Chapter 7 — Fee Structure and Fund Performance Part 6: Risk Measures — Standard Deviation, Skewness, Kurtosis, and Maximum Drawdown

An investment in an Alternative Investment Fund (AIF) cannot be evaluated solely on its nominal or compounding returns. Sophisticated investors understand that returns are a function of the risks assumed by the fund manager. Section 7.6 of the SEBI Category III AIF curriculum focuses on the quantitative tools used to measure, analyze, and interpret historical risk.

This part explores the core statistical and mathematical risk metrics—Standard Deviation, Skewness, Kurtosis, and Maximum Drawdown (MDD)—providing step-by-step calculations, Excel functions, and real-world interpretations.

1. Standard Deviation: Measuring Historical Volatility (Section 7.6.1)

In finance, risk is defined as the variability or dispersion of actual returns around an expected or average value. The most common and widely accepted measure of this absolute risk is Standard Deviation.

A. Conceptual Foundations

• Historical Risk Metric: Standard Deviation (represented by the Greek letter Sigma, 𝛔) uses historical returns (ex-post data) to measure the historical volatility of a fund.
• Mean Reversion Baseline: It measures how far the actual periodic returns of the fund deviate from its average (mean) return over a specific timeframe.
• The Dispersion Rule: A larger standard deviation indicates a wider dispersion of returns, signifying higher volatility, lower predictability, and greater historical risk. Conversely, a lower standard deviation indicates that the returns are clustered closely around the mean, representing greater stability.

B. Mathematical Formulas (Linear Simple Text Format)

1. Mean Return (R-bar):

   • Mean Return = Sum(Ri) / n
   • Where:
     • Ri = Return achieved in period/year i
     • n = Total number of periods/years taken into consideration
     • Sum = Summation of all returns from period i = 1 to n

2. Standard Deviation (Sigma):

   • Standard Deviation = SquareRoot( Sum((Ri - Mean Return) ^ 2) / n )
   • Where:
     • Ri = Return achieved in period/year i
     • Mean Return = The arithmetic average calculated in Step 1
     • n = Total number of periods/years
     • Sum((Ri - Mean Return) ^ 2) = The sum of squared deviations of each period's return from the mean

(Note: While some statistical software packages use n - 1 in the denominator for sample standard deviation, the standard NISM curriculum utilizes the population standard deviation formula dividing by n for simplicity in exam calculations).

2. Worked Example: Volatility Comparison (Example 7.7)

We calculate the mean and standard deviation of two competing Category III AIFs over a 5-year operating history to evaluate their risk-return profiles.

Historical Returns Dataset:

• Fund A: Year 1 = 12.50% | Year 2 = 11.25% | Year 3 = 13.65% | Year 4 = 17.30% | Year 5 = 14.65%
• Fund B: Year 1 = 21.50% | Year 2 = -0.50% | Year 3 = 16.75% | Year 4 = 20.25% | Year 5 = 21.90%

Step-by-Step Ledger: Standard Deviation Calculation

Step 1: Calculate the Mean Return (R-bar)

• Mean Return (Fund A) = (12.50% + 11.25% + 13.65% + 17.30% + 14.65%) / 5 = 69.35% / 5 = 13.87%
• Mean Return (Fund B) = (21.50% - 0.50% + 16.75% + 20.25% + 21.90%) / 5 = 79.90% / 5 = 15.98%

Step 2: Calculate Deviations and Squared Deviations

Fund A Ledger:

• Year 1 Deviation: 12.50 - 13.87 = -1.37 | Squared: (-1.37) ^ 2 = 1.8769
• Year 2 Deviation: 11.25 - 13.87 = -2.62 | Squared: (-2.62) ^ 2 = 6.8644
• Year 3 Deviation: 13.65 - 13.87 = -0.22 | Squared: (-0.22) ^ 2 = 0.0484
• Year 4 Deviation: 17.30 - 13.87 = 3.43 | Squared: (3.43) ^ 2 = 11.7649
• Year 5 Deviation: 14.65 - 13.87 = 0.78 | Squared: (0.78) ^ 2 = 0.6084
• Sum of Squared Deviations (Fund A): 1.8769 + 6.8644 + 0.0484 + 11.7649 + 0.6084 = 21.1630

