What it means
Statisticians describe the shape of a distribution using a measure called kurtosis. A normal distribution, the familiar bell curve, has a kurtosis of 3.
When the kurtosis is above 3, the distribution is called leptokurtic, and when it is below 3 it is platykurtic (flatter, with thinner tails). Many financial returns are leptokurtic.
Day to day, markets often move very little, which creates a tall, narrow peak in the middle of the distribution. But every so often, a crash or a surge occurs, giving the distribution thick tails, which are the extreme ends on both sides.
The business consequence is that standard risk tools can underestimate danger. If a risk model assumes a normal distribution, it treats a very large loss as almost impossible.
In a leptokurtic world, such losses are rare but far more likely than the model says, which is why they are sometimes called fat-tail events. Analysts often report excess kurtosis, which is simply kurtosis minus 3.
A positive excess kurtosis signals a leptokurtic shape, zero signals a normal one, and negative signals platykurtic. Some risk managers combine this with skewness, a measure of whether the large moves are tilted to the downside or upside.
Practically, the finding calls for caution. Stress tests, scenario analysis and holding extra capital as a buffer all help a business survive events that a normal model would label extremely unlikely.
It also reminds investors that average volatility can hide the occasional large shock.
In practice
Real-world examples.
Example
A hedge fund risk analyst reviews five years of daily returns on a share index. Most days the index moves less than 1%, but a handful of days show falls of more than 5%. The analyst finds a kurtosis of 6, and reports that the fund's value-at-risk model, which assumed a normal distribution, understated the chance of large losses.
Example
A corporate treasurer examines monthly changes in a foreign exchange rate that affects the company's imports. The data show occasional sudden jumps after policy announcements. She decides to hold a larger cash buffer than a simple bell curve would suggest.
Example
An insurance company studies the size of claims on a product line. Most claims are small, but a few are very large, producing a leptokurtic pattern. The actuary prices the policies with a margin that reflects the heavier tail.
Formula
Calculation
Kurtosis = (Average of the fourth powers of deviations from the mean) / (Variance squared)
Excess kurtosis = Kurtosis - 3
Worked example: a fund has these ten monthly returns, in percentage points: -5, -1, -1, 0, 0, 0, 0, 1, 1, 5. The mean is (-5 - 1 - 1 + 0 + 0 + 0 + 0 + 1 + 1 + 5) / 10 = 0.
Squared deviations: 25, 1, 1, 0, 0, 0, 0, 1, 1, 25. Their sum is 54, so the variance is 54 / 10 = 5.4.
Fourth powers: 625, 1, 1, 0, 0, 0, 0, 1, 1, 625. Their sum is 1,254, so the average is 1,254 / 10 = 125.4.
Variance squared is 5.4 x 5.4 = 29.16. Kurtosis = 125.4 / 29.16 = 4.30, and excess kurtosis = 4.30 - 3 = 1.30. Because the figure is above 3, the returns are leptokurtic.Case study
Seen in the real world.
Marlowe Asset Partners is a fictional investment firm that used a standard model assuming normally distributed returns to set its risk limits. The model said a loss of more than 8% in a month would occur roughly once in several centuries.
A new analyst measured the real return history and found a kurtosis of 5.8. In the previous ten years the fund had suffered such a loss twice, which the old model said was practically impossible.
The firm replaced its model with one that allowed fat tails and added regular stress tests. Position limits were reduced by 15%, and the board agreed to hold an extra cash reserve. This is an illustrative story, but it shows why recognising a leptokurtic shape matters for risk management.
Watch out
Common mistakes.
- Thinking leptokurtic means higher volatility. Kurtosis is about the shape of the distribution, and two series can have the same standard deviation but very different tails.
- Assuming a tall peak means safer returns. The sharp peak means many small moves, but the same shape comes with more extreme outcomes in the tails.
- Confusing kurtosis with skewness. Skewness describes whether the distribution leans to one side, while kurtosis describes how heavy the tails are.
Questions
People also ask.
What is the kurtosis of a normal distribution?
It is 3, which is why analysts often report excess kurtosis (kurtosis minus 3) so that a normal distribution scores zero.
What are fat tails?
They are the extreme ends of a distribution when they hold more probability than a normal curve would give. Fat tails mean that crashes and spikes happen more often than a bell curve suggests.
Why do financial returns tend to be leptokurtic?
Markets are driven by news, panic and herd behaviour, which can produce sudden large moves. In calm periods, little happens, creating the tall central peak.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%Related