About Equifax:
Equifax Inc. is a consumer credit reporting agency. It collects and aggregates information on over 800 million individual consumers and more than 88 million businesses worldwide. Headquartered in Atlanta, Georgia, Equifax operates or has investments in 24 countries in North America, Central and South America, Europe and the Asia Pacific region.
About the Equifax Retail Bullish Model:
The model identifies optimal conditions for entry based on trends in bank card and retail card transactions of US consumers. In addition, the model may consult global macro, fundamental, and technical factors from Neuravest. The goal of the strategy is to outperform the benchmark XRT (SPDR S&P Retail) in total return, Sharpe ratio, and drawdown. The model covers twelve publicly traded companies in the retail sector: ‘AMZN’, ‘DG’, ‘AZO’, ‘KR’, ‘MCD’, ‘WBA’, ‘HD’, ‘WMT’, ‘V’, ‘TGT’, ‘MA’, ‘SBUX’, ‘NKE’. Some of the Equifax and Neuravest factors are listed below:
- Delinquency Rate - Percentage of consumers who are unable to pay their debt.
- Utilization Rate - Ratio of a consumer’s debt to their total credit limit.
- Debt to Income - Ratio of a consumer’s debt to their annual income.
- Inquiries TTM - Number of credit inquiries made by a consumer over the past 12 months.
- Total Balance - The total balance across all consumers.
- Payment Ratio - The ratio of a consumer’s monthly payment to their outstanding debt.
- Available Credit - The difference between a consumer’s total credit limit and their outstanding debt.
- Payment to Due - The ratio of a consumer’s monthly payment to their payment due.
- Avg. # Accounts - Average number of accounts per consumer.
- # New Accounts - Number of newly opened accounts.
- US Industrial Production - A measure of the output of industrial establishments.
- Neuravest Macro - CPI / Inflation - Percent change in inflation based on the Consumer Pricing Index.
Here’s how the backtest is generated
- This is a long only model based on the retail sector.
- Each day models scan a list of major publicly traded companies in the retail space.
- The positions are entered based on Extreme Gradient Boosting machine learning models .
- If constituents are identified for entry, 8.33 % of available cash is allocated to each constituent
- There is conditional hold in place. Each position held as long as the signal is active
- Models are retrained and feature selection is done in walk-forward fashion using sliding window.
- Transaction costs and slippage have been considered in the report.