Fund B Ledger:

• Year 1 Deviation: 21.50 - 15.98 = 5.52 | Squared: (5.52) ^ 2 = 30.4704
• Year 2 Deviation: -0.50 - 15.98 = -16.48 | Squared: (-16.48) ^ 2 = 271.5904
• Year 3 Deviation: 16.75 - 15.98 = 0.77 | Squared: (0.77) ^ 2 = 0.5929
• Year 4 Deviation: 20.25 - 15.98 = 4.27 | Squared: (4.27) ^ 2 = 18.2329
• Year 5 Deviation: 21.90 - 15.98 = 5.92 | Squared: (5.92) ^ 2 = 35.0464
• Sum of Squared Deviations (Fund B): 30.4704 + 271.5904 + 0.5929 + 18.2329 + 35.0464 = 355.9330

Step 3: Compute Variance and Standard Deviation (Sigma)

Fund A:

• Variance = Sum of Squared Deviations / n = 21.1630 / 5 = 4.2326
• Standard Deviation (Sigma) = SquareRoot(Variance) = SquareRoot(4.2326) = 2.0573% (reported as 2.06%)

Fund B:

• Variance = Sum of Squared Deviations / n = 355.9330 / 5 = 71.1866
• Standard Deviation (Sigma) = SquareRoot(Variance) = SquareRoot(71.1866) = 8.4372% (reported as 8.44%)

Step-by-Step Risk Interpretation

• Return Perspective: Fund B generated a higher average annual return (15.98%) compared to Fund A (13.87%).
• Risk Perspective: The Standard Deviation of Fund B (8.44%) is more than four times higher than that of Fund A (2.06%).
• Decision Matrix: Fund B's strategy is highly volatile, punctuated by a severe negative drop to -0.50% in Year 2. Fund A's returns are highly stable and consistent. A prudent investor will choose Fund A over Fund B to prioritize consistency, whereas an investor with a very high risk-tolerance might accept Fund B's volatility in exchange for the potential outperformance.

3. Skewness and Kurtosis: Modeling Non-Normal Distributions (Section 7.6.2)

Standard Deviation assumes that investment returns follow a Normal Distribution (the classic "bell-shaped" curve). Under a perfectly normal distribution:

• The curve is completely symmetrical.
• The Mean, Median, and Mode are identical and lie exactly at the center of the curve.
• 68.0 percent of all observed returns fall within 1 Standard Deviation (Mean +/- 1𝛔) of the mean.
• 95.0 percent of all returns fall within 2 Standard Deviations (Mean +/- 2𝛔) of the mean.
• 99.7 percent of all returns fall within 3 Standard Deviations (Mean +/- 3𝛔) of the mean.

However, Category III AIFs (such as hedge funds and long-short funds) do not follow a normal distribution. They utilize options, leverage, short positions, and complex derivatives that create non-linear, asymmetric payoffs. To capture these non-normal behaviors and evaluate the likelihood of extreme events, investors calculate Skewness and Kurtosis.

4. Skewness: Measuring Asymmetry

Definition and Core Purpose

Skewness measures the degree to which a distribution of returns is asymmetric around its mean. It indicates whether the fund's historical performance is dominated by positive or negative outliers (extreme returns that deviate far from the average).

