• Get 30% Off with code "NOV"
  • 40% OFF GROWTH ACCOUNTS
  • ENDS 13 NOVEMBER, 11:59PM EST

Tail Risk and Black Swan Events in 20 Years of ES Data

A lone black swan on a dark lake at dusk, under the headline Tail Risk and Black Swan Events.

TL;DR: Tail risk and black swan events differ, since tail risk is a countable extreme move and a black swan is a shock no model expected. On 20 years of E-mini S&P 500 futures, four sigma days came about 123 times more often than a normal model predicts. Those days clustered into 11 episodes, sent the VIX higher every time and pulled correlations together. Size every position for the worst day on record, not the average one.

The worst day in twenty years of E-mini S&P 500 futures was a 10.38% drop. Apply that to an illustrative index level of 6,000 and it is 622 points, roughly $31,000 on one E-mini contract in a single session. Most traders size a position around an ordinary day, because ordinary days are what they see. Tail risk is the cost of the day that is not ordinary, and it does not wait until you are ready for it.

This guide separates the two ideas people blend together, tail risk and black swans. Then it measures how often extreme days actually hit ES, how they cluster, what they do to other markets, and which decisions a trader can still control before the next one.

What Tail Risk and Black Swan Events Mean

Tail risk is the chance of a return far from the average, out in the thin "tails" of a return distribution. The usual working definition is a move more than three standard deviations from the mean. The cutoff is a convention, not a law, so this guide reports two through six standard deviations side by side.

A black swan is a narrower claim. Nassim Nicholas Taleb, who popularized the term in finance with Fooled by Randomness in 2001 and extended it in The Black Swan in 2007, gives it three attributes:

  • Rarity. The event is an outlier that nothing in the past convincingly pointed to.
  • Extreme impact. It carries consequences far beyond an ordinary bad day.
  • Retrospective predictability. Afterward, people build explanations that make it look obvious.

Taleb summarizes the triplet as rarity, extreme impact, and retrospective (though not prospective) predictability.

Where the Black Swan Phrase Comes From

The phrase is older than finance. The Roman poet Juvenal wrote in his Satire VI of "a rare bird in the lands and very much like a black swan," and for centuries Europeans used "black swan" to mean something impossible. In 1697 Dutch explorers led by Willem de Vlamingh saw black swans in Western Australia, and the impossible thing turned out to be simply unobserved.

That origin is the whole lesson in miniature. Absence of evidence in your sample is not evidence the event cannot happen.

Why One Idea Is Countable and the Other Is Not

Tail risk is a property of a price series. You can count how many sessions moved more than four standard deviations, and the count is a fact about the market.

Whether an event was a black swan depends on what was in your model beforehand, which is a fact about you. Taleb's own example is the turkey fed every day until the butcher arrives: a shock for the turkey, a plan for the butcher. A trader cannot forecast the next black swan, but a trader can measure tail risk and size for it. That is where the rest of this guide spends its time.

Black Swans Compared With Grey Swans and White Swans

Black, grey and white swans differ in how foreseeable they are: black swans are only explained afterward, grey swans are conceivable but unlikely, and white swans are certain with unknown timing.

Black swans, grey swans and white swans differ mainly in how foreseeable they were. The labels matter because they change what preparation is realistic.

Swan type What it means Could it be anticipated Trading example
Black swan An outlier nothing in the past pointed to, with extreme impact No, only explained afterward A shock with no precedent in anyone's risk model
Grey swan Highly unlikely but conceivable, often without historical precedent Yes, as a known possibility with a low assigned probability A known risk, such as a pandemic, that is easy to discuss and hard to time
White swan Great impact and high certainty that it will eventually happen Yes, timing is the unknown A recession or a bear market at some point in a cycle

The grey swan label shows up most often in debates about 2020. Some researchers called the pandemic a grey swan because it was conceivable, and Taleb himself called it a white swan, an event with great impact that was bound to happen eventually.

