1. Company Fundamentals

Each metric notes which statement it comes from: IS = Income Statement, BS = Balance Sheet, CF = Cash Flow Statement.

Algo-wide warning. Fundamentals are restated and reported with a lag (often 30–90 days after quarter-end). Backtesting on values keyed to the fiscal-period date introduces lookahead. Key every fundamental to its filing/announcement date, not the period it covers.

The Hierarchy (how everything connects)

Fundamentals are a derivation tree. Don't memorize ratios in isolation — know what they're built from. Everything bottoms out in line items reported on the three statements.

Metric pyramid

Algo implication. Pull and store Level 0 line items; compute everything above yourself. That way one PIT (point-in-time) snapshot reconstructs every ratio consistently, and you're never at the mercy of a vendor's differing definition of "EBITDA" or "invested capital."

Income Statement waterfall (IS)

Balance Sheet structure (BS) — Assets = Liabilities + Equity

$$\text{Working Capital}=\text{Current Assets}-\text{Current Liabilities}$$ $$\text{Invested Capital}=\text{Total Debt}+\text{Equity}-\text{Cash}$$ $$\text{Total Debt}=\text{Short-term debt}+\text{Long-term debt}$$

Cash Flow structure (CF) — bridges Net Income to actual cash

$$\text{Free Cash Flow (FCF)}=\text{CFO}-\text{Capex}$$

Building Blocks (Level 0–1 line items)

The raw quantities the ratios above are made of. Source: IS unless noted.

Revenue (Sales / Top Line)

Def. Total value of goods/services sold in the period. Formula. $\text{Revenue}=\text{Units}\times \text{Price}$ (reported directly). Net revenue = gross − returns/allowances. Signal. The growth engine; everything downstream is a slice of it. Watch organic vs acquired growth. Algo. Recognition timing differs (ASC 606); deferred revenue (BS liability) ≠ recognized revenue. Use net, not gross.

COGS (Cost of Goods Sold)

Def. Direct costs of producing what was sold. Formula. $\text{COGS}=\text{Beginning Inventory}+\text{Purchases}-\text{Ending Inventory}$. Signal. Drives gross margin; rising COGS/Rev → input-cost or pricing pressure. Algo. FIFO vs LIFO inventory accounting changes COGS (and taxes) — note the method when comparing firms.

Gross Profit

Def. Revenue left after direct production costs. Formula. $\text{Gross Profit}=\text{Revenue}-\text{COGS}$. Signal. The first profitability checkpoint; high & stable → pricing power. Algo. Some firms bury D&A in COGS, others in opex — distorts cross-company gross margin.

Operating Expenses (SG&A, R&D)

Def. Costs of running the business not tied directly to production. Formula. $\text{OpEx}=\text{SG\&A}+\text{R\&D}+\text{other operating}$. SG&A = selling, general & admin. Signal. R&D intensity signals reinvestment (tech/pharma); bloated SG&A → inefficiency. Algo. R&D is expensed (US GAAP) not capitalized → understates asset base for R&D-heavy firms; some quants capitalize it manually.

D&A (Depreciation & Amortization)

Def. Non-cash allocation of asset cost (depreciation = tangible, amortization = intangible) over its life. Formula. Straight-line: $\frac{\text{Cost}-\text{Salvage}}{\text{Useful Life}}$. Found on CF (added back) and within IS opex/COGS. Signal. The key non-cash charge bridging accrual earnings and cash; the "DA" in EBITDA. Algo. Pull from the CF statement (cleanest source); accelerated vs straight-line schedules differ.

EBIT (Operating Income)

Def. Earnings before interest & taxes; pure operating profitability, capital-structure-neutral. Formula. $\text{EBIT}=\text{Gross Profit}-\text{OpEx}=\text{Revenue}-\text{COGS}-\text{OpEx}$. Signal. Compare operations across firms regardless of debt/tax. Numerator of ROIC (after tax) and interest coverage. Algo. "Operating income" and EBIT can differ slightly (non-operating items) — reconcile if precision matters.

