Abstract:This is a short review in honor of B. Mandelbrot's 80st birthday, to appear in W ilmott magazine. We discuss how multiplicative cascades and related multifractal ideas might be relevant to model the main statistical features of financial time series, in particular the intermittent, long-memory nature of the volatility. We describe in details the Bacry-Muzy-Delour multifractal random walk. We point out some inadequacies of the current models, in particular concerning time reversal symmetry, and propose an alternative family of multi-timescale models, intermediate between GARCH models and multifractal models, that seem quite promising.
What problem does this paper attempt to address?
The main problem that this paper attempts to solve is to explore how to use the multifractal cascade model and other related concepts to better describe the main statistical characteristics of financial time series, especially the intermittent and long - memory properties of volatility. The paper details the Bacry - Muzy - Delour multifractal random walk model and points out some deficiencies of the existing models, especially in terms of time - reversal symmetry. For this reason, the author proposes a new multi - time - scale model, which lies between the GARCH model and the multifractal model and seems to have great potential.
### Summary of the core issues in the paper:
1. **Intermittent and long - memory properties of volatility**:
- The volatility in financial time series shows significant intermittent and long - memory properties, which cannot be explained by traditional Brownian motion models (such as the Bachelier - Black - Scholes model).
- The multifractal cascade model can better capture these statistical characteristics, but the existing models still have some deficiencies in certain aspects.
2. **Time - reversal symmetry**:
- Existing multifractal models (such as the Bacry - Muzy - Delour model) assume time - reversal symmetry, that is, they cannot distinguish between forward and reverse time series.
- There is obvious time - reversal asymmetry in actual financial data, such as the "leverage effect", that is, price declines are usually accompanied by an increase in volatility. This asymmetry requires new models to explain.
3. **Multi - time - scale model**:
- The author proposes a new class of multi - time - scale models, which lie between the GARCH model and the multifractal model, aiming to better capture the statistical characteristics of financial time series, especially time - reversal asymmetry.
- These models may be used for volatility filtering and option pricing in practice and have important application values.
### Specific problems and their solutions:
- **Volatility distribution**:
- The multifractal model assumes that the local volatility is log - normally distributed, but the actual data may be more suitable to be fitted with an inverse gamma distribution or other distributions.
- The author suggests further research on the impact of different distribution forms on the performance of the model.
- **Intermittency coefficient**:
- The intermittency coefficient predicted by the BMD model is not completely consistent with the values in the actual data, but is generally close.
- Further research is needed on how to estimate and calibrate this parameter more accurately.
- **Tail index**:
- The tail index predicted by the multifractal model is quite different from the values in the actual data, which may be due to some assumptions in the model (such as Gaussian residuals).
- The author proposes the concept of "ergodicity breaking", believing that the theoretical value may not correspond to the actual observed value, but it still needs further verification.
- **Time - reversal invariance**:
- Existing multifractal models cannot explain the time - reversal asymmetry in actual financial data.
- The author suggests introducing negative correlation terms or considering other mechanisms to break time - reversal symmetry, so as to better fit the actual data.
In conclusion, through a detailed analysis of the deficiencies of existing models, this paper proposes a new multi - time - scale model, aiming to more accurately describe the complex statistical characteristics of financial time series, especially the intermittency and time - reversal asymmetry of volatility.