Peer group comparisons can provide valuable insight into the composition and performance of your bank portfolio. With that thought in mind, the purpose of this column is to provide some current metrics and observations to use in reference to the investment account of your bank. Taken from our bond accounting database, the top 10 percent total return yield bank bond portfolios register 4.11 percent with a total return of 8.21 percent. The bottom 10 percent average yield is 2.25 percent, with a total return of 3.07 percent. These are significant differences in yield and total return for bond portfolios and can be explained by portfolio duration, composition and convexity (pre-payment or call optionality).
The average duration of the top 10 percent is 3.72. Duration is defined as the weighted average maturity of the cash flows of a bond. The duration of a zero-coupon bond is the same as its maturity, because there are no periodic cash flows. The size of the coupon plays a part in determining duration in that the higher the coupon, the shorter the duration. The duration of the bottom 10 percent is 2.90. This makes perfect sense, knowing that the yield curve has been positively sloped since June 2007. Many would assume that the shorter-duration bond portfolio would outperform the longer-duration bond portfolio in a rising interest rate environment. However, due to portfolio composition and convexity, this might not be necessarily true. Importantly, there are limits to prudent durations for bank bond portfolios; we most often see durations in the two to five year range, as dictated by investment policy guidelines. Short-duration bond portfolios suggest an implied interest rate forecast for higher rates, while longer-duration bond portfolios can be earnings motivated.
Concerning composition, the top 10 percent hold 33 percent in municipal bonds, general obligation (27 percent) and revenue (6 percent), whereas the bottom 10 percent hold 16 percent muni bonds, general obligation (12 percent) and revenue bonds (4 percent). Municipal bonds in general have higher-yield, longer duration and more call protection versus taxable alternatives. The lower group holds almost three times as many callable bonds as the higher-yielding bond portfolios, 21 percent versus 8 percent. The top 10 percent hold 38 percent of portfolio in mortgage-backed securities and collateralized mortgage obligations while the bottom 10 percent hold 15 percent in those categories. An inherent advantage of mortgage-backed securities is that they provide cash flow over time and do not re-price all at once, whereas callable bonds almost always are called in their entirety. Callable bonds will run to maturity (extend) when interest rates rise, locking in lower yields in periods of otherwise higher rates.
Mortgage-backed securities will continue to pay down in principal, albeit slower, when interest rates rise. It could be that the cash flow characteristics of longer-duration bond portfolios (due to MBS holdings) could perform as well or better than shorter-duration portfolios with larger callable bond holdings. Positively convex bullets will outperform in a falling interest rate period in terms of total return.
Bond portfolio performance in terms of average yield and total return depends on getting convexity right. As mentioned earlier, the lower-performing group holds a higher percentage of callable bonds versus the higher-performing group (21 percent versus 8 percent). Keep in mind, convexity has consequences. Non-callable bonds are positively convex, whereas callable and pre-payable bonds are negatively convex due to the optionality. Said another way, the higher-yielding bond portfolios are more positively convex; they hold less optionality.
The more predictable the series of cash flows over the term to maturity, the greater the premium paid by the market on a relative basis. Buyers will not pay meaningful premiums for bonds that will be called away or pre-pay quickly. The larger profit of less negatively convex or more positively convex accounts provides a cushion when interest rates rise. The bond portfolio is not likely to go from a 4 percent profit to an immediate 4 percent loss. Convexity, especially at this point in the rate cycle, is critical.
Currently, the top 10 percent yielding bond portfolios in the database have longer durations, more tax-free bonds, less callable agencies and hold a higher percentage of mortgage-backed bonds than do the bottom-yielding bond portfolios.
John D. Jones is senior vice president in the capital markets group of Country Club Bank, Kansas City.
Copyright (c) April 2012 by BankNews Media