• Normal Distribution Baseline: A perfectly symmetrical normal distribution has a Skewness of Zero.
• Positive Skewness (Skewed Right):
  • Occurs when a dataset contains extreme positive outliers that pull the right tail of the curve outward.
  • In a positively skewed distribution, the Mean is greater than the Mode (Mean > Mode) because the positive outliers pull the average upward.
  • Investor Preference: Investors prefer positive skewness because it indicates a higher probability of occasional large, positive "super-normal" returns alongside consistent average returns.
• Negative Skewness (Skewed Left):
  • Occurs when a dataset contains extreme negative outliers that pull the left tail of the curve outward.
  • In a negatively skewed distribution, the Mean is less than the Mode (Mean < Mode) because extreme losses drag down the average.
  • Investor Warning: Negative skewness represents significant downside risk. It indicates that the fund usually generates small, consistent profits but is subject to occasional catastrophic losses.

Excel Implementation

• Excel Formula: =SKEW(range)
• Where range is the array of historical returns.

5. Kurtosis: Measuring Tail Risk and Peakedness

Definition and Core Purpose

Kurtosis measures the "peakedness" of a distribution and the thickness ("fatness") of its tails compared to a normal distribution. It directly quantifies tail risk—the probability of extreme, outlier events occurring.

• Normal Distribution Baseline: A perfectly normal distribution has a Kurtosis of exactly 3.0.
• Excess Kurtosis: To simplify comparison, analysts calculate Excess Kurtosis:
  • Excess Kurtosis = Kurtosis - 3.0
  • A normal distribution has an Excess Kurtosis of Zero.

Classification of Distributions

1. Leptokurtic Distribution (Kurtosis > 3.0 | Excess Kurtosis > 0):
   • The distribution has a very high, sharp peak and fat tails.
   • This indicates that while most returns are concentrated tightly around the mean, the fund is highly susceptible to extreme outliers (both highly positive and highly negative).
   • Systemic Risk: Leptokurtic funds are riskier because they are prone to unexpected "Black Swan" events or sudden market shocks.
2. Platykurtic Distribution (Kurtosis < 3.0 | Excess Kurtosis < 0):
   • The distribution is relatively flat with thin tails.
   • This indicates that returns are distributed more uniformly across the range with very few extreme deviations.
   • Systemic Risk: Platykurtic funds represent consistent, highly stable performance with low outlier risk.

Excel Implementation

• Excel Formula: =KURT(range)
• Where range is the array of historical returns.

6. Case Study: Multi-Dimensional Risk Analysis (Example 7.8.1 & 7.8.2)

To understand how these metrics interact, we analyze the performance of Fund A and Fund B after a sixth year of returns is added to their historical records.

Statistical Output (Calculated via Excel):

Deep-Dive Analysis of the Case Study

1. Evaluating Skewness

• Fund A (Skewness = 0.705): Fund A exhibits Positive Skewness. This confirms that the distribution is slightly asymmetrical due to positive outliers (like the 17.30% return in Year 4), which pull the mean (13.87%) above the mode. This is highly favorable for investors.
• Fund B (Skewness = -2.222): Fund B exhibits severe Negative Skewness. This indicates a long left tail caused by negative outliers—specifically the -0.50% return in Year 2. This skewness value alerts investors that Fund B's higher mean return is built on a highly asymmetric strategy that carries a severe risk of sharp, negative losses.

2. Evaluating Kurtosis

• Fund A (Kurtosis = 1.162 | Excess Kurtosis = -1.838): Since the kurtosis is well below 3.0, Fund A is Platykurtic. Its returns are highly consistent, with thin tails that indicate a near-zero probability of catastrophic, outlier-driven losses.
• Fund B (Kurtosis = 5.034 | Excess Kurtosis = 2.034): Since the kurtosis is significantly above 3.0, Fund B is highly Leptokurtic. It has fatter tails and a higher peak. This fat-tailed profile warns sophisticated investors of a high probability of extreme, sudden deviations (outliers) from the mean.

