Volatility Trading and Arbitrage Strategies › Risks of Volatility Trading
Quantitative FinanceAll chapters Volatility Trading and Arbitrage Strategies › Risks of Volatility TradingOne section · Sections in this chapter13 / 15 ↓
Reading tools Definitions & inline links
Choose your expertise level to adjust how many terms are explained. Beginners see more tooltips, experts see fewer to maintain reading flow. Hover over underlined terms for instant definitions.
Article links
Make inline references clickable
Volatility trading risks differ from directional equity or fixed income strategies. Understanding these risks is essential for proper position sizing and risk management.
Short Volatility Risk: The Picking Up Nickels Problem
Short volatility strategies typically produce frequent small gains and occasional large losses. This profile resembles "picking up nickels in front of a steamroller"; the strategy works until it doesn't.
The danger stems from several factors. Volatility is inherently mean-reverting but can spike dramatically and persistently during market stress. Historical examples include the 2008 financial crisis, the 2015 Chinese market turmoil, and the February 2018 "Volmageddon" event when inverse VIX products collapsed overnight. During Volmageddon, the VIX doubled in a single day, causing the XIV ETN (which sold VIX futures) to lose 90% of its value and be liquidated.
def simulate_short_vol_strategy(
n_months=120, monthly_premium=0.02, spike_prob=0.02, spike_loss=0.30
):
"""
Simulate returns from a short volatility strategy.
Parameters:
- monthly_premium: Average monthly return in calm periods
- spike_prob: Probability of volatility spike per month
- spike_loss: Loss magnitude during spike
"""
portfolio = 100
portfolio_history = [portfolio]
for _ in range(n_months):
if np.random.random() < spike_prob:
# Volatility spike - large loss
monthly_return = -spike_loss
else:
# Normal month - small gain
monthly_return = monthly_premium + np.random.normal(0, 0.02)
portfolio *= 1 + monthly_return
portfolio_history.append(portfolio)
return np.array(portfolio_history)
# Simulate multiple paths
n_simulations = 100
all_paths = []
for i in range(n_simulations):
np.random.seed(42 + i)
path = simulate_short_vol_strategy()
all_paths.append(path)
all_paths = np.array(all_paths)
# Calculate statistics
final_values = all_paths[:, -1]
max_drawdowns = []
for path in all_paths:
running_max = np.maximum.accumulate(path)
drawdown = (running_max - path) / running_max
max_drawdowns.append(drawdown.max())
avg_max_drawdown = np.mean(max_drawdowns)
Short Vol Strategy Statistics (10-year simulation, 100 paths): Mean Final Value: $502 Median Final Value: $439 5th Percentile: $133 Probability of Ruin (>50% loss): 0.0% Average Max Drawdown: 35.4%
The simulation demonstrates the 'tail risk' inherent in short volatility strategies. While the median path is profitable, the presence of ruinous drawdowns and high probabilities of significant loss illustrates why simple historical mean returns can be misleading for risk assessment in volatility markets.
Long Volatility Risk: The Carry Problem
Long volatility strategies face the opposite challenge: they tend to lose money slowly through carry costs but profit during volatility events. The volatility risk premium means implied volatility typically exceeds realized volatility, creating a drag on long vol positions.
The carry cost manifests in several ways:
- Theta decay: Long options value daily, requiring volatility events to offset
- Roll yield: Long VIX futures contango lose money as they converge to spot
- Variance swap negative carry: Buying variance swaps elevated implied levels loses if realized vol is lower
The behavioral challenge for you is maintaining conviction through extended periods of losses while waiting for the volatility event that justifies the strategy.
Position Sizing and Leverage
Position sizing is critical because of these asymmetric risk profiles:
For short volatility: You should size positions to survive 3-4 standard deviation moves. The February 2018 VIX spike was approximately a 7-sigma event under normal assumptions, which shows the inadequacy of normal distribution assumptions. For long volatility: You should size positions to survive extended carry costs. A long vol strategy might need to withstand 12-24 months of bleeding before a payoff event occurs.
A useful framework is for you to define the maximum acceptable drawdown and work backward to position size:
def calculate_position_size(
max_drawdown, expected_loss_in_spike, confidence_level=0.99
):
"""
Calculate appropriate position size for volatility strategies.
Parameters:
- max_drawdown: Maximum acceptable portfolio loss (e.g., 0.20 for 20%)
- expected_loss_in_spike: Expected loss on position in vol spike (e.g., 0.50 for 50%)
- confidence_level: Probability of not exceeding max_drawdown
"""
# Position size = max_drawdown / expected_loss
position_size = max_drawdown / expected_loss_in_spike
return min(position_size, 1.0) # Cap at 100%
# Example for short vol strategy
max_acceptable_dd = 0.15 # 15% max drawdown
expected_spike_loss = 0.40 # 40% loss on position in spike
recommended_size = calculate_position_size(
max_acceptable_dd, expected_spike_loss
)Position Sizing for Short Volatility Strategy: Maximum Acceptable Drawdown: 15% Expected Position Loss in Vol Spike: 40% Recommended Position Size: 37% of portfolio
This position size ensures that a 40% loss in a volatility spike does not exceed the 15% drawdown limit. This inverse sizing approach limits losses in fat-tailed markets.
Correlation with Market Stress
Volatility positions have distinct correlation properties that affect portfolio construction:
- Short volatility is implicitly long equities. When markets crash, volatility spikes, and short vol positions lose money at the worst time.
- Long volatility provides crisis alpha: profits during market stress that can offset equity losses.
This correlation structure means that short volatility strategies provide false diversification. They may show low correlation to equities during calm periods but become highly correlated during crises when diversification is most needed.