Equity drawdown depth and observed recovery duration

A decline has a depth and an observed timespan

Two account-value series can fall by the same percentage and take different amounts of time to regain their previous peaks. This worksheet keeps those questions separate. It retains the original calculator’s maximum and current drawdown and adds an episode ledger: which observation was the peak, which was the trough, and when the first recovery was actually observed.

The calculations follow the records you supply. They do not reconstruct a continuous equity curve. An account may have moved between two statements, and a monthly observation cannot establish its intraday low. A recorded recovery is a historical observation, not a promised timetable for the next decline.

Start with the value and date your statement actually reports

This guide uses US account-statement terminology as a concrete recordkeeping context. Investor.gov’s illustrated statement separates beginning and ending account values from deposits, withdrawals, fees and changes in investments. FINRA likewise describes account activity and money moving into or out of an account. Those sources help identify the input; they do not certify this calculator or create a US-only mathematical method.

Copy the statement’s valuation date, value definition and unit into your source notes. Do not substitute the date you downloaded the PDF for the date the reported value describes. If the statement names only a period-end date, retain that precision. A convenient assumption of midnight or the market close would add a time the source did not establish.

Keep a reference to the original document locally. A balance, liquidation value and a separately constructed performance index are different series. Pick one consistent definition before combining observations. The worksheet cannot detect that two identical column labels came from incompatible statement conventions.

Sources: SEC Investor.gov — Better Understanding Your Brokerage Account Statement, FINRA — Understanding Brokerage Account Statements and Trade Confirmations.

Keep cash movements out of an unsupported performance claim

A deposit can lift an account above an earlier value without recovering an investment loss. A withdrawal can create a decline without a corresponding market loss. If external cash flows are present or unknown, the arithmetic still describes the supplied values, but its result should not be labelled investment performance.

The cash-flow selector records your description only. No external flows means you have established that condition for your selected interval. Already adjusted means you supply a separately prepared series and explain its method. Neither selection makes the page verify that work. The default remains unknown, and no amount is automatically subtracted or added.

For example, a hypothetical account falling from 100 to 80 and then receiving a 25 deposit shows a raw balance of 105. The supplied-value method observes a return above 100. That alone cannot establish recovery of investment performance. Preserve this distinction before presenting the episode to another reader.

Define the peak, trough and recovery once

For each row, the running peak is the largest equity value observed so far. Drawdown is 100 multiplied by running peak minus current value, divided by running peak. Maximum drawdown is the largest such percentage in the file; current drawdown uses its final row. Research on maximum drawdown describes the underlying peak-to-trough concept; the detailed episode rules here are the worksheet’s stated convention.

An episode begins when a value falls below its saved peak. Its trough is the lowest subsequent observation before recovery. Recovery is the first supplied observation equal to or greater than that peak. If equal peak values occur before a decline, use the latest one as the start. If the same trough value occurs several times, retain its first occurrence.

A later, higher peak starts the reference for later declines. Separate episodes are not merged merely because their percentages match. These choices make the dates reproducible; another analysis using a different tie rule or observation frequency may report different durations.

Sources: Goldberg and Mahmoud — Drawdown: From Practice to Theory and Back Again.

Read three durations without mixing their endpoints

Peak-to-trough measures the interval from the selected peak observation to the selected trough. Trough-to-recovery measures the interval from that trough to the first observed recovery. Peak-to-recovery spans the full completed episode. These durations share endpoints, but they describe different parts of the record.

For the hypothetical values 1,000, 1,200, 900, 1,100 and 1,250 on 1–5 January, the peak is 2 January and the trough is 3 January. Depth is 25%. Recovery is first observed on 5 January. In date-only mode the durations are one, two and three calendar days respectively. Maximum drawdown remains 25%, while current drawdown is zero.

Choose calendar dates or actual instants

Date-only mode accepts YYYY-MM-DD and measures differences between calendar dates. It does not claim those values were measured at 00:00 UTC, and it cannot provide elapsed hours. Instant mode requires a complete timestamp with Z or a numeric offset and measures elapsed seconds after converting those timestamps to UTC. The distinction follows the separation between a full date and a date-time in RFC 3339.

A New York offset example shows why the choice matters. Hypothetical observations at 16:00 on 31 October 2026 with -04:00 and 16:00 on 2 November with -05:00 are 49 elapsed hours apart. Their calendar dates are two days apart. NIST documents the US daylight-time change on 1 November; neither answer is interchangeable with the other.

These are invented observation times, including weekend dates, not claims about exchange opening hours or trading sessions. This tool calculates no trading-day count. It also rejects leap-second timestamps and does not model leap seconds in elapsed-clock arithmetic.

Sources: RFC 3339 — Date and Time on the Internet: Timestamps, NIST — Daylight Saving Time Rules, NIST — Local Time FAQs.

An unfinished episode must stay unfinished

If the final value remains below the peak, the recovery date is unknown. Both completed recovery durations stay blank in the ledger, represented internally as null. A separate span from peak to the last observation is available. Calling that span a recovery time would turn an unresolved event into a completed one.

A missing equity cell stops the calculation rather than becoming zero. A gap between two valid dates is different: both observations can be used, but the path between them remains unknown. The worksheet reports the observed interval range and does not invent missing rows, a daily schedule or market holidays.

A return above the peak after a long gap is only the first recovery you observed. Actual recovery could have occurred earlier, and a deeper intervening trough could be absent from the file. Keep the original sampling description with every exported result.

Share the ledger with its observation limits

Keep the CSV, source note, unit, time mode and cash-flow description together. State the observation count and final supplied date. If you compare two analyses, first compare their input rows and tie conventions; a different maximum or duration does not by itself establish an error.

The formulas, worked examples and boundaries are available without JavaScript. The local calculator and CSV export make the same method easier to repeat. The existing resource comparison and learning hub remain related references; the publisher’s Telegram offer is a separate optional next step.

Worked examples

Questions before you use the result

Primary sources and scope

1. SEC Investor.gov — Better Understanding Your Brokerage Account StatementChecked 2026-10-20. Illustrates statement periods, account values, deposits/withdrawals and investment-value changes as separate information.
2. FINRA — Understanding Brokerage Account Statements and Trade ConfirmationsChecked 2026-10-20. Account activity includes trades and money going in or out; used for record provenance, not a legal compliance or performance certification claim.
3. Goldberg and Mahmoud — Drawdown: From Practice to Theory and Back AgainChecked 2026-10-20. Research defines maximum drawdown through peak-to-trough decline. The worksheet states its own discrete observation and tie rules.
4. RFC 3339 — Date and Time on the Internet: TimestampsChecked 2026-10-20. Distinguishes full-date from date-time with explicit Z/numeric offset; supports separating date labels from instants.
5. NIST — Daylight Saving Time RulesChecked 2026-10-20. US daylight time ends on November 1, 2026; used for the explicitly hypothetical New York observation example.
6. NIST — Local Time FAQsChecked 2026-10-20. Eastern daylight/standard UTC offsets provide context for recorded New York instants.

Prepared from the sources listed above. Examples and calculations explain a method; they do not predict prices or certify an investment. Geographic time references do not establish where a trading service is available. Dated schedules can change: consult the named publisher before relying on a future event.

Related resources on this site

• Historical Equity Drawdown Resources
• 10 Services for Reviewing an Equity Curve