Corporate Default Risk Classification with Complex Boundaries
An investment fund needs to classify mid-sized companies as 'Solid Solvency' or 'Insolvency Risk' using financial ratios where the boundary between the two states is very narrow and non-linear.
Where the money comes from
What to do now
- 1Gather ratios of companies that defaulted and did not default over the last 10 years.
- 2Validate the model with crisis years left out of the training.
- 3Use it as a pre-filter ahead of the investment committee for 2 quarters.
For your committee
One page with the decision, the impact, the risks, the steps and the KPI to track.
The print dialog opens: choose "Save as PDF".
Watch out: more volume is not always more profit
Seeking more return by buying debt of companies on the boundary: the extra spread rarely compensates for a default. Optimise expected return net of losses, not nominal return.
KPI for the committee
Portfolio default rate versus the benchmark index.
Detail for your data team
Proposed technique: Support Vector Machines (SVM)
Why: SVM finds the optimal boundary with the maximum possible separation margin between healthy and insolvent companies, handling complex financial relationships with great rigour.
Cost of getting it wrong: Investing in a company that enters bankruptcy proceedings means capital losses in the millions.
Data needed: Audited Balance Sheets: Liquidity ratios, financial leverage, EBITDA margin and free cash flow.; Sector Data: Sector volatility and interest-rate exposure.