EBITDA

Def. EBIT plus the non-cash D&A charge; a rough cash-operating-earnings proxy. Formula. $\text{EBITDA}=\text{EBIT}+\text{D\&A}$. Signal. Capital-structure- and depreciation-policy-neutral; basis of EV/EBITDA and Net Debt/EBITDA. Algo. Ignores capex & working-capital needs — flatters capital-intensive firms. "Adjusted EBITDA" adds back management's pet exclusions; treat skeptically.

Interest Expense (net)

Def. Cost of debt financing for the period. Formula. $\approx \text{Avg Debt}\times \text{Effective Interest Rate}$. Net = interest expense − interest income. Signal. Denominator of interest coverage; high → leverage strain. Algo. Firms net interest income against expense — read the notes before using as a pure debt cost.

Pre-tax Income (EBT)

Def. Earnings before taxes; EBIT minus net interest (and non-operating items). Formula. $\text{EBT}=\text{EBIT}-\text{Interest Expense}\pm \text{non-operating items}$. Signal. Base for the effective tax rate calc. Algo. One-off non-operating gains/losses live here — normalize them out for trend analysis.

Income Tax / Effective Tax Rate

Def. Taxes on pre-tax income. Formula. $\text{Effective Tax Rate}=\frac{\text{Income Tax Expense}}{\text{Pre-tax Income}}$. Signal. Low/volatile rate → tax credits, foreign mix, or one-offs; not sustainable if abnormal. Algo. Use effective rate to compute NOPAT for ROIC; cash taxes (CF) differ from book taxes (IS).

Net Income (Bottom Line)

Def. Final accrual profit after all costs, interest, and taxes. Formula. $\text{Net Income}=\text{EBT}-\text{Taxes}$ (− minority interest/preferred for "to common"). Signal. Feeds EPS, ROE, P/E, payout ratio. The headline, but the most manipulable (accruals). Algo. Use net income attributable to common for per-share metrics; strip preferred dividends.

Working Capital

Def. Short-term operating liquidity tied up in the business. Formula. $\text{Working Capital}=\text{Current Assets}-\text{Current Liabilities}$. Source: BS. Signal. Rising WC consumes cash (gap between earnings and CFO); negative WC can be efficient (e.g. retailers float on payables). Algo. Δ Working Capital is the reconciling item between Net Income and CFO — watch it for earnings-quality red flags.

Capex (Capital Expenditures)

Def. Cash spent on long-lived productive assets (PP&E). Formula. Reported in CFI; $\text{Capex}\approx \Delta \text{Gross PP\&E}+\text{Depreciation}$. Maintenance vs growth capex differ. Signal. High capex/sales → capital-intensive; the subtraction in FCF. Algo. FCF = CFO − Capex hinges on this; separating maintenance from growth capex is judgment, document your split.

NOPAT (Net Operating Profit After Tax)

Def. After-tax operating profit, ignoring financing — the return to all capital providers. Formula. $\text{NOPAT}=\text{EBIT}\times (1-\text{effective tax rate})$. Signal. Numerator of ROIC; the cleanest measure of operating earning power. Algo. Pair with Invested Capital $(\text{Debt}+\text{Equity}-\text{Cash})$ for ROIC; keep definitions of both consistent across the universe.

Valuation

Price-to-Earnings (P/E)

Def. Price paid per dollar of earnings. Formula. $\text{P/E}=\frac{\text{Price}}{\text{EPS}}$ (TTM or forward). Source: Price + IS (net income). Signal. High → growth expectations or overvaluation; low → value or distress/declining earnings. Algo. Undefined/negative when EPS ≤ 0 — mask or treat as NaN, never rank raw. "Forward" P/E uses analyst estimates (not point-in-time safe unless you store vintage estimates).

Forward vs Trailing P/E

Def. Trailing uses realized TTM EPS; forward uses estimated next-12-month EPS. Formula. $\text{Trailing}=\frac{P}{{\text{EPS}}_{\text{ttm}}}$, $\text{Forward}=\frac{P}{{\text{EPS}}_{\text{est}}}$. Signal. Forward < trailing → earnings expected to grow. Algo. Estimates revise constantly; only use stored-as-of estimates in backtests.