For a trader, the practical takeaway is that most market crashes behave like grey or white swans. The specific trigger surprises you, but the possibility of a large down day was always in the price history. That is good news, because it means the size of the damage can be planned for even when the cause cannot.

Are Black Swan Events Good or Bad?

A black swan can be positive. Taleb's attributes require extreme impact, not a negative one. A short position gains in a crash, and the largest single up day in the ES sample examined here (a 14.11% gain on October 13, 2008) arrived in the middle of the worst crisis in the series. Negative swans get the attention because an account has a floor at zero, so a large loss can end it while a large gain does not.

How Often Tail Events Hit 20 Years of ES Data

To put a number on tail risk, this study uses daily bars on the front-month continuous E-mini S&P 500 futures contract from September 18, 2006 to September 16, 2026, 5,032 sessions in all. The continuous series switches contracts every quarter, and on expiry days it prints the final settlement value as the close. So the change into and the change out of each of the 80 quarterly expiry sessions is left out. That leaves 4,872 clean daily changes with a mean of 0.04% and a standard deviation of 1.23%.

A normal distribution built from that mean and standard deviation predicts how many days should land beyond each threshold. The table compares that prediction with what ES actually did.

Threshold Move from the mean ES days observed Normal model expects Observed vs expected Model says one every
2 sigma 2.46% 234 221.7 1.06 times 22 sessions
3 sigma 3.69% 77 13.2 5.85 times 370 sessions
4 sigma 4.92% 38 0.31 123.1 times 15,787 sessions
5 sigma 6.15% 18 0.0028 6,444 times 1,744,278 sessions
6 sigma 7.37% 11 under 0.0001 over 1 million times 506,797,346 sessions

At two standard deviations the normal model is close. At three it undercounts by a factor of about six. At four, where ES produced 38 days, the model expects a day like that once every 15,787 sessions, which is about 62.6 years of trading. ES delivered one every 128 sessions, roughly twice a year.

That is what "fat tails" means in practice. The middle of the distribution looks normal enough to fool you, and the edges are where the model breaks.

Testing Whether One Crisis Explains the Fat Tails

A fair objection is that 2008 alone could create this pattern. Removing every session from September 2008 through January 2009 leaves 4,771 changes and a standard deviation of 1.10%. Four sigma days still arrive 82.7 times more often than the model predicts. Removing the February to April 2020 crash as well still leaves 66.9 times.

The sample's standard deviation is itself inflated by the extreme days it is trying to measure. A robust estimate built from the median absolute deviation puts daily volatility at 0.73%, and against that yardstick four sigma days arrive 486.1 times more often than a normal model expects. Change the method and the size of the gap moves, but the direction never does.

Fat Tails in the ES Return Distribution

Kurtosis measures how heavy the tails are. A normal distribution has a kurtosis of 3. Clean ES daily changes come in at 17.82.

The tails also carry an outsized share of total risk. In the clean ES sample, the largest 1% of days (48 sessions out of 4,872) account for 30.66% of all daily variance in the sample. The ten largest days alone account for 14.50%. Most of the risk in twenty years lived in a few dozen sessions.

Interestingly, the ES tails are roughly balanced. At four sigma there were 23 down days and 15 up days, and the skew of the clean series is close to zero. The violent up days tend to arrive inside the same crises as the violent down days.

Why Extreme Market Days Arrive in Clusters

Extreme market days arrive in clusters separated by long quiet stretches, so after one extreme day another often follows soon.

If extreme days were independent, the 38 four sigma days in ES would be spread out evenly, about 128 sessions apart on average. They were not.

Related lesson: Correlation risk

Grouping any four sigma days within 10 sessions of each other into one episode, the 38 days fell into just 11 episodes:

  • The largest episode, starting September 15, 2008, contained 18 four sigma days.
  • The 2020 episode, starting February 27, 2020, contained 8.
  • Four episodes had more than one day, and 7 were single sessions.
  • The median gap between consecutive four sigma days was 3 sessions, and 27 of the 37 gaps (72.97%) were 10 sessions or less.

