As a recap, Part
I dealt with bond price movements relative to interest rates. In Part
II I covered the subject of rising interest rates, and presented some data
on the 5 Year Treasury, which currently has duration of about 4.85:
- Mean: 0.89
- Standard Deviation: 0.81
- Max: 5.26
So what does this mean to a bond portfolio? Here are some observations:
- Assuming a normal distribution, in a period of rising interest rates roughly 95% of the time interest rates should not rise by more than 2.50%, meaning the fixed income portfolio value would fall by 12%, exclusive of any interest payments received. However, I am not sure a normal distribution is completely valid here (see bullet 3).
- The max rise in interest rates is 5.26; however, if purchased a year earlier the bond would have a yield of around 9.50% and a duration (note: higher interest payments = lower duration) of about 3.90. Thus, the loss that year would have been only about 11%.
- The data appear to cluster together and be prolonged. By this I mean periods of rising interest rates tend to be trending and thus prolonged. Further, larger increases in interest rates and smaller increases in interest rates appeared in the same batch (think early 1980’s and high inflation). In such a data set where observations are at the tails and not around the mean a normal distribution isn’t as useful. Further, the sample size is probably too small.
- Similarly, the large rises in interest rates came at a period of high inflation and already high interest rates. Right now we have low inflation and low interest rates. As such, data from that time period is probably not comparable.
- Most people do not have only Treasury Bonds. As a result, these calculations ignore credit risk (the risk a bond default), meaning if the aggregate credit risk of your bond portfolio rises then your bond portfolio will have additional losses; however, if it falls your portfolio will have gains.
- With regard to the above, current corporate credit spreads are elevated, but not substantially. As a result it’s possible credit risk could go either way. That said it is unlikely that interest rates and credit risk would rise at the same time. It’s more likely that credit risk would fall as interest rates rise and vice versa (i.e. credit spread narrowing and widening, respectively).
In my next post, the last in the series, I will try and put
it all together.
