I'm interested in calculating the term and credit risk for a fixed income fund using the DFA Returns program. They provide the term factor and the US default premium in their regression tools, but discourage its use. Mainly, I'm interested in assigning a numerical value to individual funds for the credit risk (since duration is so widely available).
Is there a better way?
Thank you,
HM
Regressing Term and Credit Risk for Fixed Income
-
- Posts: 279
- Joined: Fri Oct 17, 2014 9:41 pm
-
- Posts: 16022
- Joined: Thu Feb 22, 2007 7:28 am
- Location: St Louis MO
Re: Regressing Term and Credit Risk for Fixed Income
http://www.portfoliovisualizer.com/factor-analysis
Very good and easy to use site for factor analysis, for equities particularly good --and ignore the dialogue box which limits data to 2004, if you enter the date yourself directly you can go back much further
Unfortunately the bond side only goes back to 2009,
Larry
Very good and easy to use site for factor analysis, for equities particularly good --and ignore the dialogue box which limits data to 2004, if you enter the date yourself directly you can go back much further
Unfortunately the bond side only goes back to 2009,
Larry
Re: Regressing Term and Credit Risk for Fixed Income
The PV site calculates the factors using Vanguard LT Gov/Corp bond funds, as I understand it. These funds have inception dates around 2009, that explains the available start date.
For annual returns you can copy the data from Simba's Excel spreadsheet or from http://www.portfoliovisualizer.com/hist ... ss-returns (same thing). DFA's matrix book or the Ibbotson SBBI book have annual data going back even longer - mid 1920s I think. This will only give you meaningful regression results for VERY old funds or index data though. Otherwise there are just not enough data points for reliable results.
For annual returns you can copy the data from Simba's Excel spreadsheet or from http://www.portfoliovisualizer.com/hist ... ss-returns (same thing). DFA's matrix book or the Ibbotson SBBI book have annual data going back even longer - mid 1920s I think. This will only give you meaningful regression results for VERY old funds or index data though. Otherwise there are just not enough data points for reliable results.
-
- Posts: 279
- Joined: Fri Oct 17, 2014 9:41 pm