Free Value at Risk (VaR) Calculator
How bad is a typical bad day for your portfolio — and how bad is the worst day? Upload a CSV and get VaR at 95% and 99%, plus Conditional VaR (Expected Shortfall) for true tail-risk estimation. Works for any stock, ETF, or crypto portfolio.
What is Value at Risk (VaR)?
VaR estimates the maximum loss your portfolio is likely to experience over a given period at a given confidence level. “95% one-day VaR of $5,000” means: there's a 5% chance you lose more than $5,000 in a single trading day. Foliolytic computes both historical VaR (empirical worst days) and parametric VaR (assuming normal returns).
95% VaR · historical + parametric · daily / weekly / monthly horizonsWant the deeper math? Learn the math behind Value at Risk →
What Is Value at Risk?
Value at Risk (VaR) answers the most basic downside question in portfolio management: how much could I lose on a normal-but-bad day? A 1-day VaR of $10,000 at 95% confidence means: on 95 out of 100 days, you lose less than $10,000. On roughly 5 days out of 100, you lose more.
VaR entered mainstream practice through J.P. Morgan's RiskMetrics framework, published openly in 1994, which standardized confidence-level loss estimates across the industry. VaR is the metric banks report to regulators. It's the metric hedge funds cite in risk disclosures. It's imperfect — it doesn't tell you how much worse the bad 5% gets — but as a single-number summary of downside, it's the most widely used measure in finance.
VaR95% = 5th-percentile of daily return distribution × Portfolio Value
The loss at the 5th percentile — bigger losses happen 5% of the time
VaR's Blind Spot: Tail Risk
VaR tells you the threshold. It doesn't tell you what happens past the threshold. If your VaR 95% is 4%, that just means the 5th-worst day is a 4% loss. The 1st-worst day could be a 4.1% loss — or a 15% loss. VaR makes no distinction.
Conditional VaR (CVaR), also called Expected Shortfall, fixes this. It's the average loss on the days worse than VaR. If your VaR 95% is 4% but your CVaR 95% is 8%, that tells you the bad 5% of days average 8% losses — a fundamentally different risk picture than if both numbers were close.
CVaR95% = E[loss | loss > VaR95%]
Expected (average) loss conditional on exceeding the VaR threshold
CVaR is the regulatory standard under Basel III precisely because VaR alone underestimates how bad a crash can be. For fat-tailed portfolios (crypto, options, emerging markets), the CVaR/VaR gap is often 2x or more.
Typical VaR Values by Asset Class
| Portfolio | 1-day VaR 95% | 1-day VaR 99% | 1-day CVaR 95% |
|---|---|---|---|
| US Treasury Bonds (AGG) | 0.3 – 0.5% | 0.5 – 0.8% | 0.4 – 0.7% |
| 60/40 Stocks/Bonds | 0.8 – 1.2% | 1.3 – 1.8% | 1.2 – 1.7% |
| S&P 500 (SPY) | 1.3 – 2.0% | 2.0 – 3.0% | 2.0 – 3.5% |
| Nasdaq 100 (QQQ) | 1.7 – 2.5% | 2.5 – 3.8% | 2.5 – 4.2% |
| Emerging Markets (EEM) | 1.8 – 2.8% | 2.8 – 4.5% | 2.8 – 5.0% |
| Bitcoin | 5.0 – 7.0% | 8.0 – 12.0% | 8.0 – 14.0% |
| Altcoin portfolio | 8.0 – 15.0% | 13.0 – 25.0% | 13.0 – 30.0% |
Ranges reflect long-term historical data (2010-2026) and vary with market regime. VaR expands dramatically in crises — 2020's March crash briefly pushed S&P 500 daily VaR above 6%. This is the 'volatility clustering' phenomenon that static VaR can miss.
What Is the Actual 1-Day VaR of Major Assets Right Now?
As of July 2026, the S&P 500's 1-day historical VaR at 95% confidence is 1.69% — with a CVaR of 2.48% and a worst single day of −5.85% over the past five years (April 4, 2025). Bitcoin's 95% VaR of 4.13% runs about 2.5× the S&P's, and a classic 60/40 stock/bond blend cuts the S&P's VaR by roughly a third.
