Financial stress and realized volatility: The case of agricultural commodities
Thoughts
Follows @hakkio2009 definition of financial stress. Suggests a FSI can be estimated by applying the Heterogenous Market Hypothesis, which implies auto-regressing daily volatility at different time resolutions.
As indicators, uses: financial volatility, credit spreads, equity valuations, and safe assets. Suggests dynamic principal component analysis weights as the ideal aggregation method.
Determinants:
- Commodities: global financial stress predicted realized volatility of commodities in different forecast horizons (β=0.01)
Connects with: @hakkio2009 @monin2019
Annotations
bonato2024 - p. 1
In general, financial stress is considered to be capturing disruptions to the normal functioning of financial markets.
bonato2024 - p. 1
the objective of our paper is to analyze, for the first time, the predictive ability of financial stress for the second moment movement of agricultural commodities.
bonato2024 - p. 2
Given that rich information contained in intraday data can produce more accurate estimates and forecasts of daily (realized) volatility (McAleer and Medeiros, 2008), we augment the Heterogeneous Autoregressive (HAR) model developed by Corsi (2009) to include measures of global financial stress to predict, both in- and out-of-sample, the daily realized volatility (RV ), as computed from 5-minute-interval data, of 16 important agricultural commodities price returns over the period of September, 2009 to May, 2020.
bonato2024 - p. 2
Moreover, because the HAR-RV model employs RV at different time resolutions to model and predict RV , it can be interpreted as an empirical representation of the theory of the heterogeneous market hypothesis (HMH; Müller et al., 1997). The HMH posits that financial markets (in our case, the markets for agricultural commodities) are populated by various groups of market participants (such as, investors, speculators and traders), who, in turn, differ in their sensitivity to information flows at different time horizons
bonato2024 - p. 3
The global FSI is constructed from 33 financial market variables, such as yield spreads, valuation measures, and interest rates.
bonato2024 - p. 3
The FSI incorporates five categories of indicators: credit, equity valuation, funding, safe assets, and volatility.
bonato2024 - p. 3
weighted average capturing global financial stres
bonato2024 - p. 4
The upper panels of the figures show results for the FSI coefficient β4, and the lower panels show results for the p-values (based on robust standard errors) of this coefficient. In both figures, we use boxplots to summarize the results, where the solid horizontal line denotes the median coefficient (p-value) and the boxes represent the interquartile range of the results.