Expected Default Frequency

Anyone got any idea on how I can get the EDF’s for greater than 5 years Seems moodys only does up to 5. Maybe S&P or fitch have this, or there is some fancy mathematical trick to help calculate it?

You can calculate it mathematically using probability theory. For example, the 10-year EDF estimated from the 5-year EDF is 1-(1-EDF_5year)^2. In this case, you are assuming that the default probability for the 5-10y period is the same as in the 0-5y period. You could also looks at credit spreads and calculate the spread-implied default probability.

You could just assume the probability of default in each subsequent year is the same as the 1year default rate in year 5 and then rejigger the total calculation.

It would be better if you could derive the default rate from credit spreads/CDS, since might be a “term structure” for default rates. For instance, when GM was near bankruptcy, their CDS for 1-2 years traded much higher than 3-5 years. This means that the market thought that if GM could get through the next 1-2 years, their default risk would be much lower. So if you were to extrapolate the 2-year (shorter term) default risk to 5-years (longer term), you would get an inaccurate answer. Depending on the company, you might have some difficulty finding market data for corporate debt of more than 5 years. I’m not sure if many companies issue bonds of such long maturity. I don’t know. Maybe someone else has more information.