PEG Ratio

Def. P/E normalized by growth rate. Formula. $\text{PEG}=\frac{\text{P/E}}{\text{annual EPS growth \%}}$. e.g. P/E 20, growth 20% → PEG 1.0. Signal. <1 → cheap relative to growth; >2 → expensive. Algo. Extremely sensitive to the growth input (historical vs forecast). Document which you use.

Price-to-Book (P/B)

Def. Price vs net asset (book) value. Formula. $\text{P/B}=\frac{\text{Price}}{\text{Book Value per Share}}$, where $\text{BVPS}=\frac{\text{Total Equity}-\text{Preferred}}{\text{Shares}}$. Source: BS. Signal. <1 → trades below accounting net worth (value or impaired assets). Most useful for banks/financials. Algo. Meaningless for asset-light/IP-heavy firms (book understates value). Negative equity → garbage ratio.

Price-to-Sales (P/S)

Def. Price per dollar of revenue. Formula. $\text{P/S}=\frac{\text{Market Cap}}{{\text{Revenue}}_{\text{ttm}}}$. Source: IS. Signal. Used when earnings are negative (early growth/biotech). High → rich. Algo. Ignores margins entirely — compare only within sector.

EV/EBITDA

Def. Enterprise value per dollar of pre-D&A operating earnings; capital-structure-neutral. Formula. $\text{EV}=\text{Market Cap}+\text{Total Debt}-\text{Cash \& Equivalents}$, $\text{EBITDA}=\text{EBIT}+\text{D\&A}$, so $\text{EV/EBITDA}=\frac{\text{EV}}{\text{EBITDA}}$. Signal. Lower → cheaper. Preferred for cross-company comparison because it's debt- and tax-neutral. Algo. Compute EV from PIT debt/cash. EBITDA excludes capex — bad proxy for capital-intensive firms.

EV/Sales, EV/FCF

Def. EV scaled by revenue or free cash flow. Formula. $\text{EV/Sales}=\frac{\text{EV}}{\text{Revenue}}$, $\text{EV/FCF}=\frac{\text{EV}}{\text{Free Cash Flow}}$. Signal. EV/FCF is the cash-honest valuation multiple; lower = better. Algo. EV/FCF blows up when FCF near zero/negative.

EV/EBIT

Def. EV per dollar of operating profit; like EV/EBITDA but charges for D&A (capital intensity). Formula. $\text{EV/EBIT}=\frac{\text{EV}}{\text{EBIT}}$. Signal. Lower → cheaper. Penalizes capex-heavy firms that EV/EBITDA flatters — a stricter comparison. Algo. Prefer over EV/EBITDA for capital-intensive sectors (telecom, industrials, utilities).

P/FCF & P/OCF (Price to Free / Operating Cash Flow)

Def. Equity-level cash-flow multiples (vs the EV-level ones above). Formula. $\text{P/FCF}=\frac{\text{Market Cap}}{\text{FCF}}$, $\text{P/OCF}=\frac{\text{Market Cap}}{\text{Operating Cash Flow}}$. Signal. Cash-based alternatives to P/E that sidestep accrual/earnings manipulation. Lower = cheaper. Algo. P/OCF avoids the capex-definition ambiguity in FCF; use it when capex classification is noisy.

PEGY

Def. PEG that also credits dividend yield — for income+growth names. Formula. $\text{PEGY}=\frac{\text{P/E}}{\text{EPS growth \%}+\text{dividend yield \%}}$. Signal. <1 → cheap relative to total shareholder return. Fairer than PEG for dividend payers. Algo. Same growth-input sensitivity as PEG; document the growth source.

