Amid the recent rally, the mortgage market is wondering when or if convexity hedgers will re-enter the flow. Most participants agree that convexity-hedging flows will probably be back soon, though nowhere near the levels seen last summer.
"For most of us, the term convexity hedgers brings back painful memories of the carnage during the summer," said a recent Lehman Brothers report. "We don't expect to see that anytime soon. However, that doesn't mean that convexity flows will be absent from the market."
After some time, call risk has once again become an important issue for the mortgage market - which is particularly significant given that before the rally, the market was geared toward hedging a downtrade.
Although risks from origination pipelines are still slight, mortgage portfolios are at their worst convexity point, analysts said. Furthermore, mortgage servicers heighten the risks from a rally. Lehman noted that not only are asset prices higher, fewer offsets from the origination side should cause a heavier dependence on the capital markets for hedging. Analysts expect mortgage convexity hedging to begin at about 3.9% on 10-year Treasurys. This should improve the market bid or duration/spread products, they said.
The convexity danger zone
Citigroup Global Markets considers 3.9% the start of the "convexity danger zone." The mortgage and swap markets could be affected significantly if the long end of the Treasury curve rallies further, Citigroup analysts said.
"The strong bond market rally thus far in 2004 has caused mortgage durations to shorten considerably, and, if sustained, could lead to a reemergence of the fabled convexity trade,'" wrote analysts. The effective duration of Citigroup's mortgage index was 3.42 years as of Jan.1. This dropped to 3.1 years on Jan. 13, and to 3.05 years on Jan 15.
The change in effective duration is the important factor, wrote analysts, and not the starting level. They note that the duration change experienced is equivalent to $205 billion in 10-year Treasurys.
In its report on the recent rally, Merrill Lynch estimated that the duration of the mortgage universe was reduced by about $212 billion in 10-year equivalents (as of Jan. 16). Merrill concurs with other firms that some delta hedging related to the recent move in rates has occurred, but nothing in the vicinity of last summer.
The firm concluded that the recent rally has increased the mortgage market's negative convexity, with both the large 5% and 5.5% coupon becoming more negatively convex. Contraction risk is likely to be greater now compared to when rates were about 50 basis points higher ($344 billion versus $316 billion, as of Jan 16). The implication: Mortgage hedging would still be a factor if the market were to rally further.
Extension potential has also increased - up from $237 billion to $316 billion as of Jan 16 - bringing it to the level of contraction risk in the market.
"This tells us that once the mortgage hedgers purchase the duration needed given the rally, a backup could force the selling of what would then be extra duration," wrote Merrill analysts. The convexity might still be more of a risk in a rally; however, convexity could exacerbate the movements in either direction.
The market's hedging needs
In a separate report, UBS presented a methodology for determining the relative importance of each of the hedging needs of the different mortgage participants - investors, servicers and originators. Current convexity hedging needs are significantly less compared to last July, analysts concluded in the article.
UBS stated that convexity hedging should only have a minimal impact at these rate levels. "While the convexity level of the mortgage index is high, both originators and servicers have muted hedging needs at this juncture," analysts wrote.
UBS added that convexity hedgers might have to make up large amounts of duration if rates move either way, but do not need additional hedging if the market trades in a narrow range, citing the mid-2002 experience. During that time, even with the convexity needs of the mortgage market comparatively high, there was really no "catalyst" that might have ignited a large rate move.
There are no "magic triggers" that will inspire a deluge of convexity hedging, UBS argues. Though the convexity needs of the market are measurable, significant volumes of such does not necessarily mean a considerable amount of buying and selling is going to happen. In other words, "there has to be a catalyst triggering a market move, which could then be exacerbated by the convexity hedgers," said analysts.
Breaking through 4%
Although many investors remain bearish or may have recovered their shorts in the market, the vast majority of the market is still discounting the effects of the recent rally, as mortgages remain far too rich.
"Many investors who would normally be rebalancing their durations based on the recent move down in rates and uptick in prepayments are maybe holding off," said an MBS analyst with a primary dealer. "However, if the market went through the 4% mark on the 10-year Treasury decisively, you would see something of a capitulation trade." This would probably cause rates to move down to the 3.85% to 3.75% level, he said.
No matter, convexity hedging would not be as rampant today compared to what the market saw last summer. For one, Street inventories were significantly larger back then. As of Jan 7, for instance, Bloomberg data showed that the MBS inventories were only up to roughly $19 billion, compared with approximately $65 billion in July.
"In the summer, you had a tremendous whipsaw effect where many investors were long mortgages and were trading them extremely short, and all of a sudden rates were backing up. I don't think you have that now," said the analyst. However, if the market moves decisively through the 4% mark on the 10-year, expect some portfolio rebalancing. This would be worth about $40 billion in 10-year equivalents for every 10 basis points that the market moves, he said.
In terms of prepayments, moving through the 3.85% level on the 10-year would bring a significant amount of the mortgage market in the money. In this scenario, the market would have to adjust to a whole new prepayment environment, especially because at this level, mortgages tend to lag badly and prices of MBS don't keep up. This would compel the market to adjust back to mortgage valuations reflecting a Refinancing Index in the area of 7,000 and speeds reflective of last spring's prepayments.
Copyright 2004 Thomson Media Inc. All Rights Reserved.