Given the modest size of the UK IG corporate bond market, we believe a globally-focused and actively managed IG corporate portfolio and a careful portfolio construction can help UK DB schemes.
In an ideal world, a CDI portfolio would be void of currency risk and foreign interest rate risk. However, the relatively limited size of the U.K. IG corporate bond market makes it challenging to construct a purely domestic CDI portfolio, because:
1. A smaller opportunity set would likely lead to a less diversified portfolio, which in turn could increase potential default exposure.
2. The lower yield relative to the higher yield available in other countries.
3. More limited liquidity may translate into higher transaction costs and therefore reduce the amount of liability cash flows covered for a given capital commitment. It would also make future adjustments to the CDI portfolio more challenging.
4. A significant uptake of CDI strategies within the U.K. DB community could be disruptive, given the modest size of the U.K. IG corporate bond market (around GBP 500 billion) relative to the size of the U.K. DB pension industry (around GBP 2 trillion).
We therefore believe a global scope provides a more desirable foundation to build a CDI portfolio. We also believe the active selection of IG corporate bonds is more desirable as it can default-adjust the cash flows of a CDI portfolio at inception (e.g., using Moody's cumulative losses by rating).
To assess the efficiency of a passive CDI approach, we backtested the performance of a simple hypothetical default-adjusted, buy-and-hold investment grade (IG) CDI portfolio strategy initiated monthly between January 2004 and December 2008.
At the beginning of each month over the 2004 to 2008 period, we solve for the budget-minimising1 IG corporate bond portfolio2 (within the ICE Merrill Lynch USD Corporate bond universe3), which matches on a default-adjusted basis4 a specified 12-year annual liability stream (in USD). The portfolio is held over the duration of the 12-year liability stream, during which time liability cashflows are met primarily with portfolio cashflows. The portfolio is self-funded - i.e., a cashflow shortfall in any given year is funded by selling assets from the remaining portfolio on a pro rata basis. Equally, any surplus is reinvested.
We found that over the observation period (60 portfolios initiated between January 2004 and December 2008), on average portfolios experienced higher losses than were expected at the outset, equivalent to 18 basis points (bps) per annum, with a median of 16 bps. In over 25% of the samples, annualized losses exceeded approximately 27 bps.
While an average annual excess loss of approximately 20 bps may not seem significant, the potential impact this would have had on a scheme over this period is surprisingly large. The chart below shows the impact on the expected path to full funding.
We now contrast these results with a hypothetical actively managed CDI portfolio, with an assumed symmetrical annual outperformance of 20 bps versus the "Baseline". In this case, we found that the expected time to full funding was accelerated by three years, from 13 years to 10 years. In other words, the difference in the expected time to full funding between a passive CDI portfolio and an active portfolio with a modest alpha target (20 bps) can be highly significant: as much as seven years (17 years versus 10 years) in our example.
Given the findings from our analysis above, we believe a globally focused and actively managed IG corporate portfolio and careful portfolio construction can help U.K. DB schemes better plan and meet future liabilities.
1 We chose this simple approach because of its perceived appeal. However, we would recommend for various reasons a more careful portfolio construction.
2 Subject to maximum issuer concentration limit of 1.50% and sector concentration limits (40%)
3 Excluding callable and sinkable bonds
4 We apply the then prevailing Moody's rating-dependent cumulative average default rates. Specifically, we bootstrap an annual default intensity and enforce a flat forward intensity at the point where term structure of cumulative annual defaults becomes flat.