- The report constructs a bottom-up social financing (社融) prediction framework, which predicts each sub-item of social financing based on their respective characteristics[7][11] - The framework includes various sub-items such as RMB loans, enterprise bill financing, government bonds, corporate bonds, foreign currency loans, trust loans, entrusted loans, undiscounted bank acceptance bills, domestic stock financing of non-financial enterprises, loan write-offs, and asset-backed securities of deposit-taking financial institutions[12] - For RMB loans, the model uses PMI and Tangshan steel plant capacity utilization rate as independent variables for rolling regression predictions[12] - Enterprise bill financing is predicted using the rediscount rate as an exogenous variable with a 5-year rolling autoregression[12] - Government bonds are tracked using high-frequency issuance and maturity data, with adjustments for discrepancies in data scope[12] - Corporate bonds are predicted using a 5-year rolling regression to reweight sub-items, effectively reducing discrepancies in data scope[12] - Foreign currency loans are predicted using the average of the past three months[12] - Trust loans are approximated by tracking the issuance and maturity of collective and single trust disclosures[12] - Entrusted loans use the average of the past 12 months, with additional judgment for infrastructure-related increments[12] - Undiscounted bank acceptance bills use the average of the same period over the past three years due to the cessation of high-frequency data publication by the Shanghai Bills Exchange[12] - Domestic stock financing of non-financial enterprises is predicted by deducting the financial enterprise portion from the monthly net financing data of equity financing (including IPOs, additional issuances, rights issues, and preferred stocks)[12] - Loan write-offs use the value from the same period last year due to significant seasonal effects[12] - Asset-backed securities of deposit-taking financial institutions are tracked using high-frequency data of credit ABS net financing[12] Model Evaluation - The bottom-up prediction framework allows for detailed depiction and accurate prediction of both the total amount and structure of social financing[7][11] Model Testing Results - Predicted new social financing for March 2026 is approximately 6.06 trillion yuan, an increase of 0.16 trillion yuan year-on-year[7][11] - Social financing TTM (Trailing Twelve Months) is expected to increase by approximately 0.48% month-on-month[7][11] - Social financing stock is expected to grow by approximately 8.16% year-on-year[7][11] - New RMB loans are predicted to be approximately 3.95 trillion yuan, an increase of 0.13 trillion yuan year-on-year[7][11] - Government bond net financing is expected to be approximately 1.17 trillion yuan, a decrease of 0.32 trillion yuan year-on-year[7][11] - Corporate bond net financing is expected to be approximately 0.35 trillion yuan, an increase of 0.44 trillion yuan year-on-year[7][11]