Earnings Yield

Def. Inverse of P/E; earnings as a % of price. Comparable to bond yields. Formula. $\text{Earnings Yield}=\frac{\text{EPS}}{\text{Price}}=\frac{1}{\text{P/E}}$. Signal. Compare to 10Y Treasury yield (the "Fed model"): higher earnings yield → equity relatively cheap. Algo. Well-defined even when EPS negative (just negative) — more robust to rank than P/E.

Dividend Yield

Def. Annual dividend as % of price. Formula. $\text{Div Yield}=\frac{\text{Annual Dividends per Share}}{\text{Price}}$. Signal. High → income or a falling price (yield trap). Check payout ratio for sustainability. Algo. Adjust for special dividends; "indicated" (forward) vs trailing yield differ.

Profitability & Returns

Gross / Operating / Net Margin

Def. Profit at three levels as % of revenue. Formula. $\text{Gross}=\frac{\text{Rev}-\text{COGS}}{\text{Rev}}$, $\text{Operating}=\frac{\text{EBIT}}{\text{Rev}}$, $\text{Net}=\frac{\text{Net Income}}{\text{Rev}}$. Source: IS. Signal. Rising margins → pricing power / scale. Net margin reflects everything incl. taxes & interest. Algo. Margins are sector-relative; software ~80% gross, retail ~25%. Never compare cross-sector.

Return on Equity (ROE)

Def. Net income earned per dollar of shareholder equity. Formula. $\text{ROE}=\frac{\text{Net Income}}{\text{Avg Shareholders’ Equity}}$. Source: IS + BS. Signal. High → efficient; but can be inflated by leverage. Algo. Decompose via DuPont (below) before trusting it. Negative equity → meaningless.

DuPont Decomposition

Def. Breaks ROE into operating, efficiency, and leverage drivers. Formula. $\text{ROE}=\text{Net Margin}\times \text{Asset Turnover}\times \text{Equity Multiplier}=\frac{\text{NI}}{\text{Rev}}\times \frac{\text{Rev}}{\text{Assets}}\times \frac{\text{Assets}}{\text{Equity}}$. Signal. Identifies why ROE is high — margins (good) vs leverage (risky). Algo. Each factor sourced from different statements; align periods (all TTM or all FY).

Return on Assets (ROA)

Def. Profit per dollar of total assets. Formula. $\text{ROA}=\frac{\text{Net Income}}{\text{Avg Total Assets}}$. Signal. Leverage-neutral profitability; high → efficient asset use. Algo. Banks have very low ROA by nature — sector-relative.

Return on Invested Capital (ROIC)

Def. After-tax operating return on all invested capital (debt + equity). Formula. $\text{ROIC}=\frac{\text{NOPAT}}{\text{Invested Capital}}$, where $\text{NOPAT}=\text{EBIT}\times (1-\text{tax rate})$, $\text{Invested Capital}=\text{Debt}+\text{Equity}-\text{Cash}$. Signal. ROIC > WACC → company creates value; ROIC < WACC → destroys it. The cleanest quality metric. Algo. Definitions of invested capital vary — pick one and apply consistently across the universe.

Efficiency / Activity Ratios

How hard the balance sheet works — turnover of assets, inventory, and receivables. These feed DuPont (asset turnover) and the cash cycle.

Asset Turnover

Def. Revenue generated per dollar of assets. Formula. $\text{Asset Turnover}=\frac{\text{Revenue}}{\text{Avg Total Assets}}$. Source: IS + BS. Signal. High → asset-light/efficient (retail, software); low → capital-intensive (utilities). The middle term of DuPont ROE. Algo. Use average assets (begin+end)/2 to match the flow (revenue) to the stock (assets); strongly sector-dependent.

Inventory Turnover & Days Inventory (DIO)

Def. How many times inventory is sold and replaced per year. Formula. $\text{Inv Turnover}=\frac{\text{COGS}}{\text{Avg Inventory}}$, $\text{DIO}=\frac{365}{\text{Inv Turnover}}$. Signal. High turnover → lean operations; low/falling → overstocking or weakening demand. Algo. Use COGS (not revenue) in the numerator — inventory is carried at cost. Seasonal businesses distort point-in-time inventory.

