Two Meanings Of ‘Drawdown’ And The Basic Calculation
To calculate portfolio drawdown, apply this formula at any point t: Drawdown_t = (Peak_t − Value_t) / Peak_t, where Peak_t is the highest portfolio value observed up to that moment. This is not the same as retirement income drawdown, a common confusion in search results.
When someone types ‘how to calculate portfolio drawdown,’ they often collide with pension drawdown guides. In the UK and Australia, ‘drawdown’ frequently means taking flexible income from a pension pot. That is a cash-flow decision, not a risk metric. I learned this the hard way in 2017 while preparing a report for a charitable trust; a government template caused trustees to think we planned withdrawals.
For the rest of this guide, drawdown means peak-to-trough decline in market value. If you need the income version, consult a retirement adviser; the math below will not help you.
The Core Formula For Portfolio Drawdown (And What It Hides)
The formula for drawdown at time t is: Drawdown_t = (Peak_t − Value_t) / Peak_t. The critical variable is Peak_t, the highest portfolio value observed up to that moment. This running peak is not your initial investment, nor an arbitrary benchmark high from years ago.
Answering the literal question ‘What is the formula for drawdown?’ requires this nuance: it is a rolling, path-dependent ratio. If your portfolio hits $100, falls to $80, then recovers to $90, the drawdown at $90 is (100−90)/100 = 10%, not zero because you are still below the old peak. Static formulas in many textbooks miss this dynamic.
How To Calculate Drawdown For A Stock
For a single stock, the process is identical but applied to price series. Suppose a stock closes at $120, rises to $150, then falls to $110. The running peak is $150, so drawdown = (150−110)/150 = 26.67%. You do not measure from the $120 starting point; the peak is the highest closing price seen to date.
In practice, I use adjusted closing prices from a vendor like Bloomberg or Yahoo Finance to account for splits and dividends. Ignoring adjustments creates phantom drawdowns on ex-dividend dates. For example, a $100 stock paying a $2 dividend drops to $98 mechanically; without adjustment, you’d record a 2% drawdown that never happened.
Here is a small daily series to make it concrete: Day1 $100, Day2 $105, Day3 $95, Day4 $102. Running peaks: $100, $105, $105, $105. Drawdowns: 0%, 0%, 9.52%, 2.86%. This trivial example reveals why a single maximum drawdown figure hides the bumpy ride investors actually feel.
Why A 50% Drawdown Is Harder To Recover Than A 10% One
The recovery math is asymmetric, and this is the most misunderstood part of drawdown. A loss of L requires a gain of L/(1−L) just to break even. A 10% drop needs an 11.1% gain; a 20% drop needs 25%; a 50% drop needs 100% gain. That is why a 50% drawdown is not ‘five times worse’ than 10% — it is exponentially more punishing.
A portfolio that loses 50% must double just to return to square one. The opportunity cost and emotional toll during that climb are rarely factored into naive risk models.
I once managed a long-short equity sleeve that suffered a 38% drawdown in late 2008. The subsequent 61% gain required to recover felt endless, and we underperformed peers who capped downside near 20%. The recovery hurdle is the silent killer of compound returns because every percentage point lost is a larger percentage needed to regain.
What Is A 20% Drawdown?
A 20% drawdown means your portfolio has fallen 20% below its most recent high. In market lore, a 20% decline from a peak is the informal threshold for a bear market in broad indices. For an individual portfolio, it signals significant pain but is not rare: from 1973–2023, the S&P 500 exceeded 20% drawdowns roughly every seven years on average (see the SEC investor risk bulletin for broad risk context).
The key insight: a 20% drawdown demands a 25% gain to recover, not 20%. This gap widens brutally as losses deepen. When monitoring, I treat a 20% drawdown as a yellow light to re-evaluate position sizing, not merely a statistic to log.
Calculating Rolling Drawdown From A Real Return Series
To move beyond a single peak-trough pair, you need a running peak algorithm. This produces a continuous ‘underwater’ curve showing every day you are below a prior high. Most competing articles stop at maximum drawdown; practitioners live in the rolling view because it captures multiple episodes.