The conditional numbers are the ones that matter for a plan. In the ES study, after a four sigma day, another one followed within 10 sessions 27 times out of 38, or 71.05%. If those days arrived independently, the chance of seeing one in any 10-session window would be about 7.53%.

At three sigma the same thing happens: 77 days, 22 episodes, a median gap of 3 sessions and 55 of 76 gaps (72.37%) inside 10 sessions. The first extreme day is usually the start of a sequence, not the end of one. That changes the decision after a big loss. The urge is to win it back on the next session, but the next session is exactly when another extreme day is most likely.

Recovery Time After a Four Sigma Drop

There were 23 clean four sigma down days. For each one, the study counts how many sessions it took for ES to close back above the close before the drop.

  • In this ES study, the median wait was 6 sessions.
  • In the same ES sample, 10 of the 23 (43.48%) recovered within 5 sessions.
  • In the same ES sample, 3 of the 23 (13.04%) took more than 100 sessions.
  • The longest, after September 29, 2008, took 530 sessions.

A plan that assumes the storm passes in a week is right more often than not. The trouble is that you cannot tell in advance which drop is the 6-session kind and which is the 530-session kind.

Why Forecasting Models Miss Tail Risk

Most standard risk models start from the assumption that returns follow a normal distribution with stable volatility. Both assumptions fail at the same moment, which is why forecasting models tend to be most accurate when you need them least.

Three problems compound each other:

  • The distribution is wrong in the tails. In 20 years of ES data, a normal model is close at two standard deviations and off by a factor of more than 100 at four.
  • Volatility is not stable. ES's 14-day average true range ran from 0.72% of price at its 5th percentile to 3.08% at its 95th, a factor of about 4.3. A model calibrated in a calm year is calibrated to the wrong market when a crisis starts.
  • Correlations are not stable. Assets that move independently in calm markets tend to move together in a crisis, which breaks the diversification math many models rely on.

Value at Risk Compared With the Real ES Tail

Value at risk (VaR) answers a narrow question: how much could I lose on a normal day at a given confidence level? A normal-distribution VaR fitted to the clean ES sample looks reasonable at 95% and drifts further from reality the deeper into the tail you go.

Related lesson: Value at risk

Confidence level Normal VaR (daily loss) Actual ES loss at that level Actual vs model Average loss beyond that level
95% 1.98% 1.80% 0.91 times 3.00%
99% 2.82% 3.57% 1.27 times 5.24%
99.9% 3.75% 7.91% 2.11 times 9.39%

The last column is the more honest number. VaR tells you where the bad days start, not how bad they get. The average of the worst 1% of ES days was 5.24%, nearly double the 99% VaR figure. Bank regulators made the same point: the Basel Committee's revised market risk framework shifted from value at risk to an expected shortfall measure of risk under stress, which looks at the average loss beyond the cutoff instead of the cutoff itself.

The LTCM Lesson for Model Risk

Long-Term Capital Management is the classic warning about trusting a model's tails. According to the Federal Reserve's history of the episode, the hedge fund was known for sophisticated mathematical models, held about $30 in debt for every $1 of capital at the end of 1997, and bet that price spreads would converge. In 1998 they diverged instead. In September 1998, 14 banks and brokerage firms put $3.6 billion into the fund to prevent its collapse.

The models were not stupid. They were built on history that did not contain the event that broke them, and the leverage turned a model error into an existential one.

Historical Black Swan Events in the Stock Market

The events most often called black swans in markets share one pattern: a known category of risk arriving through a trigger few people priced. Here are the major ones, with the measured ES moves for those inside the 2006 to 2026 sample.