Receivables Turnover & Days Sales Outstanding (DSO)

Def. How fast the firm collects cash from credit sales. Formula. $\text{Rec Turnover}=\frac{\text{Revenue}}{\text{Avg Receivables}}$, $\text{DSO}=\frac{365}{\text{Rec Turnover}}$. Signal. Rising DSO → customers paying slower (cash-flow / credit-quality warning), or aggressive revenue recognition. Algo. A jump in DSO vs revenue growth is a classic earnings-quality red flag.

Payables Turnover & Days Payable Outstanding (DPO)

Def. How long the firm takes to pay suppliers. Formula. $\text{Pay Turnover}=\frac{\text{COGS}}{\text{Avg Payables}}$, $\text{DPO}=\frac{365}{\text{Pay Turnover}}$. Signal. High DPO → using suppliers as free financing (good for cash), but too high strains relationships. Algo. Pairs with DIO/DSO into the cash conversion cycle below.

Cash Conversion Cycle (CCC)

Def. Days from paying for inventory to collecting cash from its sale — the working-capital cash gap. Formula. $\text{CCC}=\text{DIO}+\text{DSO}-\text{DPO}$. Signal. Lower → less cash tied up; negative CCC (Amazon, Dell) = customers pay before suppliers do = self-funding growth. Algo. The single best operational-efficiency summary; track its trend, not just the level.

Leverage, Liquidity & Coverage

Debt-to-Equity (D/E)

Def. Leverage relative to equity. Formula. $\text{D/E}=\frac{\text{Total Debt}}{\text{Total Equity}}$. Source: BS. Signal. High → more financial risk; norms vary wildly by sector (utilities high, software low). Algo. Use "total debt" (short + long), not "total liabilities" — be explicit which.

Net Debt / EBITDA

Def. Years of EBITDA needed to repay net debt. Formula. $\frac{\text{Total Debt}-\text{Cash}}{\text{EBITDA}}$. Signal. >4–5 → highly levered; covenant territory. Lower → safer. Algo. Key leverage screen for credit risk; negative net debt = net cash position.

Current Ratio / Quick Ratio

Def. Short-term liquidity: ability to cover current liabilities. Formula. $\text{Current}=\frac{\text{Current Assets}}{\text{Current Liabilities}}$, $\text{Quick}=\frac{\text{Current Assets}-\text{Inventory}}{\text{Current Liabilities}}$. Signal. >1 → can cover near-term obligations. Quick is stricter (excludes hard-to-sell inventory). Algo. Too high can signal idle cash; sector-relative.

Cash Ratio

Def. Strictest liquidity test — covers current liabilities with cash alone. Formula. $\text{Cash Ratio}=\frac{\text{Cash}+\text{Cash Equivalents}}{\text{Current Liabilities}}$. Source: BS. Signal. >1 → can extinguish all short-term debt immediately; rarely needed that high. The most conservative of the liquidity ratios. Algo. Excludes receivables/inventory entirely; useful as a distress/solvency screen.

Interest Coverage

Def. How many times operating earnings cover interest expense. Formula. $\text{Interest Coverage}=\frac{\text{EBIT}}{\text{Interest Expense}}$. Signal. <1.5 → distress risk; high → comfortable. Algo. Interest expense can be netted against interest income in filings — read the notes.

Altman Z-Score

Def. Composite bankruptcy-risk score. Formula. $Z=1.2A+1.4B+3.3C+0.6D+1.0E$, where $A=\text{WC}/\text{TA}$, $B=\text{RE}/\text{TA}$, $C=\text{EBIT}/\text{TA}$, $D=\text{Mkt Equity}/\text{Tot Liab}$, $E=\text{Sales}/\text{TA}$. Signal. Z >2.99 safe; 1.81–2.99 grey; <1.81 distress. Algo. Original coefficients are for manufacturers; use Z'' variant for non-manufacturers/EM.

