Though Ginnie Maes have been trading at the richer end in recent months, a report by UBS Warburg said that they have reached their richest levels, and would eventually cheapen up.
"We don't believe that GNMAs can richen further," said UBS. "So while GNMAs could remain rich' for the next few months, they are far more likely to cheapen over the medium term."
According to the report, GNMAs are currently at "the very richer end of the price and OAS ranges," which is seen most in the 6 and 6.5 coupons. Though 7s, 7.5s and 8.0s are now also in the fuller end of the price range, their particular ranges are comparatively much narrower than those in the lower coupons.
The factors that drive richness
UBS's report cited four reasons why GNMAs have achieved great levels of richness. However, the bank added that these factors "cannot drive GNMAs much richer, plus are apt to erode over the next few months anyway."
One reason cited by the report is the possibility of the Fed's inclusion of GNMAs in its System Open Market Account (SOMA) portfolio. The report said however that judging from the Fed's priority structure shown in its Jan. 1st meeting, this event would be a long way from happening.
And even if the Fed does include GNMAs in its portfolio, UBS said that this may not create the price distortion that MBS players are anticipating. Given the size of the Fed's portfolio compared to Fannie and Freddie combined, even if GNMA comprises a significant part of the Fed's portfolio, "it is unlikely to create demand even close to that contributed by Fannie and Freddie."
Furthermore, the country's diminishing surpluses will probably delay any action from the Fed on the GNMA front.
Another factor in the GNMA rise to richness is the refi boom. Typically, Ginnie Mae issuance declines during the height of refi activity. However, as we get to the tail end of the refi wave, UBS expects GNMA production share to rise. And as refi activity slows down, production tapers off as well. Conventionals usually experience a greater decline in production compared to GNMAs, thus "GNMAs will become a larger part of new issuance over the next few months, alleviating the temporary supply and demand imbalance," said UBS.
The other factors that have driven GNMA richness - adverse selection in the conventional market and the collapsing conventional rolls - are expected to wane as well. UBS said that with "conventional rolls down at carry, they have little room to collapse further."
"We think that GNMAs could richen a little more, even though on fundamentals they are fully priced, " said Art Frank, director of fixed-income research at Nomura Securities.
This is mainly because Ginnie production has been low in recent months.
For instance, since GNMA 6s have seen very little production in 2001, if you do a large TBA trade in GNMA 6s, the chances of getting 1999 production for a substantial portion of the pool is very good. This is because since you are getting more seasoned production, the chances for housing turnover is increased. And since 6s are a discount coupon, the possibility of faster prepays are a good thing.
"The GNMA-FNMA swap in 6s looks rich but is misleading because in a Fannie 6 trade you are most certain to get 2001 production," said Frank. "Whereas in a GNMA 6 trade you are probably going to get some 1999 production."
The supply in terms of the higher coupons 6.5s, 7s, 7.5s have also been a bit sparse. These coupons are trading around a half a point, 14-16 ticks up from Fannies.
"Even though on fundamentals GNMAs are on the rich side, we think they can get a few 32 nds richer before correcting so we would not short GNMAs at this point," said Frank.
Merrill says shift to GNMA IIs
In a conference call held last week, analysts from Merrill Lynch suggested that it would be a good trade idea to go from GNMA I discounts into GNMA IIs, because both GNMA II 6s and 6.5s are cheap relative to GNMA Is. Merrill believes that this is better than shorting on GNMAs, since there is such a dearth in GNMA supply.
"I think a better idea would be to go from GNMA I discounts into GNMA IIs, both GNMA II 6s and 6.5s are cheap relative to GNMA Is," said Akiva Dickstein, a director at Merrill. "And in terms of supply, you have actually seen Ginnie IIs start being produced significantly less than GNMAs Is probably because the prices have diverged so much. So we think that's probably a good trade in the Ginnie Mae discount arena."
GNMA Is are fundamentally more expensive than GNMA IIs because while they have a lower weighted average coupon, and have a shorter delay (GNMA Is have a fourteen day delay, while GNMA IIs have a nineteen day delay), the current pay-ups are excessive.