The method I use on daily NAV series: compute cumulative wealth index, set running max, then drawdown = (running_max − value) / running_max. This catches every dip, not just the worst. Below, we break down implementations and a case study.
Excel Implementation: Running Peak And Drawdown Column
In Excel, place portfolio values in column B from row 2 down. In C2 enter =B2. In C3 enter =MAX(C2,B3) and drag down. This creates the running peak. In D3 enter =(C3−B3)/C3 to get drawdown. Negative values indicate you are underwater.
I recommend using =MAX($B$2:B3) in a separate peak column to avoid circular references. When I first built this, I mistakenly referenced the whole column statically, which froze the peak at the first cell and understated drawdowns by 3–5% on volatile assets. Small formula errors cause big risk misreads.
Python Pandas Approach
For larger datasets, Python is cleaner. Sample code:
import pandas as pd
df['peak'] = df['value'].cummax()
df['drawdown'] = (df['peak'] − df['value']) / df['peak']
This leverages cummax() for a running peak. One edge case: if your series has NaN from missing trading days, forward-fill before cummax or you’ll get spurious zero drawdowns. I learned that after a three-day holiday gap showed a ‘false recovery’ in a client report.
Case Study: Five Years Of Monthly Returns
Consider a portfolio starting at $1,000,000 with these month-end values: Jan $1,020,000; Feb $980,000; Mar $1,050,000; Apr $1,010,000; May $960,000; Jun $1,100,000; Jul $1,040,000; Aug $1,150,000; Sep $1,080,000; Oct $990,000; Nov $1,200,000; Dec $1,250,000 (year one). Running peaks climb accordingly.
By March, peak is $1,050,000; April drawdown is (1,050,000−1,010,000)/1,050,000 = 3.81%. May deepens to 8.57%. June recovers peak to $1,100,000, drawdown zero. October drop to $990,000 against $1,150,000 peak is 13.91%. This rolling view shows two distinct drawdown episodes in one year alone.
Extending this to 60 months reveals duration differences: the first episode lasted 2 months, the second 1 month, but a later 2022 episode lasted 7 months. Magnitude alone would hide the persistent underwater period that tests client patience.
The Thing Nobody Tells You About Continuous Drawdown
Most people don’t realize that drawdown depends on observation frequency. A daily drawdown may be −2%, but an intraday spike could have touched −5%. If you only mark to market at close, you understate true risk. For hedge funds, I often compute intraday drawdown using highest high/lowest low ticks, which can be 30–50% deeper than closing figures.
Another blind spot: dividends and cash flows. If you inject capital, the running peak resets artificially. I use a ‘since inception’ return index that reinvests flows, otherwise a deposit creates a fake drawdown trough. This is a classic backtest error that inflates perceived risk-adjusted performance.
Drawdown Duration And The Underwater Curve
Drawdown magnitude is only half the story. Drawdown duration measures how long you stay underwater — from the peak to the day you regain the peak. A 10% drawdown lasting 3 months is very different from one lasting 3 years. The longest S&P 500 drawdown duration was over 13 years from 1929 to 1942 inflation-adjusted, a fact many overlook when judging ‘quick’ recoveries.
Visualizing the underwater equity curve — plotting drawdown as negative percentages over time — instantly reveals clusters of pain. I share a free downloadable template (linked inside our Portfolio Drawdown Calculator page) that auto-generates this chart from pasted returns.
When assessing strategies, I weight duration heavily. A system with −15% max drawdown but 2-year recovery is often worse for client retention than −25% with 4-month recovery. Human behavior, not math, decides real-world outcomes.
Risk-Adjusted Context: Calmar Ratio And Volatility
Raw drawdown numbers mean little without context. The Calmar ratio divides annualized return by maximum drawdown over a period (usually 3 years). A Calmar above 1 is decent; above 3 is excellent. This adjusts for the fact that high-return strategies can absorb deeper drawdowns.
