In this blog, we discuss the relationship between realized volatility (RV) and implied volatility (IV), focusing on BTC as a case study. The goal is to introduce a simple option trading strategy that takes advantage of potential mispricing between RV and IV. We'll then compare the performance of this strategy against an alternative approach that doesn't incorporate RV information, highlighting the value of using RV in volatility-based trading strategies.

Disclaimer:

This analysis is for educational purposes only and is not financial advice. The strategies discussed, particularly those involving shorting options, carry significant risks, especially in volatile markets like cryptocurrencies. This blog does not cover the impact of margin, which can increase risk for traders.

What is Realized Volatility?

Realized Volatility (RV) measures the historical volatility of an asset by calculating the standard deviation of its spot price over a defined period. This gives traders insight into how much an asset’s price has fluctuated in the past. For more details on how RV is calculated, refer to here.

The key parameters in RV calculation are the frequency of observations and the lookback period. For instance, daily RV uses daily spot price observations, while weekly RV uses weekly observations. The lookback period determines how far back the analysis goes, providing an average volatility over that time frame. A well-chosen lookback period smooths out short-term fluctuations but remains relevant to current market conditions. If the period is too long, older data may skew the results and fail to reflect the latest market dynamics. Therefore, the lookback period must be long enough to ensure accuracy but not too long to lose relevance.

Daily Realized Volatility Observed at 8 AM UTC with Various Lookback Periods.

The graphic above shows the daily Realized Volatility (RV) using different lookback periods (15, 30, 60, and 90 days). As the lookback period increases, RV becomes smoother, reducing the influence of short-term fluctuations. This highlights the importance of selecting an appropriate lookback period when analyzing volatility—it's a crucial parameter for traders.

In this analysis, we focus on daily RV, as our strategy centers around daily expiring options. By aligning the RV calculation with the options' expiration timeframe, we ensure that the volatility we're analyzing is relevant to the short-term nature of the trades.

What is Implied Volatility?

Implied Volatility (IV) reflects the market's expectations for future volatility. In the options market, IV represents the volatility implied by the market prices of options contracts. For this analysis, we will focus on the 1-day At-The-Money (ATM) IV, which estimates how volatile the market expects the asset to be over the next 24 hours.

1D ATM Implied Volatility of BTCUSD at 8AM UTC.

The graphic above shows 1-day ATM implied volatility (IV) observed at 8 AM UTC daily, from July 2023 to October 2024. The IV is highly volatile, with significant fluctuations occurring from one day to the next. For example, on August 3rd, 5th, and 6th, 2024 (highlighted by coloured circles), IV surged from 35% to over 110%, then fell back to 75% the following day.

These significant changes in implied volatility (IV) are the result of short-term market dynamics. To demonstrate this, we’ve plotted the corresponding volatility smiles for the highlighted dates. Since these smiles are well-calibrated, it confirms that the observed IV fluctuations are directly influenced by shifts in market conditions.

1D BTCUSD Volatility Smile at 8AM UTC on 03-08-2024, 05-08-2024 and 06-08-2024.

Comparing Realized Volatility to Implied Volatility

When Realized Volatility (RV) is lower than Implied Volatility (IV), it suggests that the market may be overestimating future volatility based on historical data. This can create opportunities for traders to sell volatility through strategies like selling straddles or strangles.

However, since RV depends on the lookback period, it's important to determine which RV to compare against IV. In the graphics below, we plot daily ATM IV against RV for different lookback periods. The correlations between the two are weak, ranging from 15% to 24%, indicating that no single lookback period consistently outperforms the others in capturing this relationship.

Daily Realized Volatility as a function of 1D ATM Implied volatility between July 2023 and October 2024.

The percentage of days when daily Implied Volatility (IV) is lower than daily Realized Volatility (RV) for various lookback periods, spanning from 1st August 2023 to 20th October 2024, varies between 50% and 60%.

Trading Strategy Based on RV and IV

Our strategy takes advantage of the difference between Realized Volatility (RV) and Implied Volatility (IV). We’ll compare RV, using different lookback periods, with the 1-day ATM IV at 8 a.m. UTC. If RV is lower than IV, we will sell a 1-day ATM straddle, aiming to profit from the possible market’s overestimation of future volatility.

Why Sell a Straddle?

A straddle is an options strategy where both a call and a put option are sold at the same strike price, with the expectation that the asset's price will remain close to that level. In this strategy, the straddle is held until maturity, without dynamic hedging. The options are sold when implied volatility (IV) is high, particularly when IV exceeds realized volatility (RV), allowing the seller to collect a higher premium.

The strategy profits if the asset's price remains stable, enabling the seller to retain part of the premium. However, it’s important to note that selling options carries significant risk if there are large price movements away from the strike price, as this can result in substantial losses for the seller.

Backtesting the Strategy

We will backtest this strategy and compare it to a simpler approach where we sell the 1-day ATM straddle every day, without considering the relationship between RV and IV. This comparison will allow us to assess how using RV as a signal in our strategy impacts overall performance. Since none of the lookback periods showed a clear advantage in the earlier analysis, this parameter will need to be calibrated during the backtest. It is a crucial element of the strategy that requires optimization.

For this analysis, we will use the average mid 1d-ATM volatility observed in the market. The straddle will be sold at the bid price, reflecting the highest price buyers are willing to pay. We will apply a 5% volatility spread as a proxy, which is a conservative estimate within the typical bid-ask volatility spread range of 3-8%, accounting for the difference between buying and selling volatility. This adjustment ensures that the strategy factors in market liquidity and spread costs.

Additionally, trading commissions will be included, similar to the fees charged by major exchanges. A fee will be applied for opening positions, calculated as a percentage of the spot index price of the underlying asset (typically 0.03%) at the time of the transaction, multiplied by the number of contracts. These fees will be capped in line with industry-standard fee structures as a percentage of the option price.

Backtesting the 1D ATM Straddle Short-Selling Strategy with a Signal Based on Realized Volatility (RV) using various Lookback periods, and comparing it to the approach of Selling the Straddle Daily, regardless of RV.

Conclusion

The backtesting results demonstrate that the strategy is sensitive to the chosen lookback period, with the 90-day lookback providing the best PnL profile while reducing the number of trades. These findings underscore the value of using Realized Volatility (RV) as a filter in options trading strategies. Incorporating RV as a volatility indicator can enhance the base strategy of selling daily 1D ATM straddles by mitigating risk during periods of high volatility and optimising overall performance.

By combining historical volatility measures, such as RV, with options trading strategies, traders can gain deeper insights and make more informed decisions. This approach highlights the potential to improve traditional options strategies by integrating reliable volatility metrics into the trading process.