Standard & Poor's recently announced a revised methodology to calculate the convexity risk in the mortgage-backed portfolios of life insurance companies.
Though in the near term this would have a minimal impact on the amount of MBS holdings by these firms, it is still a step in the right direction, analysts said.
"For the foreseeable future I do not see any significant change in life companies' appetite for MBS," wrote Linda Lowell, an MBS analyst from RBS Greenwich Capital, in a recent report.
She explained that there is anecdotal evidence that many life companies increased MBS allocations in response to the poor performance in their corporate bond portfolios. However, with the improving outlook in some credit product sectors, as well as the partial rebound of corporate spreads, managers have been enticed to go back to corporates. Simultaneously, extension risk in the mortgage market has been amplified by the extended foray into low rates and the drawn-out prepayment wave, necessitating insurance company investors to factor extension risk into their asset allocation decisions.
"Net, net I would expect life companies simply maintained a near 20% share of total MBS outstanding over 2002," added Lowell. "However, in the event that the economy picks up, the interest rate cycle turns and the mortgage market returns to a more symmetrical risk profile, I would expect to see judicious and substantial increases in life company allocations into mortgages."
In a previous report published back in May 2002 entitled Does Outmoded Thinking About MBS Influence Insurer Ratings? Lowell criticized S&P's old method of recognizing the option risk in MBS (dubbed the S&P 300 test). She said there was no better time for the rating agency to reassess this interest rate risk "model" for insurers than at the time the article was written -when the investment performance of insurers was undermined by historically high levels of corporate defaults, downgrades and headline risk, and when MBS, which has been tested in volatile rate environments, provided a good alternative to corporates.
The S&P test, which was adopted in 1995, was implemented during a precarious time in the mortgage-backed market. During this period, advances in CMO structuring prompted the influx of new investors who were attracted by the generous base-case yields but who were also less familiar with the obstacles to achieving those yields, according to Lowell. At the same time, accounting and other conventions allowed buysiders to book, advertise and live off the great expectations that were captured in those base case yields. Thus, there was a potential for abuse.
However, since then, Lowell explained that accounting practices have been reformed, which has led most financial institutions to carry their MBS at market value. Aside from this, mortgage-backeds are currently the largest component of high-grade bond indices. "Pass-throughs enjoy an incomparably deep market and liquidity comparable to on-the-run Treasurys," wrote Lowell. "As a result, professional fixed-income investors can no longer afford to bypass the sector. Accordingly, mortgage sophistication is now widespread."
Clearly, it has become evident that these factors have rendered S&P's old methodology "outmoded," and perhaps unnecessarily hindering insurance company appetite for MBS. The numbers are telling.
In the report, Lowell mentioned that according to aggregate holdings data reported by insurers to the American Council of Life Insurance (ACLI), insurers actually reduced their concentration of MBS from a peak of 24.5% in 1996 down to 19.7% by 2000.
Out with the old
in with the new
S&P's previous way of measuring convexity against a benchmark non-callable corporate portfolio has been changed to more accurately reflect the risk in negative convexed assets such as MBS. The rating agency has adopted a part of its financial product capital model (FPC) as the new methodology to calculate the convexity risk when applying its risk-based adequacy model to the life insurance industry.
This new method, which is effective immediately, will be applied not only to MBS portfolios (including passthroughs and CMOs) but also to callable corporate bonds and related hedge instruments.
Under the old method, the rating agency created a synthetic portfolio from a group of A' rated non-callable corporate bonds. This portfolio was duration-matched to the effective duration of the company's MBS portfolio. This synthetic A' asset was then priced based on the same parallel shifts in the yield curve, which were usually plus 300 basis points and minus 150 basis points. The results were compared to the MBS portfolio at the same levels to get the level of capital necessary.
This means the market value of the MBS portfolio was calculated at year-end for a plus-300 basis points parallel shift in interest rates, which was then subtracted from the corresponding market value of the synthetic non-callable A' corporate bond for the same scenario. This same equation was calculated for minus 150 basis points. The resulting greater number was used for the capital charge.
The new rule will allow insurance companies to specify the DV01 (dollar value of a basis point) of their mortgage-backed portfolio. They will also indicate the market value of their mortgage portfolio and the optional hedges in the base case. The interest rate scenarios that were suggested are down 125 basis points, down 100 basis points, up 100 basis points, up 150 basis points, up 200 basis points and up 250 basis points. Assuming these shifts in rates, insurance companies are supposed to specify the expected market value of their mortgage portfolio and the associated hedges. This would suggest the underperformance or the outperformance of the combined portfolio during incremental shifts in rates, relative to their original duration measure. The size of the shifts used in the calculation can change from year to year, depending on year-end curves and applied interest rate volatility.
In other words, the capital charge would actually be the cumulative underperformance under the down rate or the up rate scenarios relative to the original duration of the mortgage portfolio.
Analysts from JPMorgan Securities said that given the negative convexity of an MBS portfolio, capital charges under the revised methodology would not be significantly different from the old one.
They stated that under the old convexity test, insurance companies were able to buy deep-out-of-the-money options that kicked in under extreme-interest-rate environments and therefore minimized the convexity-related capital charge. But by specifying the incremental changes in rates under the new test, it actually becomes harder to game the system by buying deep out-of-the-money options.
"As we have suggested in the past, using a static duration to arrive at potential convexity charge on MBS is too onerous," wrote researchers. "Investors in MBS understand the duration drift on mortgages with shifts in interest rates." They added that insurance companies with adequate monitoring systems should be allowed to reconsider the duration of their mortgage portfolios.
RBS Greenwich's Lowell said that this is a very positive step for life companies and the rating agency. It takes into consideration that insurance companies purchasing MBS actually use advanced prepayment and evaluative techniques to buy, hedge and manage assets and are able to anticipate the behavior of liabilities.
She said that more importantly, the new method does not really impose an extreme extension risk scenario. Companies she had spoken with could accommodate an up 250 basis points shift without having to make difficult changes in their asset allocation strategies.
However, she acknowledges that the changes do not go all the way to remove some of the more arbitrary features of the risk-based capital computation for mortgage convexity.
S&P analysts said that they have opened the communication lines with the insurance industry and the asset managers in terms of revising S&P's methodology to calculate the convexity risk on their portfolios. They also look at other quantitative and qualitative factors that affect their views on the company's capitalization strength and the quality of the company's risk management. "We are not trying to limit the investment allocation in mortgage-backed securities or callable corporate bonds for life insurance companies," stated Jose Siberon , a director at S&P. "What we really are trying to capture is the convexity risk in these asset classes." He said, however, that they are continually discussing matters with people in the industry to see what improvements could be made to the calculation of the convexity risk in MBS. In this light, JPMorgan and S&P will be holding a webcast on the issue on Monday, Feb. 24 at 1 P.M.
Copyright 2003 Thomson Media Inc. All Rights Reserved.