DCF valuation, excluding SBC from EBITDA while also failing to account for the dilutive impact of future option exercises systematically overstates intrinsic value. The analytical compromise is to exclude SBC from adjusted EBITDA but apply the treasury stock method to the fully diluted share count, capturing the dilutive economic cost through the share count rather than the earnings measure.

**Restructuring charges**: Embedded within SG&A or operating expenses per GAAP. Cash restructuring charges represent real economic costs and should be excluded from adjusted EBITDA only if genuinely non-recurring. A company with restructuring charges in three of the last four years is undergoing continuous restructuring; these charges are operational, not extraordinary. The non-recurring classification requires that the event be both unusual in nature and infrequent in occurrence.

**Litigation gains and losses**: Classified as operating items under GAAP. The adjustment treatment depends on whether the litigation relates to ongoing business operations (e.g., recurring patent defense costs for a pharmaceutical company -- do not adjust) or discrete events outside normal operations (e.g., a one-time antitrust settlement -- adjust). The burden of proof for non-recurring classification rests with the analyst, not management.

**Foreign exchange gains and losses**: Generally considered operating in nature for multinational companies and should not be adjusted out of EBITDA. Exception: hyperinflationary economies where currency translation produces extreme period-over-period distortions that mask operating trends; in such cases, the adjustment should be clearly footnoted with the rationale.

**Pension-related charges**: Settlement charges and curtailment gains/losses are typically adjusted as non-recurring. Ongoing service cost and interest cost components of pension expense are operational and should remain in adjusted EBITDA.

**Acquisition-related costs**: Transaction advisory fees, integration costs, and contingent consideration re-measurement adjustments are generally classified as non-recurring and adjusted out of EBITDA. However, serial acquirers with acquisition costs in every period raise the same recurring-vs-non-recurring question: if the company is on an acquisition program, these are effectively recurring operating costs of the growth strategy.

### Tax Effect Consistency
When pre-tax non-recurring items are excluded from EBITDA, the associated tax shield must also be excluded. The logic: if the adjustment pretends the item did not occur, the tax effect of the item must also be eliminated. The standard approach applies the company's normalized marginal tax rate (not the effective rate, which may embed the item's own tax effect) to pre-tax adjustments. Companies that present pre-tax adjustments without corresponding tax-effect adjustments are overstating adjusted net income by the amount of the unadjusted tax benefit.

### Creditor vs. Sponsor Perspectives
Credit analysis demands the most conservative EBITDA definition: all cash costs remain in EBITDA, non-cash items are scrutinized for future cash conversion, and the sustainability of adjustments is challenged. Private equity sponsors evaluating a platform acquisition adopt a more permissive adjustment framework, anticipating that post-acquisition operational improvements will make certain current costs genuinely non-recurring. The analyst must identify which perspective is relevant to the specific analytical context and apply adjustment criteria consistently.

### Prevalence and Earnings Quality Signals
Adjusted EBITDA disclosures in public filings have increased substantially, with the number of large-cap companies referencing the metric more than doubling between 2010 and 2015. While disclosures are technically transparent (the reconciliation from GAAP to non-GAAP is provided), the proliferation of adjusted metrics has created a de facto alternative reporting framework that many analysts accept without sufficient scrutiny. A company with adjusted EBITDA exceeding GAAP net income by more than 50% is exhibiting a significant adjustment burden; companies whose adjusted EBITDA is more than double GAAP net income warrant heightened skepticism regarding earnings quality. The appropriate institutional response is not to reject adjusted EBITDA but to independently verify each adjustment against the recurring/non-recurring classification framework and apply a valuation multiple discount proportional to the adjustment burden.