Q's re Morningstar's Maturity Analyses of GNMA funds

[docneil88]

Hi all, I just looked at the "Average Effective Maturity"on Morningstar's Portfolio Summary page for Vanguard's GNMA fund, VFIIX , and Payden's, PYGNX . The numbers were 4.50 years and 6.35 years respectively. But then I looked at the "Bond Maturity Breakdown" at the bottom of the same page and found that VFIIX had 89% in 20-30 year maturities, 8% in "over 30 year" maturities, 3% in 15-20 year, and 1% in 10-15 year. PYGNX had 52% in 20-30 year, 36% in over 30 year, 5% in 15-20 year, 2% in 10-15 year, and 3% in 7-10 year. Here's Morningstar's definition "Average Effective Maturity" found at http://www.morningstar.com/InvGlossary/ ... at_is.aspx :

Average effective maturity is a weighted average of all the maturities of the bonds in a portfolio, computed by weighting each bond’s effective maturity by the market value of the security.

Average effective maturity takes into consideration all mortgage prepayments, puts, and adjustable coupons. (Because Morningstar uses fund company calculations for this figure and because different companies use varying interest-rate assumptions in determining call likelihood and timing, we ask that companies not adjust for call provisions.)

Longer-maturity funds are generally considered more interest-rate sensitive than their shorter counterparts.

A few sets of questions I'd appreciate some help on:

(1) Anyone care to speculate on why there is that apparently huge discrepancy between "Average Effective Maturity" and "Bond Maturity Breakdown"? In the maturity breakdown do they have to go by the longest dated mortgage in the each mortgage-pooled holding even though the rest of the mortages have earlier maturities? Or do they go with the average maturity of the pool? To what degree do the "Average Effective Maturity" numbers factor in extension risks (i.e. risks that if interest rates go up, people will hold onto their mortgages longer)? It seems odd that they say they factor in pre-payments without mentioning extension risk.

(2) I wasn't aware of there being any significant number of mortgages, much less GNMA guaranteed mortgages, that have maturities over 30 years. Are there? And if so, what are some of the banks or mortgage companies that offer them?

(3) In the 1970's and early 1980's interest rates went way up. Does anyone know any statistics on the degree of extension risk that manifested during that period or the effect extension risk had on average effective maturity?

[DaveS]

I don't really know the answers to the exact questions you posed. But I do know that mortgage bonds are affected by refinancing. That is to say that your typical mortgage bond pays off faster depending on whether the people who got the mortgage proceeds refinance or sell their houses. This negative convexity is a disadvantage to holding mortgage debt. In a market when rates go down people will refinance to get loans with a lower rate, meaning your bond with high rates will pay off faster - just when you don't want it to. In a market where rates are rising people won't refinance because doing so would get them a higher payment at the higher rates - just when you would hope they would pay you off so you could reinvest the proceeds at a higher rate. Right now rates are at historic lows with the most likely direction up at some unknown point in the future. At that point in time the duration of the bonds will extend, because people won't be refinancing. I admit I don't know the figures based on history, but I don't like the likely direction the future my bring. Dave