Month: November 2025

Fractal Market Hypothesis: From Theory to Practice

Fractal Market Hypothesis is an alternative framework that models financial markets through long-memory and multi-scale dynamics. There is a growing trend in the industry to incorporate it—first in analyzing the behavior of underlying assets, and more recently in the pricing of financial derivatives such as futures. In this post, we will examine these developments.

Fractal Market Hypothesis: Quantification and Usage

The Fractal Market Hypothesis (FMH) is a theory that suggests that financial markets behave in the same way as natural phenomena and are subject to the same physical laws as found in nature. It suggests that financial markets are composed of similar patterns which repeat over and over again at different scales. These patterns can be used to identify market trends and can help investors make more informed decisions.

The Fractal Market Hypothesis is one of the alternatives to the Efficient Market Hypothesis (EMH) which states that all available information is already factored into the price of a security. The other alternative is the Adaptive Market Hypothesis (AMH).

Reference [1] examined how the fractal nature of the financial market can be quantified and used in investment analysis.

Fractal Market Hypothesis (FMH):

-Suggests financial markets mimic natural phenomena, governed by the same physical laws.

-Identifies repeating patterns at different scales in financial markets.

-Offers a quantitative description of how financial time series change.

Comparison with Efficient Market Hypothesis (EMH):

-FMH contrasts with EMH, which claims all available information is already reflected in security prices.

-FMH, along with Adaptive Market Hypothesis (AMH), presents alternatives to EMH.

Quantification and Usage of FMH:

-The paper quantifies the fractal nature of developed and developing market indices.

-FMH posits self-similarity in financial time series due to investor interactions and liquidity constraints.

-Market stability is influenced by liquidity and investment horizon heterogeneity.

Market Dynamics and Stability:

-FMH suggests that during normal conditions, diverse investor objectives maintain liquidity and orderly price movements.

-Under stressed conditions, herding behavior reduces liquidity, leading to market destabilization through panic selling.

Reference

[1] A. Karp and Gary Van Vuuren, Investment implications of the fractal market hypothesis, 2019 Annals of Financial Economics 14(01):1950001

Fractional Geometric Brownian Motion and Its Application to Futures Arbitrage

While the previous paper discusses the FMH from an investment perspective, Reference [2] reflects a recent trend in quantitative research—namely, incorporating the FMH into the pricing of financial derivatives.

The paper proposed an extension based on fractional Brownian motion (FBM), which incorporates trend fractal dimensions (FTD)—distinguishing between upward (D⁺) and downward (D⁻) dimensions—combined with momentum lifecycle theory.

The authors developed a pricing framework for futures under this setup. Because FBM is not a semi-martingale in the classical sense, they adjusted the drift of the log-price process to reconcile fractal dynamics with approximate arbitrage-free pricing.

Afterward, they constructed a futures pricing model and designed an arbitrage strategy based on the futures–cash basis. The strategy operates as follows:

-Rule 1: Execute a positive arbitrage (sell futures, buy spot/ETF) when the basis series enters the low reversal phase, as identified by the conditions on D⁺ and D⁻.

-Rule 2: Close the positive arbitrage position (buy futures, sell spot/ETF) when the basis series enters the high reversal phase, or, depending on market rules and strategy design, open a negative arbitrage position.

Findings

-The study challenges traditional futures pricing models based on the efficient market hypothesis, noting their limitations in capturing complex market behavior and their tendency to produce significant pricing errors.

-It introduces the fractal market hypothesis (FMH) as a more effective framework that accounts for long memory and multi-scale market dynamics.

-A fractal futures pricing model is developed by incorporating the Hurst exponent and a cash-futures arbitrage strategy that uses trend fractal dimensions (D⁺ and D⁻) and momentum lifecycle logic to generate dynamic trading signals.

-Empirical testing using CSI 300 data shows that the fractal model substantially reduces pricing errors relative to the traditional cost-of-carry model.

-The proposed fractal-based arbitrage strategy achieves higher returns, stronger risk-adjusted performance, and lower drawdowns compared to conventional static-threshold approaches.

-Backtesting results indicate a total return of 12.71% versus 7.06% for the traditional strategy, with a positive Sharpe ratio of 0.32 compared to a negative −0.61.

-The strategy demonstrates exceptional resilience during market stress, such as the 2015 crash, limiting losses to −0.83% while traditional approaches lost −5.82%.

-This robustness under extreme conditions highlights the model’s effectiveness for both profitability and capital preservation.

Overall, the findings validate the practical value of the fractal market hypothesis for developing adaptive, accurate, and profitable pricing and arbitrage tools.

Reference

[2] Xu Wu and Yi Xiong, A fractal market perspective on improving futures pricing and optimizing cash-and-carry arbitrage strategies, Quantitative Finance and Economics, Volume 9, Issue 4, 713–744.

