Metrics catalogue

This page lists all 100 metrics that Metrics serves, in ten categories. Request metrics by name in fields, or a whole category by the name given under each heading.

Input fields in the formulas are the statement fields described on Fundamentals fields. They are in millions of the row's currency (millions of USD on a USD row), except eps_diluted and eps_basic, which are in that currency per share. A metric never sets figures labelled in two currencies against each other; see the :currency reasons on Metrics. Metrics read EPS on the share basis of the row's price date where the basis it was filed on is confirmed, and as filed otherwise; see "Stock splits" on Metrics. The market inputs close, market_cap and shares are described under "Market inputs".

Units and conventions

Unit labels

• "ratio": a plain multiple, such as 12.5 times. Not a percentage.
• "decimal fraction": 0.25 means 25%. Not a percentage.
• "millions of USD": 4977636.9726 is USD 4.98 trillion.
• "USD per share": dollars per share.
• "days": calendar days.
• "score": a unitless number on the scale of its published model.

Values are not rounded. Every metric is nullable: it is null when an input is unavailable (the row's missing_inputs then names the input), when a denominator is zero, or when the formula has no real value. Metrics lists the reasons missing_inputs names besides an input, such as eps_diluted:zero.

Market inputs

• close: the requested share class's close, as traded, in USD per share, on the last trading day on or before the row's date. The row's date is period_end on annual and quarterly rows, priced_on on ttm rows and the trading date on daily rows.
• market_cap: the whole company's market capitalisation, in millions of USD. See market_cap below.
• shares: the company's share count on the row's date, summed over the same share classes as market_cap and consistent with that day's close across splits. Per-share metrics divide millions of USD by shares in millions, giving USD per share. When shares are unavailable, missing_inputs names shares_outstanding.

Flows, balances and TTM

Flow inputs are income-statement and cash-flow items: revenue, cost_of_revenue, gross_profit, operating_income, ebit, ebitda, pretax_income, net_income, eps_diluted, eps_basic, interest_expense, research_and_development, selling_general_admin, depreciation_amortization, operating_cash_flow, investing_cash_flow, capital_expenditures, free_cash_flow and dividends_paid_common. Every other input is a balance taken at the period end.

• annual rows: flows cover the fiscal year.
• quarterly rows: flows cover the quarter, except in the fourteen metrics marked "TTM on quarterly rows", which read trailing-twelve-month (TTM) flows: the quarter and the three before it, summed.
• ttm and daily rows: every flow is TTM.

A TTM flow needs four consecutive quarterly statements with the item present in each. A trailing EPS is the four quarters' EPS summed; where one of them is null, it is the EPS filed on an annual report, named in the row's ttm_eps_basis, or null.

Preceding and comparison periods

The preceding period is used by averaged balances, piotroski_f_score, beneish_m_score and shares_buyback_ratio:

• annual rows: the fiscal year ending a year earlier.
• quarterly rows: the quarter ending a quarter earlier.
• ttm and daily rows: the statement ending a year earlier, and the share count 365 days before the row's date.

"Average" means the value at this period end plus the value at the preceding period end, divided by two.

The comparison period is used by the growth metrics, peg_ratio and market_cap_performance:

• annual rows: the fiscal year ending a year earlier.
• quarterly rows: set by growth_basis. yoy (default) compares with the quarter ending a year earlier, qoq with the quarter ending a quarter earlier, and ttm_yoy compares TTM flows with the TTM flows ending a year earlier.
• ttm and daily rows: TTM flows against the TTM flows ending a year earlier, and market_cap 365 days before the row's date.

A year earlier is the period ending closest to 365 days before the end of the period the row reads (period_end, or statement_period_end on daily rows), within 45 days either way. A quarter earlier is the latest period ending 60 to 125 days before it. Derived fourth quarters count as periods. Where no period ends in that window, the row has no preceding or comparison period, and a metric that needs one is null, named prior:<field>. See Metrics.

Days in the period

days_inventory, days_payable and days_sales_outstanding multiply by 365 on annual, ttm and daily rows, and by 91.25 on quarterly rows.

Valuation metrics

Category name: valuation. 18 metrics.

Profitability metrics

Category name: profitability. 13 metrics.

Leverage metrics

Category name: leverage. 16 metrics.

Liquidity metrics

Category name: liquidity. 4 metrics.

Efficiency metrics

Category name: efficiency. 11 metrics.

Per-share metrics

Category name: per_share. 11 metrics. Each divides by the company's shares on the row's date, in millions.

Dividend metrics

Category name: dividends. 4 metrics.

Accounting scores

Category name: scores. 9 metrics.

Piotroski tests

Each test scores 1 when it passes. "Preceding" is the preceding period as defined under "Units and conventions".

• net_income is above zero.
• operating_cash_flow is above zero.
• net_income / total_assets is higher than in the preceding period.
• operating_cash_flow is above net_income.
• long_term_debt / total_assets is lower than in the preceding period.
• current_assets_total / current_liabilities_total is higher than in the preceding period.
• Shares are no higher than in the preceding period.
• gross_profit / revenue is higher than in the preceding period.
• revenue / total_assets is higher than in the preceding period.

Beneish indices

A suffix "p" marks the preceding period's value.

• DSRI = (accounts_receivable / revenue) / (accounts_receivable p / revenue p)
• GMI = (gross_profit p / revenue p) / (gross_profit / revenue)
• AQI = (1 − (current_assets_total + property_plant_equipment_net) / total_assets) / (1 − (current_assets_total p + property_plant_equipment_net p) / total_assets p)
• SGI = revenue / revenue p
• DEPI = (depreciation_amortization p / (property_plant_equipment_net p + depreciation_amortization p)) / (depreciation_amortization / (property_plant_equipment_net + depreciation_amortization))
• SGAI = (selling_general_admin / revenue) / (selling_general_admin p / revenue p)
• LVGI = ((current_liabilities_total + long_term_debt) / total_assets) / ((current_liabilities_total p + long_term_debt p) / total_assets p)
• TATA = (net_income − operating_cash_flow) / total_assets

Growth metrics

Category name: growth. 11 metrics. Each is (current − comparison) / absolute value of comparison, against the comparison period defined under "Units and conventions". null when the comparison value is zero (prior:<field>:zero in missing_inputs) or missing (prior:<field>).

Volatility metrics

Category name: volatility. 3 metrics. Each is the sample standard deviation of daily log returns, ln(close on a day / close on the previous trading day), multiplied by the square root of 252. Closes are the requested class's as-traded closes, so dividends are not added back and a window that includes a split date includes that day's price change. The window ends on the row's date as defined under "Market inputs".

A value needs the window's full count of returns, spanning no more than 46, 137 or 548 calendar days respectively, with a close on every trading day in the window. Otherwise it is null and missing_inputs names insufficient_price_history. A null volatility means the price history is too short or too sparse to measure, not zero volatility.