If defined-benefit pension plan sponsors are looking for arespite from the extreme fluctuations in funded status seen overthe last few months of 2018, they probably shouldn't hold theirbreath. After the fourth-quarter selloff, money managers seemresigned to the idea that financial markets will remain volatilefor some time. The transition back to trend-like growth in theUnited States after a year of stimulus was never likely to besmooth sailing, but the recent weakness in China, concerns overtrade relations, and the ongoing tightening by the Federal Reservehave made for an extremely uncertain outlook.

Fed Chairman Powell has managed to engineer a recovery frommarkets' December lows by suggesting the Fed will likely hold rates steady for much ofthe year. The question now becomes how long the rally mightlast.

Few plan sponsors expect valuations to jump all the way up towhere they were at the end of the third quarter, when the averagecorporate defined-benefit plan's funded status hitnearly 92 percent. Frankly, the overriding sentiment among manystrategists and managers of risk is that while there may be atactical opportunity to base investment decisions on the Fed's“prolonged pause” and the possibility of a U.S.-China trade detenteto come, rallies are there to be sold, not bought, as the businesscycle advances in age.

That means volatility will likely remain a fact of life fordefined-benefit plan sponsors. In the fourth quarter alone,plans' average funded status dropped by 8percent, which erased the entirety of the gains made in2018.

As funded status recovers, it isworth reflecting on how quickly markets can move in adverse ways.Indeed, the Q4/2018 hit to funded status could have been much worsehad yields rallied (declined) more. Most of the damage to fundedstatus came from the drop in equity prices. The effects weredampened by corporate yields inching higher because spreads widenedslightly more than Treasury yields dropped. This increasedliability discount rates for plan sponsors. In a more severerisk-off event, where the Fed was forced to cut the policy rate,longer-dated corporate yields might fall, causing the present valueof plan liabilities to increase.

If anything, the volatility of the fourth quarter reiteratedthat there is a strong case for a disciplined approach to pensionplan de-risking. But those decisions should not be driven solely bythe interplay between equities and yields. As the credit cycleenters its later stages, issuer selection and avoiding downgradesare set to become much more important. Corporate plan sponsors whodecide to adopt a liability-driven investment (LDI) strategy willneed to exhibit discipline in how they implement and conduct thatplan.

Explosion of Downgrades in Q4

The nature of the shareholder-friendly behavior going on duringthis credit cycle—i.e., debt-funded share buybacks, dividendincreases, and mergers and acquisitions (M&A)—is that it tendsto be driven by the objective of maintaining investment-graderatings. However, downgrades are still a concern. In the fourthquarter of 2018 alone, more than 10 percent of the universe of longbonds with a rating of A or better were downgraded to BBB in theUnited States—with General Electric, Anheuser-Busch, Altria,Lowe's, AstraZeneca, and PG&E among them. In fact, well overhalf of the investment grade index-eligible universe are now ratedBBB.

The impact of this trend on companies that remain A or bettershould not be overlooked, particularly from the perspective of aliability hedger. First off is the concentration risk that stemsfrom lack of diversity in corporate bond indices. Back when FAS 87 came into effect in the mid-1980s,AA or higher bonds comprised nearly 50 percent of theinvestment-grade bond market. Today, AA or higher corporate bondscomprise less than 20 percent. In fact, there aren't even enough AAcorporate bonds outstanding (around $1.1 trillion, according toBarclays) to cover the value of pension liabilities outstandingtoday (around $3.1 trillion). Figure 1 shows a summary of thecorporate bond market, split by full maturities and long-duration(10+ year) maturities.

The multi-decade downward migration of higher-rated creditsmeans that companies which are able to avoid being downgraded whileengaging in mergers and acquisitions (M&A) or debt-fundedbuybacks are becoming an ever-larger presence in the higher-qualitycorporate indices. Among long-duration AA-or-higher bond issuers,for instance, the top 10 issuers make up more than 50 percent ofthe market, and the top five encompass just over 40 percent of themarket.

In other words, there is a tremendous amount of issuerconcentration. Take the example of Microsoft: The technologycompany did not issue any debt in 2018, yet it makes up 13 percentof the long-duration AA-or-higher index. Adding A-rated creditsinto the investment mandate does not completely eliminate theproblem. For instance, since Comcast purchased Sky in the U.K. lastyear, its debt now accounts for 5.25 percent of all U.S. longcorporate A-or-better debt.

