The Financial Health Rating (FHR®) is RapidRatings’ proprietary and innovative measure of risk that estimates the financial strength and weakness of firms. This is achieved by combining the Core Health Score (CHS) models that focus on financial health over the medium term with 11 additional ratios that are specifically focused on short-term (within the next year) default estimation.
The FHR system is sensitive to balance sheet performance by providing additional emphasis on total debt and liquidity, and has a sensitivity to managing obligations through earnings performance, which together help sharpen the default prediction accuracy and improve the discrimination between high-risk survivors and high-risk defaulters in the next 12 months.
|
For more information on our modeling process, we recommend reviewing our backtesting section, as well as our Advanced Model FAQs |
Calculating Core Health
The Core Health model is a multivariate econometric pseudo panel data model which measures the financial health of a firm along a continuum from 0 to 100. The FHR model is designed to fine-tune the Core Health model and sharpen our estimates of the probability of default by combining the Core Health model with 11 Resilience Indicators. All 11 resilience ratios are dynamically interactive and a change in one changes the others.
The Core Health model, our traditional rating approach, measures a firm’s efficiency and competitiveness with a horizon of the next 2 to 3 years. The model utilizes 24 industry models, each with 62 bounded scoring distributions (one for each ratio). The resulting ratio scores are weighted on an industry-specific basis.
Calculating the CHS is a four-step process:
- Use the financial statements to calculate 62 ratios
- Benchmarks each of the 62 ratios on an industry-specific basis for each firm against a global distribution of peers in 24 industries covering up to 40 years of performance data. This produces a 'ratio score'
- Apply an industry-specific weighting to each ratio score. These weights are based on the ratio’s predictive significance (determined by econometric modeling)
- Produce a single weighted risk score, our Core Health Score (CHS)
Resilience and the FHR
The FHR Model is a probabilistic statistical classification model that estimates a firm’s probability of default along a continuum from 0 to 100. The inputs for the FHR are the CHS and 11 resilience ratios.
- Leverage
- Total Debt to Total Assets
- Liquidity
- Working Capital to Total Assets
- Cash to Current Liabilities
- CFO to Current Liabilities
- Earnings Performance
- COGS to Sales
- Operating Profit to Total Assets
- NPAT to Total Assets
- Retained Earnings to Total Assets
- CFO to Current Debt Service
- Operating Profit to Interest Expense
- Operating Profit to Current Debt Service
The resilience ratios vary in significance. The output of combining the CHS and resilience ratios is transformed into an estimated probability of default and an FHR.
Estimated Probability of Default (EPD)
The output of the model, from which the probability of default is also derived, is monotonically transformed into the FHR. This transformation ensures that FHR is a score between 0 and 100, and that lower values correspond to a higher probability of default. Based on testing of over 1,300 defaults between 1991 and 2013, a particular transformation was chosen so that the probability of default is about 1% for an FHR of 40 and 0.1% for an FHR of 60. The threshold at 40 between Medium Risk and High Risk is an important workflow lever for many users, and the group of companies whose probability of default is greater than 1% encompasses over 90% of defaults rated since 1991.
Once a company passes the High Risk threshold with a rating below 40, it demonstrates not only a material level of default risk, but that default risk increases geometrically with each subsequent reduction in FHR, approximately doubling for each 5-point FHR deterioration. The relationship between probability of default and FHR is presented with a logarithmic scale in figure 1 and with a linear scale in figure 2.
Figure 1: Estimated Probability of Default and the FHR (Logarithmic Scale
Figure 2: Estimated Probability of Default and the FHR (Linear Scale)
Please reach out to our support team if you have further technical questions.