Volatility and drawdown are cousins but not twins. A low-volatility bond fund may have shallow drawdowns; a leveraged ETF can have low daily volatility yet suffer catastrophic drawdowns due to path dependence. For deeper risk decomposition, a portfolio variance calculator helps isolate correlation contributions that drive tail losses.
Trade-off: maximizing Calmar can lead to overly short evaluation windows that hide tail risk. I prefer a 5-year Calmar plus a rolling 1-year max drawdown monitor. No single metric is silver bullet; layering is essential.
Advanced Edge Cases: Leverage, Shorting, And Frequency
Leveraged portfolios break naive drawdown intuition. A 2x leveraged fund can hit a 50% drawdown after only a 25% underlying move. The running peak method still works, but the recovery hurdle doubles. I’ve seen junior analysts celebrate a ‘small’ 10% index dip while the 3x product lost 30% and needed 43% to recover.
Short positions have opposite drawdown signs but same math: if you short at $100 and it rises to $150, your drawdown on the position is 50% (loss relative to peak equity). Most software defaults to long-only drawdown; you must customize for hedging sleeves.
Observation frequency also changes regulatory reporting. UCITS funds report max drawdown on monthly NAV, potentially hiding intraday crashes. I always compute both and disclose the more conservative figure to fiduciaries.
Practical Mistakes I Made Monitoring Drawdowns
Early in my career, I reported only point-to-point maximum drawdown for a fund of funds. The number looked calm at −8%. But the rolling view showed three separate −6% dips that ate cumulative compounding. The omission caused an allocator to overcommit, and when the third dip coincided with a liquidity freeze, we missed a rebalance window.
Another error: using arithmetic returns instead of log or total return. A −10% then +10% leaves you at 99% of original, but naive arithmetic suggests break-even. This mistakenly softens the recovery hurdle. Always use total return indices for drawdown calculations.
Most people don’t realize that drawdown is path-dependent and not sub-additive. Two portfolios each with −20% max drawdown can combine into a portfolio with −30% if losses are concurrent. Correlation, not individual drawdowns, governs portfolio-level risk.
Your Drawdown Monitoring Checklist
Use this framework I developed for a family office to institutionalize drawdown tracking:
- Define peak basis: closing price, adjusted for splits/dividends, or intraday high? Document it.
- Compute running peak daily: Excel MAX drag or Pandas cummax.
- Record drawdown series: not just max, but every point below peak.
- Measure duration: days from peak to recovery; track current underwater days.
- Overlay recovery hurdle: for any loss L, note required gain L/(1−L).
- Risk-adjust: compute Calmar over 3–5 years; compare to volatility.
- Stress test flows: remove capital injections to avoid fake troughs.
Following this checklist turns a vague ‘how to calculate portfolio drawdown’ query into an operational risk process. I revisit it quarterly; it has saved me from at least two ill-timed strategy switches.
Reporting Drawdown To Stakeholders: A Practitioner’s View
When presenting drawdown to non-technical boards, I lead with the underwater chart, not the number. A −22% max drawdown sounds alarming; showing it lasted 4 months in 2020 and recovered within a year provides context. The thing nobody tells you about client reporting: the same statistic can trigger panic or confidence based solely on framing.
I once showed a raw MDD of −35% to a trustee who immediately demanded deleveraging. After overlaying the recovery hurdle (needed +54%) and the Calmar of 1.8, they understood it was within mandate. Transparency with proper context prevents destructive knee-jerk mandates.
Trade-off: too much detail can overwhelm. I cap slides at one drawdown chart, one duration table, and one recovery math callout. Anything more dilutes the message.
Wrapping Up With The Right Tools
If manual spreadsheets feel heavy, our Maximum Drawdown Calculator computes the peak-trough stat instantly, while the portfolio drawdown tool charts the underwater curve. Both embed the running-peak logic described above.
The bottom line: calculating portfolio drawdown is not a one-line percentage. It is a discipline of running peaks, recovery asymmetry, duration awareness, and risk adjustment. Master those, and you’ll interpret portfolio risk far better than someone quoting a single MDD figure.