Cash Flow

Free Cash Flow (FCF)

Def. Cash left after funding operations and capex; what's actually distributable. Formula. $\text{FCF}=\text{Operating Cash Flow}-\text{Capital Expenditures}$. Source: CF. Signal. Positive & growing → self-funding; the hardest number to fake (vs accrual earnings). Algo. "Levered" FCF (after interest) vs "unlevered" FCFF differ — match to the discount rate you use.

FCF Yield

Def. FCF as % of market cap or EV; cash-based valuation. Formula. $\text{FCF Yield}=\frac{\text{FCF}}{\text{Market Cap}}$, or $\frac{\text{FCF}}{\text{EV}}$. Signal. High → cheap on cash terms; a robust quality+value signal. Algo. More reliable than earnings yield for screening (less accounting noise).

Cash Conversion / Accruals

Def. How much of reported earnings shows up as actual cash. Formula. $\text{Cash Conversion}=\frac{\text{Operating Cash Flow}}{\text{Net Income}}$, $\text{Accruals}=\frac{\text{NI}-\text{OCF}}{\text{Assets}}$. Signal. Ratio <1 persistently → earnings quality concern (aggressive accruals). High accruals predict underperformance. Algo. A classic quant quality factor — low accruals = higher-quality earnings.

Growth & Per-Share

EPS (Basic vs Diluted)

Def. Earnings allocated per share. Diluted assumes options/converts exercised. Formula. $\text{Basic}=\frac{\text{NI}-\text{Pref Div}}{\text{Wtd Avg Shares}}$, $\text{Diluted}=\frac{\text{NI}}{\text{Shares}+\text{dilutive securities}}$. Signal. Use diluted for valuation — it's the conservative, realistic figure. Algo. Watch buybacks: EPS can rise while net income is flat (shrinking denominator).

Revenue / EPS Growth (YoY, QoQ, CAGR)

Def. Rate of change of revenue or earnings. Formula. $\text{YoY}=\frac{X_t}{X_{t-4q}}-1$ (quarterly data), $\text{CAGR}={\left(\frac{X_{\text{end}}}{X_{\text{start}}}\right)}^{1/\text{years}}-1$. Signal. Acceleration matters more than level for momentum strategies. Algo. Use YoY (not QoQ) to neutralize seasonality. CAGR hides volatility of the path.

Sustainable Growth Rate (SGR)

Def. Max growth a firm can fund from retained earnings without new debt/equity — the bridge from profitability to growth. Formula. $\text{SGR}=\text{ROE}\times \text{retention ratio}$, where $\text{retention}=1-\text{payout ratio}$. Signal. Growing faster than SGR → must raise external capital (dilution/leverage); slower → accumulating cash. Algo. Links the Profitability (ROE) and Growth sections; compare SGR to actual revenue growth to infer financing needs.

Shares Outstanding / Buyback Yield

Def. Change in share count from buybacks/issuance. Formula. $\text{Buyback Yield}=-\frac{\Delta \text{Shares Outstanding}}{\text{Shares}}$. Combined w/ dividend = total shareholder yield. Signal. Net reduction → returns capital, boosts per-share metrics; net issuance → dilution. Algo. Use weighted-average diluted share count; SBC (stock-based comp) offsets buybacks.

Book Value / Tangible Book Value

Def. Net assets per share; tangible strips goodwill & intangibles. Formula. $\text{BVPS}=\frac{\text{Equity}}{\text{Shares}}$, $\text{TBVPS}=\frac{\text{Equity}-\text{Goodwill}-\text{Intangibles}}{\text{Shares}}$. Signal. TBV is the conservative liquidation-ish floor; key for financials. Algo. Negative TBV common after debt-funded acquisitions — handle as NaN in ratios.

Payout Ratio

Def. Share of earnings paid as dividends. Formula. $\text{Payout}=\frac{\text{Dividends}}{\text{Net Income}}$, or $\frac{\text{Dividends}}{\text{FCF}}$ for the cash version. Signal. >100% → unsustainable (paying from reserves/debt); low → room to grow dividend. Algo. FCF-based payout is the honest sustainability check.

Quick reference: which statement?