Dynamic_Factor_Model

By shrijeetverma13 · February 10, 2026

Data source:

The analysis examined how debt levels relate to various risk measures and state transitions in your dataset. A comprehensive visualization was created showing the relationship between debt and Risk.2F across different states, along with debt distribution patterns by state.

The analysis examined 192 data points and found that credit spread (信用利差) has very weak correlations with both systemic risk measures. The correlation with MES is 0.0219 and with CoVaR is -0.0100, indicating virtually no linear relationship between credit spread and these systemic risk indicators.

CoVaR evolved significantly from 2005 to 2020, ranging from a minimum of 0.0069 to a maximum of 0.0474. The absolute peak risk period occurred on September 30, 2015, when CoVaR reached its highest value of 0.0474. The visualization shows the complete time series with both the actual CoVaR values and a 6-month rolling average trend line, making it easy to identify periods of elevated systemic risk.

The four risk measures show distinct characteristics across the 2005-2020 period. Risk.2F operates on a different scale (averaging 0.24) compared to the other three measures which center around zero. Risk.PC and Risk.Entropy are very closely aligned with a 0.956 correlation, while Risk.3F also moves in sync with them (0.974 correlation with Risk.PC). All measures exhibited heightened volatility during major financial stress periods in 2008, 2015, and 2020.

Factor1 is the dominant contributor to overall risk assessment, accounting for 81.3% of the total contribution compared to Factor2's 18.7%. This analysis examined both factors' mean absolute values, variance contributions, and correlations with the overall Risk.2F measure.

The analysis reveals a negative relationship between volatility and returns, meaning higher volatility tends to be associated with lower returns. This relationship varies significantly across different market conditions, with a stronger negative correlation during normal periods (-0.33) compared to stress periods (-0.19).

The analysis reveals a weak negative relationship between term spread (期限利差) and subsequent risk state changes. When risk states are about to change, the term spread averages 0.750, compared to 0.936 during stable periods - a difference of 0.186. This suggests that lower term spreads tend to precede transitions between risk states.

The SFR.Index showed significant variation over the 2005-2020 period, with an overall average of 91.35 and values ranging from 49.52 to 135.96. The index peaked at 135.96 in December 2016. A clear trend visualization shows the index fluctuating throughout the period, with red markers highlighting stress periods when the financial stress indicator (状态) was active.

The analysis reveals the relationship between M2 money supply and seven systemic risk indicators using 192 observations. Two visualizations were created: a correlation bar chart showing the strength and direction of relationships between M2 and each risk indicator, and a time series chart comparing M2 trends with Risk.3F over time.

The analysis identified periods when multiple risk measures were simultaneously elevated during 2008, 2015, and 2020. A visualization was created showing the time series of elevated risk counts across all periods, with vertical markers highlighting the three key crisis years. The threshold was set at 3 or more elevated measures (out of 7 total risk metrics) to identify significant simultaneous risk events.

The analysis reveals when Risk.Equal and Risk.Entropy diverge significantly from Risk.PC estimates by calculating absolute differences over time. A visualization shows the divergence patterns, with threshold lines marking significant deviations (above the 75th percentile). The chart displays two trend lines tracking how much each risk measure differs from Risk.PC throughout the time period.

The analysis examined how OECD leading indicators relate to seven domestic risk measures across 192 observations. Two visualizations were created: a correlation bar chart showing the strength of relationships between OECD and each risk measure, and a time series comparison displaying normalized trends of OECD indicators versus Risk.3F over time.

The analysis reveals clear patterns in Industrial Value Added (IVA) across different risk regimes. During low risk periods, IVA averages 12.24%, which is notably higher than the 9.19% average during high risk periods. This 3.05 percentage point difference indicates that industrial activity is significantly stronger during stable economic conditions.

The analysis tracked MES (Marginal Expected Shortfall) over 16 years from January 2005 to December 2020. MES values ranged from 0.0110 to 0.0738, with 25 significant peaks identified throughout this period. The visualization shows the complete evolution of MES over time, with peaks highlighted in red and a threshold line marking when MES reaches elevated levels.