3. Holistic Risk-Return Conclusion

If an investor looks only at the Mean and Standard Deviation, they might select Fund B if they have a high risk appetite, reasoning that they are being paid a premium (16.78% return) for taking on volatility. However, once Skewness and Kurtosis are factored in, Fund B is revealed to be significantly more dangerous. Its highly negative skewness (-2.222) and leptokurtic tail risk (Excess Kurtosis of 2.034) mean that its volatility is asymmetric toward the downside. Fund A represents a robust, predictable, and safe investment with stable growth, making it the superior choice for institutional investors.

7. Maximum Drawdown (MDD): Quantifying Downside Risk (Section 7.6.3)

Standard Deviation measures symmetric volatility, and Skewness/Kurtosis evaluate the shape of the return distribution. However, during market crises, investors need to know: “What is the absolute maximum loss this fund could have suffered from its peak valuation?”

The metric used to answer this is Maximum Drawdown (MDD).

A. Definition and Operational Scope

• Peak-to-trough Decline: Maximum Drawdown measures the largest cumulative drop in the Assets Under Management (AUM) or Net Asset Value (NAV) of an AIF, from its peak (highest value) to its trough (lowest value), during a specific reporting period.
• Percentage of Peak: It is always quoted as a percentage of the peak value.
• Risk Window: It measures the greatest potential loss an investor could have experienced if they bought at the absolute peak and sold at the absolute bottom.

B. The Formula (Linear Simple Text Format)

• Maximum Drawdown = (Trough Value - Peak Value) / Peak Value

Where:

• Peak Value: The highest valuation (AUM or NAV) achieved by the fund in the specified reporting period.
• Trough Value: The lowest valuation (AUM or NAV) reached by the fund after the peak value was established, but before a new peak is created.

C. Major Limitations of Maximum Drawdown

While MDD is a powerful measure of downside risk, it has two critical limitations:

1. Ignores Frequency of Loss: It only measures the single worst peak-to-trough drop. It does not indicate how many times the fund experienced large drops during the period.
2. No Indication of Recovery Time: It does not capture how long it takes for the fund to recover from the trough back to its previous peak (known as the Recovery Period).

8. Case Study: Fund PQR vs. NIFTY50 (Example 7.9)

To master MDD calculation and benchmarking for the exam, we analyze the performance of Fund PQR during the highly volatile first quarter of 2020 (the COVID-19 market crash), benchmarking it against the NIFTY50 Index.

Dataset (Q1 2020):

• Dates: Jan 01 | Jan 15 | Feb 01 | Feb 14 | Mar 01 | Mar 23 | Mar 31
• Fund PQR AUM (INR crore): 560.45 | 585.96 | 584.23 | 632.33 | 596.29 | 564.35 | 545.61
• NIFTY50 Index Value: 12202.15 | 12430.50 | 11661.85 | 12113.45 | 11132.75 | 7610.25 | 8597.75

Step-by-Step MDD Calculations

1. Fund PQR Calculation:

• Identify the Peak AUM: The highest value achieved by the fund was on Feb 14, 2020, at INR 632.33 crore.
• Identify the Trough AUM: The lowest value reached after the peak was on March 31, 2020, at INR 545.61 crore.
• Maximum Drawdown (Fund PQR) = (Trough Value - Peak Value) / Peak Value
  • MDD = (545.61 - 632.33) / 632.33
  • MDD = -86.72 / 632.33 = -0.1371 or -13.71%

2. NIFTY50 Benchmark Calculation:

• Identify the Peak Index Value: The highest value achieved by Nifty was on Jan 15, 2020, at 12430.50.
• Identify the Trough Index Value: The lowest value reached after the peak was on March 23, 2020, at 7610.25.
• Maximum Drawdown (NIFTY50) = (Trough - Peak) / Peak
  • MDD = (7610.25 - 12430.50) / 12430.50
  • MDD = -4820.25 / 12430.50 = -0.3878 or -38.78%

Interpretation and Benchmarking Insights

During the Q1 2020 market crash, NIFTY50 suffered a severe drawdown of -38.78%. In contrast, Fund PQR experienced a maximum drawdown of only -13.71%.

