I was disappointed to see Pensions & Investments inadvertently giving credence to two of the myths of stable value investing in your Sept. 1, page 38 article about Sanford C. Bernstein & Co.'s entrance into stable value management.
Contrary to your article, the objective facts do not support the statement that active bond management outperforms guaranteed investment contract returns, and expenses of GIC funds are actually lower, not higher, than those of actively managed bond funds.
Diversification and investment liquidity are two strengths of synthetics, but there is likely to be a significant performance price to be paid for these attributes. This price can be observed both in making spot yield comparisons of the most comparable bond and GIC and can also be observed over time in comparing the performance of the Ryan GIC indexes with comparable bond indexes over relevant time frames.
The Sept. 8 issue of Barron's included the summary of net of insurance company expense quotes to FCM's traditional GIC Rate Desk for the week ended Sept. 5. The high three-year, simple-interest $5 million quote was 6.54%, for five years it was 6.79%, and for seven years, 6.92%. In each case these quotes were from AA rated issuers.
The Sept. 5 weekly edition of Salomon Brothers Bond Market Roundup quoted the comparable AA rated new issue bond yields as 6.37% for three years, 6.54% for five years, and 6.67% for seven years. The premiums offered by the GIC were 17, 25 and 25 basis points respectively for three-, five- and seven-year investments.
In 10 years making these observations, I have never observed the comparable bond to offer a premium. Last fall, I looked over the previous 36 months comparing rates in the last week of every month and found the average GIC premium over that time period was as much as 30 basis points.
Active bond portfolios will not always benefit from the capital gains generated by falling interest rates, just as they will not always be hampered by the capital losses produced by rising interest rates. Eventually, the capital gains and losses will cancel each other out and the active bond portfolio will be left with an income return plus or minus the value added or lost by the manager's decisions.
So in making a fair asset class comparison between bonds and GICs, it is necessary to pick a time frame in which interest rates ended up approximately the same place where they started, because GICs are not subject to capital gains and losses.
The 10 years ended March 31, 1997, was just such a time period. Although the three-year Treasury did decline somewhat from 6.91% to 6.58% and the five-year from 7.16% to 6.76%, the capital gains generated for bond portfolios by such a decline were not dramatic. But the comparison between the 10-year return of the Ryan G5 traditional GIC index with the spectrum of bond indexes is an eye opener.
As reported in the April 29 issue of DC Plan Investing, with an annualized return of 8.64%, the 21/2-year duration Ryan G5 beat every fixed-income index illustrated over the 10-year period, including the considerably longer duration Salomon BIG and the Lehman Government/Corporate. This was so despite the small capital gain from which these bond indexes benefited as well as being further out on the yield curve.
As far as fees are concerned, the return comparisons above already are net of insurance company and bond underwriting expenses, and so the only valid comparison should be limited to additional portfolio management and wrap fees. Separate account GIC fees are about 0.20% to 0.25% or less for larger accounts. Most active bond managers start at 0.35% to 0.40%. Synthetic wrap fees on an actively managed bond portfolio are about 0.12% to 0.15%. So in total, active bond management synthetic GIC fees take about 0.47% to 0.55% off a performance already likely to be lower than that of a traditional GIC portfolio.