  • October 19, 1987 (Black Monday). The Dow Jones Industrial Average fell 22.6% in one session, still the largest one-day percentage decline in its history, according to the Federal Reserve's history of the crash. Circuit breakers that halt trading after large declines were introduced afterward.
  • 1998, Long-Term Capital Management. A highly leveraged fund's model failure ended in a $3.6 billion rescue funded by its creditors and arranged with the New York Fed's help.
  • 2008 financial crisis. Lehman Brothers filed for bankruptcy on September 15, 2008. ES posted its first four sigma day of the crisis that session, the start of an 18-day cluster of four sigma moves.
  • August 2011 downgrade. Standard & Poor's lowered the long-term US credit rating to AA+ from AAA on August 5, 2011. ES fell 7.22% on August 8, the next session.
  • August 24, 2015. The SEC's staff research note on that day describes unusual price volatility in equity and equity-related futures markets around the 9:30 AM ET open. ES fell 5.08% that session.
  • February 5, 2018. ES fell 5.40% and the VIX index more than doubled, up 115.6% in a day.
  • March 2020, COVID-19. WHO characterized COVID-19 as a pandemic on March 11, 2020. ES posted its two worst sessions in the sample on March 12 and March 16, 2020, and three of its seven worst between March 9 and March 16.
  • April 2025 tariffs. A White House fact sheet dated April 2, 2025 announced a 10% baseline tariff on all countries plus higher reciprocal tariffs. ES fell 4.89% on April 3 and 5.94% on April 4.

The Ten Worst ES Sessions Since 2006

Date ES daily change Index points lost Context
March 16, 2020 -10.38% 279.75 COVID-19 crash
March 12, 2020 -9.90% 271.25 Day after WHO pandemic assessment
October 15, 2008 -9.88% 99.00 2008 financial crisis
December 1, 2008 -8.88% 79.50 2008 financial crisis
November 20, 2008 -7.91% 64.25 2008 financial crisis
September 29, 2008 -7.88% 95.75 Two weeks after Lehman's bankruptcy
March 9, 2020 -7.30% 216.25 COVID-19 crash
August 8, 2011 -7.22% 86.50 First session after the S&P downgrade
October 9, 2008 -6.98% 68.50 2008 financial crisis
November 19, 2008 -6.23% 54.00 2008 financial crisis

Eight of the ten worst ES days since 2006 came from just two episodes, 2008 and 2020. Extreme days cluster, and this list shows it.

Seven ES Declines of 10% or More

Measured from a closing high to the lowest close before a new high, ES fell 10% or more seven times in the sample.

Peak close date Decline to the low Sessions peak to low Sessions peak to new high
October 9, 2007 57.11% 355 1,384
May 21, 2015 14.26% 183 286
January 26, 2018 10.42% 44 146
September 21, 2018 20.43% 64 150
February 19, 2020 34.45% 23 129
January 3, 2022 25.02% 195 493
February 19, 2025 18.54% 34 88

The 2020 decline was the fastest: 34.45% in 23 sessions. The 2007 to 2009 decline was the deepest and took 1,384 sessions, more than five years, to reach a new closing high.

What Black Swan Events Do to Volatility and Other Markets

A black swan event changes more than the price of one index. Volatility jumps, related markets move with it, and the usual safe havens behave less reliably than their reputations suggest.

Related lesson: What is volatility

The table shows what four other markets did on the 23 clean four sigma down days in ES, using the same daily data source.

Market on ES four sigma down days Median daily change Days it rose out of 23
VIX index +29.85% 23
E-mini Nasdaq-100 futures (NQ) -5.28% 0
10-year Treasury note futures (ZN) +0.44% 17
Gold futures (GC) 0.00% 11

The VIX rose on every single one of those days. On March 16, 2020 it closed at 82.69. Nasdaq futures fell on every one, by a median of more than 5%. Ten-year note futures rose on most of them, the classic flight to quality. Gold was a coin flip, rising on 11 of the 23 days and falling 3.17% on March 12, 2020 and 2.76% on April 4, 2025.

How Concentrated the Damage Is Inside a Crash

A crash is not a steady slide. Inside each decline, the study ranks the down days by points lost and asks what share of all the down-day points came from the worst few sessions.

