Volatility Demystified: From Theory to Practice

Market volatility is a fundamental aspect of financial markets, influencing investment & trading decisions, risk management strategies, and asset pricing. Investors often associate volatility with uncertainty and risk but understanding its attributes can help in turning volatility into opportunity and making informed trading decisions. This article explores the key statistical properties of volatility, challenges in estimating it, and practical strategies to manage volatility in portfolios.

‍

Statistical Attributes of Volatility

Measuring Price Returns

Volatility is commonly assessed through price returns, which measure the percentage change in an asset’s price over a given period (e.g., daily, monthly, yearly). There are two primary ways to calculate returns:

• Simple Returns: The percentage change between the current price and the previous price.
• Log Returns: The natural logarithm of the price ratio, often preferred in financial modeling for statistical convenience.

Key Distribution Characteristics

Market returns follow a distribution that can be described using four key statistics:

• Mean: The average return over a given period.
• Variance (or Standard Deviation): The dispersion of returns, commonly used as a measure of volatility.
• Skewness: A measure of asymmetry in the distribution; positive skew implies frequent small gains and occasional large losses, while negative skew indicates the opposite.
• Kurtosis: Measures the presence of extreme values (fat tails); high kurtosis suggests a greater likelihood of large, unexpected price moves.

Comparing Distributions and Standardization

Understanding how different asset returns behave requires comparing their distributions. A statistical technique called standardization makes it easier to compare distributions across different assets. Standardization is achieved by subtracting the mean return from each observed return and dividing by the standard deviation. This process transforms distributions into a common scale where the mean is 0 and the standard deviation is 1. Standardization allows for direct comparisons between assets with varying levels of risk and return.

A commonly referenced distribution in financial markets is the normal distribution, which has a symmetrical bell shape. A normal distribution has:

• A mean of 0
• A standard deviation of 1
• A skewness of 0 (indicating no asymmetry)
• A kurtosis of 3 (suggesting moderate tail behavior)

However, real-world market returns often deviate from this idealized form. Some assets exhibit skewness, meaning that extreme gains or losses occur more frequently on one side of the distribution. Others display kurtosis, meaning they experience more extreme price movements than a normal distribution would predict.

By analyzing these characteristics, investors can determine whether an asset is prone to extreme price swings or behaves more predictably.  

‍

Market Volatility in Practice

Beta as another Risk Measure

Typically when market analysts or financial reporters are speaking of financial market risk, they are referring to either the standard deviation of log returns or the beta of the asset or portfolio in question. Beta measures an asset’s sensitivity to market movements:

• High Beta (>1): The asset moves more than the market.
• Market Beta (≈1): The asset moves in line with the market.
• Low Beta (<1): The asset moves less than the market, providing some downside protection.

The Role of Normality Assumptions in Financial Models

Many financial models assume that price returns follow a normal distribution because it simplifies mathematical calculations. The reasons for this go back over 120 years. In 1900 Louis Bachelier wrote a dissertation for a Phd in mathematics on modeling financial markets. In this paper he assumed that each price change was independent of the previous change. And he assumed that the price return distribution was a normal distribution. The independence assumption says that markets are unpredictable.  The normality assumption was chosen for several reasons.  

• The true underlying price distribution for any market is unknowable. The only information we have as to the true distribution is by observing the price changes that occur. 
• The math of manipulating a normal distribution was well known in the 1900’s.  Additionally, normal distributions were common in a variety of phenomena. Bachelier worked out the math using a normal distribution and discovered that his theoretical results were fairly close to actual market results.  

These normality assumptions underpin widely used models such as:

• The Capital Asset Pricing Model (CAPM): Assumes returns are normally distributed when calculating expected asset returns based on systematic risk (beta).
• The Black-Scholes Option Pricing Model: Assumes asset prices follow a geometric Brownian motion with normally distributed log returns.

However, real-world market returns often exhibit fat tails (greater likelihood of extreme price movements) and volatility clustering (periods of high and low volatility tend to cluster together). These deviations from normality can lead to significant risk miscalculations. Relying too heavily on these models without stress-testing for tail risks may expose investors to unexpected losses during market crises.