Closing Thoughts

In summary, both articles underscore the growing relevance of the Fractal Market Hypothesis as an alternative framework for understanding modern financial markets. The first article outlines FMH’s theoretical foundation, emphasizing its focus on multi-scale behavior, liquidity, and investor horizon heterogeneity. The second article extends this perspective into practical applications, demonstrating how fractal-based pricing models and arbitrage strategies can outperform traditional approaches and remain resilient under stress. Together, they show that FMH is evolving from a descriptive theory into a useful quantitative tool for pricing, risk management, and strategy design.

Volatility vs. Volatility of Volatility: Conceptual and Practical Differences

Volatility and volatility of volatility are highly correlated and share many similar characteristics. However, there are subtle but important differences between them. In this post, we will examine some of these differences and explore an application of volatility of volatility in portfolio management.

Improving Portfolio Management with Volatility of Volatility

Managing portfolios using volatility has proven effective. Reference [1] builds on this research by proposing the use of volatility of volatility for portfolio management. The rationale behind using volatility of volatility is that it represents uncertainty.

Unlike risk, which refers to situations where future returns are unknown but follow a known distribution, uncertainty means that both the outcome and the distribution are unknown. Stocks may exhibit uncertainty when volatility or other return distribution characteristics vary unpredictably over time.

Practically, the author used a stock’s daily high and low prices to derive its volatility of volatility.

Findings

-The study investigates how volatility-managed investment strategies perform under different levels of uncertainty across stocks and over time.

-A new measure of volatility-of-volatility (vol-of-vol) is introduced as a proxy for uncertainty about risk, capturing a unique dimension distinct from traditional volatility.

-Results show that abnormal returns from volatility management are concentrated in stocks with low uncertainty and during periods of low aggregate uncertainty.

-The effectiveness of sentiment-based explanations for volatility-managed returns is conditional on the level of uncertainty.

-Cross-sectional differences in uncertainty help explain why volatility-managed factor portfolios perform unevenly across stocks and time.

-Theoretical analysis extends a biased belief model, showing that higher vol-of-vol reduces volatility predictability and belief persistence, weakening the benefits of volatility timing.

-The study hypothesizes that volatility management is most effective for low-uncertainty stocks and in low-uncertainty market environments.

-Empirical tests use realized vol-of-vol derived from intraday high and low prices as the measure of uncertainty.

-Consistent with prior literature, uncertainty is positively related to future returns and contains unique predictive information not explained by other stock characteristics.

-Volatility management significantly improves risk-adjusted performance in low-uncertainty stocks and during low aggregate uncertainty periods, while uncertainty also helps explain performance variation across asset pricing factor portfolios.

In short, using the volatility of volatility as a filter proves to be effective, particularly for low-uncertainty stocks.

We find it insightful that the author distinguishes between risk and uncertainty and utilizes the volatility of volatility to represent uncertainty.

Reference

[1] Harris, Richard D. F. and Li, Nan and Taylor, Nicholas, The Impact of Uncertainty on Volatility-Managed Investment Strategies (2024), SSRN 4951893

Beyond volatility of volatility

This section is written by Alpha in Academia

The Volatility of Volatility Index (VVIX) is a composite measure, driven by both short-term market panic and long-term risk expectations.

For years, the VVIX, often dubbed the “fear of fear” index, was treated primarily as a measure of the volatility of volatility (VOV), but new research reveals it contains a second, equally critical component: Long-Run Variance (LRV).

Figure 1: Time series of the squared VVIX Notes: This figure reports time series of the squared VVIX from April 4, 2007, to August 31, 2023; these are all reported on a logarithmic scale for the vertical axis, while the horizontal axis remains linear. The squared VVIX corresponds to the daily closing value retrieved from CBOE. The shaded areas indicate periods of financial distress, such as the GFC, the European debt crisis, and the COVID-19 pandemic. Note that financial distress does not correspond to the NBER recession.

Using a sophisticated model and leveraging a novel technique involving risk-neutral cumulant data extracted from VIX options, researchers decomposed the VVIX dynamics. Their analysis reveals that the factors driving the index change dramatically depending on market conditions. Specifically, the short-term panic measure, VOV, significantly contributes only during acute periods of financial distress, which aligns with intuition. However, during stable or bull markets, the VVIX is primarily driven by the LRV component, reflecting persistent, underlying risk expectations.

In fact, when testing the explanatory power on market-neutral straddle portfolios using S&P 500 options, combining LRV and VOV produced an adjusted explanatory power up to three times greater than baseline models. The finding shows that the index provides “a clear answer to the question of the informational content of the VVIX, showing that it reflects not only the VOV but also an additional important component—the LRV”. Investors should thus view the VVIX not just as a fear gauge, but as a dual-sensor monitoring immediate market stress and long-term risk.

Reference

[2] Bacon, Étienne and Bégin, Jean-François and Gauthier, Geneviève, Beyond volatility of volatility: Decomposing the informational content of VVIX, 2025, SSRN 5611090

Closing Thoughts

In summary, both studies emphasize the role of volatility-of-volatility in understanding risk and market behavior. The first shows that volatility management is most effective in low-uncertainty environments, while the second reveals that the VVIX reflects not only short-term market stress but also long-term risk expectations. Together, they suggest that volatility-of-volatility offers deeper insight into both portfolio performance and the broader dynamics of market uncertainty.