  • February to March 2020 in ES: the single worst day was 17.06% of all down-day points, the worst five days 63.78%, and the worst ten 91.55%.
  • February to April 2025 in ES: worst day 17.76%, worst five 54.01%, worst ten 79.13%.
  • 2007 to 2009 in ES: worst day 2.78%, worst five 11.42%, worst ten 19.31%. A slow bear market spreads the damage out.

Fast crashes concentrate the loss into a handful of sessions. If you are on the wrong side for those few days, the rest of the episode barely matters.

Why Diversification and Portfolio Theory Break in a Crash

Modern portfolio theory builds portfolios from expected returns, volatilities and correlations, and it assumes those correlations hold steady. Diversification works when assets that would hurt you at the same time actually move independently. Tail events are when that assumption is tested hardest.

Related lesson: Diversification

The study compares the correlation of daily changes with ES on calm days (ES within one standard deviation of its mean) against stress days (ES beyond three standard deviations).

Pair Correlation on calm days Correlation on stress days
ES and NQ 0.81 0.98
ES and 10-year note futures -0.13 -0.47
ES and gold futures 0.04 0.06
ES and VIX -0.69 -0.76

Two lessons come out of that table:

  • Same-category positions merge into one bet. Long ES and long NQ already look similar on a calm day. On a stress day they are nearly the same position, at 0.98, so two contracts in different index products give you no tail diversification at all.
  • Some offsets strengthen, but none are guaranteed. Ten-year note futures became a stronger offset on stress days, and gold showed almost no relationship with ES either way. A hedge that works on average can still fail on the specific day you needed it.

A portfolio is only as diversified as its correlations on the worst day, not on an average one.

Managing Tail Risk With Stops, Size and Stress Tests

You cannot predict the next black swan, so tail risk management is about surviving the measured worst case before the shock arrives. Three tools do most of the work: stop placement, position size and stress testing.

How Stops Behave on Tail Days

A stop order becomes a market order when its price trades. If a session opens beyond the stop, the fill comes at the open, not at the stop price.

The study tests stops placed below the prior close at one, two and three times the 14-day average true range (ATR), across 4,857 clean sessions:

  • In the ES study, a one ATR stop was touched on 684 sessions (14.08%). On 5 of those, the session opened beyond the stop, so the fill would have been worse than planned. The worst such open was 0.65 of a stop width past the level, on June 29, 2015.
  • In the same ES study, a two ATR stop was touched on 99 sessions (2.04%), and none of those opened beyond the stop in this sample.
  • In the same ES study, a three ATR stop was touched on 17 sessions (0.35%).

The absence of a two ATR gap-through is a fact about this ES sample, not a promise. In the same sample, 189 sessions (3.89%) had a high-to-low range wider than a two ATR stop.

On the 23 clean four sigma down days, the distance from the prior close to the session low was a median 1.12 times a two ATR stop, and 13 of the 23 went at least one full stop width. A wider stop does not make you safe on a tail day. It turns an unbounded loss into a known one, as long as the size was set with that wider stop in mind. Widening the stop without cutting size just makes the same position more expensive to be wrong.

Related lesson: Stop-loss orders and placement

Worked Example on the Worst ES Session

The worst clean session in the sample was March 16, 2020, when ES fell 10.38% from the prior close. Imagine a trader held one long E-mini contract into that session with a stop two ATR below the prior close. ES is $50 per index point.

  1. The prior close was 2,696.00 and the 14-day ATR was 143.90 points, so a two ATR stop sat 287.79 points lower, at 2,408.21. That stop risks $14,389.71 on one contract.
  2. The session opened at 2,673.75, only 22.25 points lower, so the open was above the stop.
  3. The low was 2,360.25, which is 335.75 points below the prior close, or 1.17 times the stop width. The stop level traded during the session.
  4. If the stop filled at its level, the loss was about $14,389.71. Holding to the close instead would have lost 279.75 points, or $13,987.50, because ES closed 8.04 points above the stop level.