| Asset | VaR 95% | VaR 99% | CVaR 95% | Worst Day (5Y) |
|---|---|---|---|---|
| SPY (S&P 500) | 1.69% | 2.95% | 2.48% | −5.85% · Apr 4, 2025 |
| QQQ (Nasdaq-100) | 2.37% | 3.91% | 3.29% | −6.21% · Apr 4, 2025 |
| VT (Total World) | 1.60% | 2.73% | 2.29% | −5.96% · Apr 4, 2025 |
| IWM (Russell 2000) | 2.24% | 3.50% | 3.02% | −6.42% · Apr 3, 2025 |
| AGG (US Bonds) | 0.63% | 0.98% | 0.84% | −1.64% · Jun 13, 2022 |
| GLD (Gold) | 1.80% | 3.12% | 2.73% | −10.27% · Feb 2, 2026 |
| BTC (Bitcoin) | 4.13% | 7.58% | 6.35% | −15.97% · Jun 13, 2022 |
| ETH (Ethereum) | 5.73% | 10.02% | 8.37% | −17.46% · Nov 9, 2022 |
| 60/40 (SPY/AGG blend) | 1.10% | 1.89% | 1.58% | −3.48% · Apr 4, 2025 |
Historical (empirical-percentile) 1-day VaR computed by Foliolytic from daily total returns in its own price database, 5-year window ending July 22, 2026 (roughly 1,290 trading days per equity asset; Bitcoin and Ethereum use all calendar days). Same method the calculator applies to your portfolio — see methodology. Figures refresh when this page is next updated.
Two lessons hide in this table. Gold's worst day (−10.27%) was 3.3× its own 99% VaR — the loss beyond the threshold is exactly what VaR alone conceals and CVaR partially captures. And five of the nine assets logged their worst day within the same 24 hours (April 3–4, 2025): on a genuine crash day, equity diversification barely helps, which is why the 60/40 blend — whose bond sleeve had its worst day in a different year entirely — shows the shallowest equity-linked drawdown.
See Your Portfolio's True Tail Risk
Upload a CSV. Get VaR 95%, VaR 99%, CVaR, and 70+ other risk metrics. Free, no signup.
Open the AnalyzerFrequently Asked Questions
What is Value at Risk (VaR)?
Value at Risk is the maximum loss your portfolio could experience over a specified time horizon at a given confidence level. A 1-day VaR of $10,000 at 95% confidence means: on 95% of days, you should lose less than $10,000. On roughly 1 out of every 20 days, you should expect to lose more than that. VaR is the most common single-number measure of downside risk used by banks, hedge funds, and regulators.
What is the difference between VaR 95% and VaR 99%?
VaR at 95% catches the "typical bad day" — the 5th percentile of the loss distribution. VaR at 99% catches the "rare bad day" — the 1st percentile. VaR 99% is always larger in magnitude than VaR 95% because you're looking further into the tail of the distribution. Regulators typically require banks to report VaR 99% because it captures more severe scenarios. Retail investors often find VaR 95% more useful for day-to-day expectations.
What is CVaR (Conditional VaR / Expected Shortfall)?
CVaR answers a different question: "when the worst happens (beyond VaR), how bad is it on average?" CVaR 95% is the average loss on the worst 5% of days. It's always worse than VaR 95% and is a better measure of tail risk because it accounts for the severity of the extreme losses, not just their frequency. Basel regulations shifted toward CVaR precisely because VaR alone underestimates fat-tail risk.
Historical VaR vs parametric VaR — which does Foliolytic use?
Foliolytic reports both. Historical VaR computes the percentile directly from your actual return distribution — it makes no assumption about normality and correctly reflects fat tails. Parametric VaR assumes returns are normally distributed and uses the mean and standard deviation. For equity portfolios, the two methods give similar answers; for crypto and leveraged portfolios, historical VaR is usually meaningfully larger (and more accurate) because return distributions have fatter tails than a normal curve predicts.
What is the 1-day VaR of the S&P 500 right now?
As of July 2026, SPY's 1-day historical VaR is 1.69% at 95% confidence and 2.95% at 99% confidence, with a CVaR (expected shortfall) of 2.48% — computed by Foliolytic from five years of daily total returns ending July 22, 2026. The worst single day in that window was −5.85% on April 4, 2025. On a $100,000 S&P 500 position, that VaR means roughly 1 trading day in 20 loses more than $1,660.
Why is crypto VaR so much larger than equity VaR?
Crypto daily returns are typically 4-6x more volatile than broad-market equities. Bitcoin's historical daily VaR 95% runs around 5-7%, versus roughly 1.5-2% for the S&P 500. Altcoins are worse still — 10-15% daily VaR 95% is not uncommon. This is why crypto allocations should be sized with VaR in mind: a 10% crypto allocation may contribute 30-40% of portfolio-level VaR.