5
$\begingroup$

Does anyone know of any papers about credit rating development or probability of default estimation done based on financial ratios that also include methodology and maybe good/bad criteria?

Something like they have some financial ratios and then they have some methodology that reduces it to a few financial ratios and then they make a regression model out of it or something.

$\endgroup$
1
  • 1
    $\begingroup$ I've relatively exhaustive research in this particular area, never came across anything similar to what you're asking for though. I'm interested in the literature as well if you find any. Best of luck! $\endgroup$ Dec 27, 2014 at 12:53

4 Answers 4

3
+50
$\begingroup$

Most of the papers concern CDS spreads which you will need to convert to a PD.

Paper using country specific fundamentals: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2517018

This paper uses leverage: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2361872

Another one that decomposes them against peer groups: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2413011

Comparing spreads and ratings: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1551406

$\endgroup$
1
  • $\begingroup$ Haven't gone through them much but you provided the most links so I'll pick you for bounty. Does accepting your answer award you the bounty automatically? $\endgroup$
    – BCLC
    Jan 3, 2015 at 13:56
2
$\begingroup$

I am also not aware of any papers in this area. But having developed many such models, I can list the important steps:

  1. Decide on the target variable: usual choices are historical default data, agency ratings and expert rankings
  2. Create a sample containing the possible predictors
  3. Reduce the list with the help of some expert, e.g. exclude all the predictors deemed to be irrelevant
  4. Analyse the predictors standalone, e.g. with a rank correlation
  5. Discuss the poor predictors with the experts and possibly eliminate some. At this stage one usually has about 10 to 20 predictors
  6. Run regression analyses ((ordinal) logistic regression) after standardisation with all the possible combinations (e.g. consider combinations with at most 8 factors and more than 2 predictors).
  7. Check for each combination: the coefficients are intuitive (e.g. higher asset/debt better the rating etc.) and that they are high enough (e.g. more than 5% weight).
  8. List e.g. top 10 models using rank correlation and discuss them with the experts.
$\endgroup$
3
  • 1
    $\begingroup$ Isn't this a general guide to predictive modelling? $\endgroup$ Dec 31, 2014 at 16:07
  • $\begingroup$ Though useful the steps are problematic. In "Reduce the list with the help of some expert, e.g. exclude all the predictors deemed to be irrelevant" some previously ignored but potentially important predictors can get ignored again. In "Discuss the poor predictors with the experts and possibly eliminate some," independence does not indicate conditional independence. For example, if Y and X1 has nonsignificant correlation, it does not suggest that Y and X1 also are uncorrelated when X2 is controlled for. "Analyse the predictors standalone" should be for description not inference. $\endgroup$
    – DrJerryTAO
    Nov 4, 2023 at 9:45
  • $\begingroup$ In "standardisation with all the possible combinations" it appears you mean variable selection, and you suggest that among 10 to 20 predictors screened by supposed "experts", 2 to 8 should remain in the model based on whether the estimated coefficients have the hypothesized signs and large magnitudes. This is quite problematic. Variable selection has been controversial. In practice, stepwise (usually based on AIC or BIC) and regularization (LASSO, ridge, and elastic net) are used. There is also multimodel inference, which can exhaustively examine all possible predictor inclusion scenarios. $\endgroup$
    – DrJerryTAO
    Nov 4, 2023 at 9:53
1
$\begingroup$

Bloomberg has a Default Risk model, which is similar to what you are querying. You can see a screenshot in this PDF. There you can also see the kind of variables they use.

You can access it by typing DRSK at the CDS screen is Bloomberg. (If the screenshot in the PDF is not clear enough, let me know and I can post one with better resolution from Bbg)

This model uses fundamental data, and obviously they have calibrated and backtested it; however it is a proprietary model, therefore you might have a hard time finding the details. You can try googling for it, there might be a white paper on it.

$\endgroup$
1
  • $\begingroup$ Is that the one supposedly in the 2nd page? It's not clear...Also where is the CDS screen in Bloomberg? Can't seem to find it. Googling "DRSK Bloomberg" doesn't help I think? $\endgroup$
    – BCLC
    Jan 4, 2015 at 18:08
0
$\begingroup$

Altman Z Score - http://en.wikipedia.org/wiki/Altman_Z-score - From Wikipedia - "The Z-score formula for predicting bankruptcy was published in 1968 by Edward I. Altman, who was, at the time, an Assistant Professor of Finance at New York University. The formula may be used to predict the probability that a firm will go into bankruptcy within two years. Z-scores are used to predict corporate defaults and an easy-to-calculate control measure for the financial distress status of companies in academic studies. The Z-score uses multiple corporate income and balance sheet values to measure the financial health of a company.

$\endgroup$
3
  • $\begingroup$ How do you know those ratios are applicable universally? $\endgroup$
    – BCLC
    Dec 31, 2014 at 13:41
  • $\begingroup$ The ratios are NOT applicable universally. The Z score is meant to work for industrial, non-financial companies. $\endgroup$
    – Bikenfly
    Jan 2, 2015 at 13:31
  • $\begingroup$ I think I'm not understanding this right. Sorry for confusion. Altman proposes a model that estimates default prob using certain ratios, cmiiw? How do we know that we are supposed to use those ratios (on those industrial companies)? $\endgroup$
    – BCLC
    Jan 2, 2015 at 13:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.