A daily bar cannot show the path inside the session. On a day like that, a fast market or a trading pause can push a fill past the stop price, so treat the stop figure as the best case, not a ceiling. The same arithmetic on the Micro E-mini at $5 per point gives a planned loss of $1,438.97, one tenth of the E-mini figure.

The headline is not that the stop "cost" $402.21 more than holding. It is that nobody knew at the open that the low would stop at 2,360.25 rather than somewhere far lower.

Sizing for the Worst Day on Record

Index points scale with the price level, so the sizing table uses the ES moves measured as a percentage of price and applies them to an illustrative index level of 6,000.

Related lesson: Volatility-based position sizing

Daily move scenario (illustrative ES at 6,000) Size of move Points One ES contract One MES contract One ES as a share of $50,000
Median 14-day ATR 1.28% 76.52 $3,825.80 $382.58 7.65%
95th percentile ATR 3.08% 184.61 $9,230.60 $923.06 18.46%
Four sigma day 4.92% 294.97 $14,748.42 $1,474.84 29.50%
Worst day in the sample 10.38% 622.59 $31,129.45 $3,112.95 62.26%

At that illustrative level, a two ATR stop at the median ATR is 153.03 points, or $7,651.59 per E-mini and $765.16 per Micro. Staying inside a 1% risk limit at that stop needs about $765,159 of equity per E-mini contract and about $76,516 per Micro.

That is not an argument against trading the E-mini. It is the reason the Micro exists, and the reason a size chosen in a quiet month can be far too big when volatility triples.

Stress Testing a Trading Plan

A stress test asks one question: what happens to this account if the worst measured day happens tomorrow? A simple version takes four steps:

  1. Write the worst session for your instrument in points and dollars per contract at today's price.
  2. Multiply by every contract you could hold at once, across all correlated products, because on a stress day ES and NQ move almost as one.
  3. Compare the result with your daily loss limit and your maximum drawdown.
  4. If the worst day ends the account, cut size until it does not, or accept that the plan depends on that day never coming.

Hedging Tail Risk and Black Swan Funds

Hedging tail risk means holding something that gains, or loses much less, when the main position takes a tail hit. Investors use a few common routes:

  • Put options. Out-of-the-money puts on an index pay off in a crash and cost a premium every month nothing happens.
  • Volatility exposure. Long volatility positions tend to gain when the VIX spikes, as it did on all 23 four sigma ES down days in the sample.
  • Government bonds. In the 20-year ES study, ten-year note futures rose on 17 of the 23 four sigma ES down days, a partial offset rather than a guaranteed one.
  • Tail-risk funds. So-called black swan funds hold mostly cash or safe assets plus a sleeve of crash protection, aiming to lose small amounts in normal years and gain sharply in a crash.

Related lesson: Hedging risk

Pros and Cons of Positioning for Black Swans

Approach What it can do What it costs
Buying crash protection Caps the loss on a portfolio in a sharp decline A steady premium drag in calm years, which can be most years
Holding bonds as an offset Often rises when equities crash Can fall alongside stocks when inflation or rates drive the selloff
Black swan funds Large gains in rare crashes Long stretches of small losses and the discipline to keep paying
Smaller position size Cuts every outcome, including the tail Lower returns in normal markets

The hard part of any tail hedge is behavioural. Paying for protection that expires worthless for years feels like a mistake right up until the day it is the only thing that worked. For most active futures traders, position size and stop discipline are the hedge they can actually afford to keep. They cost nothing in a calm month, and they are already in place on the day it matters.

Psychological Biases Around Black Swan Events

Black swan events are hard to manage partly because of how people think about them before and after. Three biases do most of the damage.

Related lesson: Recency and hindsight bias

  • Hindsight bias. After the event, the explanation feels obvious, so traders conclude they should have seen it and try to predict the next one instead of sizing for it. This is the third attribute in Taleb's definition, built into human nature.
  • Normalcy bias. In a calm market, a trader assumes tomorrow will look like the last few months. Size creeps up because nothing bad has happened lately, just as the ATR study shows volatility can swing by a factor of about four between quiet and stressed periods.
  • Recency bias after a shock. Right after a crash, the opposite happens: every dip feels like the next collapse, and traders exit at the worst moment.