• Downside Protection: This benchmark comparison reveals that even though Fund PQR lost INR 86.72 crore in AUM, the manager successfully contained the downside risk. The fund’s loss was only about a third of the broader market's decline.
• Strategy Validation: This confirms that the fund manager’s hedging strategies, long/short structures, or cash management policies worked effectively during a systemic market shock, providing robust capital protection when investors needed it most.

Summary of Statistical Risk Measures

Important Terms for the Exam

• Standard Deviation (Sigma): A statistical measure of the dispersion of a fund's returns around its arithmetic mean.
• Skewness: A measure of the asymmetry of a distribution, indicating whether outliers are concentrated on the positive or negative side.
• Kurtosis: A statistical measure of the "peakedness" and tail weight of a distribution compared to a normal distribution.
• Leptokurtic: A distribution with Kurtosis greater than 3 (Excess Kurtosis > 0), characterized by high peaks and fat tails (extreme outliers).
• Platykurtic: A distribution with Kurtosis less than 3 (Excess Kurtosis < 0), characterized by flat peaks and thin tails (consistent returns).
• Maximum Drawdown (MDD): The largest percentage drop in a fund’s AUM or NAV from its historical peak to its trough during a specific period.
• Downside Risk: The probability of experiencing a financial loss due to a decline in asset prices under changing market conditions.

Chapter 7.6 Review Questions (Exam Practice)

1. Which risk measure is specifically designed to analyze the extent to which a distribution of returns is asymmetric across its mean?

• (a) Standard Deviation
• (b) Mean Return
• (c) Skewness
• (d) Kurtosis
• Correct Answer: (c) — Skewness measures the degree of asymmetry of a distribution around its mean, reflecting the presence of outliers.

2. A Category III AIF has a return distribution with a calculated Kurtosis of 5.034. How is this distribution classified, and what is its Excess Kurtosis?

• (a) Leptokurtic, Excess Kurtosis = 2.034
• (b) Platykurtic, Excess Kurtosis = -2.034
• (c) Mesokurtic, Excess Kurtosis = 5.034
• (d) Symmetrical, Excess Kurtosis = 0.000
• Correct Answer: (a) — A distribution with Kurtosis greater than 3 is Leptokurtic (fat-tailed). Excess Kurtosis = Kurtosis - 3 = 5.034 - 3 = 2.034.

3. If a hedge fund usually generates small, highly consistent gains but is subject to occasional, sudden catastrophic losses, what are its likely Skewness and Kurtosis characteristics?

• (a) Positive Skewness and Kurtosis less than 3
• (b) Negative Skewness and Kurtosis greater than 3
• (c) Zero Skewness and Kurtosis equal to 3
• (d) Positive Skewness and Kurtosis equal to 1
• Correct Answer: (b) — A profile with consistent small gains interrupted by extreme negative losses is characterized by negative skewness (long left tail) and high kurtosis (fat tails due to outliers).

4. What is a primary limitation of using Maximum Drawdown (MDD) as a standalone measure of portfolio risk?

• (a) It is highly complex to calculate without specialized financial calculators.
• (b) It does not account for the frequency of large losses or indicate the time required to recover from those losses.
• (c) It can only be calculated for listed equities and cannot be applied to debt funds.
• (d) It is heavily distorted by the Goods and Services Tax (GST) charged on management fees.
• Correct Answer: (b) — Maximum Drawdown only measures the single worst peak-to-trough decline. It does not capture how often drops occur or how long the fund takes to recover its value.

5. During a market correction, Fund XYZ's AUM fell from a peak of INR 632.33 crore to a trough of INR 545.61 crore. What is its Maximum Drawdown?

• (a) -11.62%
• (b) -13.71%
• (c) -15.90%
• (d) -38.78%
• Correct Answer: (b) — MDD = (Trough - Peak) / Peak = (545.61 - 632.33) / 632.33 = -86.72 / 632.33 = -13.71%.