‍

Challenges in Estimating Volatility

The Unknowability of True Asset Return Distributions

As mentioned above, modern quantitative finance is based upon the assumption that market and asset returns are largely normally distributed. This assumption is not as reckless as one might believe.  The Central Limit Theorem tells us that combined repeated draws from even non-normal distributions will converge to a normal distribution. Moreover, the normal distribution reveals itself in a variety of data.  However, not every phenomenon is normally distributed. Quite an assortment of distributions is possible, and to make matters worse we will never know what the true distribution is for any process.

The true underlying price distribution for any market is unknowable because:

• We Can Only Observe Past Prices, Not the True Distribution: The actual probability distribution governing asset returns is never explicitly observable. Historical price movements provide only an estimate, which may not hold in the future.
• Financial Markets Are Dynamic: The statistical properties of returns change over time due to market events, policy shifts, and investor behavior, making past data an imperfect predictor.
• Sampling Limitations: Any dataset used for estimation is only a subset of all possible market conditions. If the sample is too small or not representative, risk estimates may be misleading.

Implications for Risk Management

Since the true distribution of returns for any asset is unknown, investors rely on historical data samples to estimate volatility. This introduces two key challenges:

• Sample Size Limitations: Small sample sizes may misrepresent true volatility.
• Incorrect Normality Assumptions: Assuming a normal distribution when reality exhibits fat tails can lead to underestimating extreme risks.

If volatility estimates are too low, investors may take on excessive risk, leading to significant losses during market downturns. Conversely, overestimating volatility may result in overly conservative strategies and missed opportunities.

‍

Managing Volatility in Portfolios

The previous sections defined the statistical attributes of volatility, discussed some of the practical implications of using it as a risk measure and highlighted the challenges in estimating it. In this section we list other aspects of volatility for investors to keep in mind. 

Volatility Regimes and Market Phases

• Volatility is not static; markets transition between low-volatility periods (complacency) and high-volatility periods (crises). 
• Recognizing these shifts can help investors adapt their strategies accordingly.

Implied vs. Realized Volatility

• Realized Volatility: Historical price fluctuations.
• Implied Volatility: Market expectations of future volatility, derived from options pricing.
• Comparing these can provide insights into potential mispricing or volatility spikes.

Behavioral Aspects of Volatility

• Investor psychology influences volatility through fear, greed, and panic selling.
• Herd behavior and feedback loops can amplify market movements.
• Monitoring volatility indices (e.g., VIX) can help gauge market sentiment.

Alternative Volatility Tools

• GARCH models and stochastic volatility models offer different ways to assess market risk.
• Comparing these alternative measures can enhance risk assessment.

The Role of Leverage

• Leverage amplifies the impact of volatility, increasing both potential gains and losses.
• When markets become highly volatile, leveraged positions can lead to rapid liquidation and forced selling, exacerbating market declines. 
• Investors should carefully assess their use of leverage and ensure they have adequate risk controls in place to prevent excessive exposure.

Tail Risk and Extreme Market Events

• Black swan events and fat-tailed distributions highlight the need for robust risk management.
• Tail risk hedging strategies (e.g., options and structured products) can mitigate extreme losses.

Volatility-Based Investment Strategies

• Options Strategies:  Various strategies including, straddles, strangles, condor, butterfly spread etc. 
• Dispersion Trading: Exploiting differences in volatility between an index and its components
• Volatility Targeting: Adjusting portfolio exposure based on changing volatility levels.
• Volatility Arbitrage: Trading the spread between implied and realized volatility.
• Short Volatility Strategies: Selling options or VIX-related products can generate income but comes with high risk.

‍

Conclusion

Market volatility is an unavoidable reality, but it does not have to be a source of fear. Instead, by understanding its fundamental characteristics and employing thoughtful risk management strategies, investors & traders can turn volatility into opportunity. By recognizing the true nature of volatility, investors can better navigate uncertain markets and optimize risk-adjusted returns. 

This article is by necessity a high level review of both the theoretical and practical aspects of market volatility. We will dive into the details of some of these issues in future research.   

‍