The ES data shows why that last one is costly. 13 of the 20 best days in the sample fell within 10 sessions of one of the 20 worst days. The largest gain, 14.11% on October 13, 2008, came in the same crisis as the largest losses. Selling in a panic often means missing the rebound days that arrive in the same cluster.

The fix is the same for all three biases. Decide the rules before the shock, when you are calm, so you are not inventing them in the middle of one.

A Tail Risk Checklist for Traders

A tail risk checklist: write down your worst session, recalculate the stop with today's ATR, use Micros when one contract is too big, count correlated products as one and expect another extreme day.

A tail risk plan works only if it is written down before the bad day. This checklist turns the ES measurements into rules a trader can apply:

  • Write the worst session for your instrument, in points and dollars per contract at today's price, at the top of your trading plan.
  • Recalculate your stop in dollars using today's ATR, not the ATR from the week you wrote the plan.
  • Check whether one contract already breaks your risk limit before adding a second, and use Micros when it does.
  • Count correlated products as one position. In the 20-year ES study, NQ moved with a correlation of 0.98 to ES on stress days.
  • After an extreme day, assume another may follow within 10 sessions. In the 20-year ES study, 71.05% of four sigma days were followed by another within that window.
  • Decide in advance what you will do after a third stop-out in a week, while the sequence may still be running.
  • Never use a daily loss limit as a stop. Your stop should be placed long before you get there.

Related lesson: Maximum drawdown

On our accounts, every position must be closed by 4:45 PM ET, and no overnight or weekend holds are allowed. Because positions cannot be carried between sessions, our traders are not exposed to an overnight gap on an open position, but they still face the full intraday range of a tail day. Our Growth evaluations carry a daily loss limit that puts the account into liquidation-only until the next trading day, and our trailing drawdown is enforced in real time on net liquidation value. We do not compensate losses during CME exchange halts, and all of our funded accounts are simulated.

Tail Risk and Black Swan Events FAQ

What Is a Black Swan Event in Risk Management?

In risk management, a black swan is an event outside the model, with extreme impact, that looks predictable only afterward. Risk managers cannot forecast it, so they plan for the damage instead. That means stress tests, position limits and hedges sized for losses far beyond normal days.

Is Tail Risk the Same as a Black Swan?

No. Tail risk is the measurable chance of an extreme move, and it can be counted on any price history. A black swan is a specific surprise that nobody's model included. Every black swan is a tail event, but most tail events are not black swans, because the possibility was already visible in the data.

How Often Do Tail Events Happen in Futures Markets?

On 20 years of E-mini S&P 500 futures, days beyond four standard deviations arrived about once every 128 sessions, roughly twice a year. A normal model predicts one every 15,787 sessions. Those days also clustered, so several often arrived within a few weeks of each other.

What Is the Most Recent Black Swan Event in the Stock Market?

The most recent four sigma episode in 20 years of ES data came in April 2025, after a tariff announcement, with drops of 4.89% and 5.94% on back-to-back sessions. There is no official list of black swans, and whether that episode qualifies depends on whose model you ask. Many would call it a grey swan, since tariffs were a known possibility.

Was the 2020 Crash a Black Swan Event?

The 2020 crash was a black swan for investors with no pandemic in their models and a grey or white swan for those who saw pandemics as a known possibility. The market damage was familiar in shape. ES fell 34.45% in 23 sessions, with 91.55% of the down-day points in the worst ten sessions.

Risk Disclosure for Tail Risk Education

Trading futures involves substantial risk of loss and is not suitable for everyone. Leverage means losses can exceed what you expected, and tail events can move prices past stop orders. The figures in this guide are historical measurements and illustrative arithmetic, not a forecast or a promise of any result. Evaluation and funded accounts are simulated, and simulated results do not reflect live market fills. This is educational content, not investment advice.