The string of downgrades that major U.S. corporates experiencedin the fourth quarter of 2018 illustrates how more-concentratedindices can lead to disruptive outcomes when rating agencies decideto act on larger issuers. General Electrics' loss of its A ratingcaused its benchmark long bonds to widen 125 basis points (bps)from the beginning of October to year-end, while A-rated long bondsin general widened just 30 bps over the same time period. Likewise,bonds from Anheuser-Busch and Lowe's widened by 80 bps to 85bps.

This is the second problem with the Q4/2018 downgrades: Thescarcity of debt rated A or better leads to overvalued prices thatare susceptible to large corrections upon a downgrade. It'sEconomics 101. If demand remains constant while the supply ofhigher-rated credit decreases, then prices will adjust higher.Reality is not that simple, of course—demand is never constant fromyear to year—yet the basic narrative is correct. The universe ofhigher-quality investment-grade bonds is not growing nearly asquickly as BBB-rated bonds. At the same time, buyers based in Asia,who have become a very important source of demand, continue toprefer higher-quality debt. The result is that bonds rated A orhigher tend to trade at ever tighter spreads.

For liability-driven investors, the pricing effects ofdowngrades of formerly A-rated debt are a real concern.Historically, when an A-rated company was downgraded to BBB, thespread of the company's bonds vs. U.S. Treasuries tended to widenan average of 25 bps to 30 bps. Yet the current environment isresulting in larger transition costs; downgrades are resulting inspread widening of 50 bps to 100 bps. Active management thatresults in underweighting these companies beforehand can reduce theimpact, but at the end of the day, downgrades lead to negativepricing performance.

More insidious for pension plans, downgrades can have asubstantial effect on discount curves. When dealing withdefined-benefit plans in the United States, there two principaldiscount rate methodologies that plan sponsors focus on in theirannual calculations of liabilities. One they use to discountliabilities for accounting purposes, and the other they use indetermining their required contributions to the plan under theEmployee Retirement Income Security Act (ERISA). Discounting foraccounting purposes requires liabilities to be discounted using“high-quality corporate bond yields,” which has historically beentaken to mean bonds rated AA or higher. Discounting to determinecontribution requirements is explicitly based on bonds rated A orhigher.

Thus, when a large issuer with an A or AA rating is downgraded,that change can increase liability valuations, which in turn lowersfunded status. Downgraded bonds almost always leave the A or AAindex at a wider spread than the bonds that remain in the index, sotheir removal reduces the spread and average yield of the index.For plan sponsors, this increases the value of the plan'sliabilities while the value of assets that are hedging thoseliabilities doesn't change.

When larger issuers are downgraded, this type of“uninvestability drag” can really add up. Generally,uninvestability introduces a drag on pension funding ratios ofabout 1 percent annually. This is a key factor supporting the useof active management when investing in corporate bonds.

The Need for Customized Strategies

The volatility in funded status that can result from downgradesis meaningful, but failing to properly hedge the plan's liabilitiescan create even more volatility. As such, it makes sense for plansponsors to continue seeking ways to lock in gains in funded statuswhile looking for ways to minimize downgrade risk in theirfixed-income portfolios.

One counterintuitive—but increasingly accepted—approach is toadd BBB-rated companies into the defined-benefit plan'sfixed-income allocation. Because BBB issuers make up such a largeproportion of the universe of investment-grade credits, investingin BBB-rated debt has large enough benefits to issuer and industrydiversity, trading liquidity, and the ability to create alpha thatthey appear to outweigh the increased risk of holding lower-rateddebt. This is particularly true when the plan's investment mandateis paired with a portfolio of government securities, to improve theweighted-average rating.

Arguably, adding BBB-rated securities to fixed-income portfoliosis just an exercise in replacing one risk with another. Clearly,BBB-rated debt is at higher risk of a downgrade to high-yield,which tends to be even more punitive than the shift from A to BBB.However, many BBB-rated companies are actively looking todeleverage after issuing debt to fund acquisitions. If they fail toreduce debt in the coming years, these companies may be at risk ofa downgrade, a situation that closely aligns the incentives of themanagement team and the bondholders. In other words, disruptive BBBdowngrades could become an even larger problem than A-rateddowngrades, but the risk and reward are well-balanced. Activemanagers who invest in BBB-rated bonds can create value byselecting winners and avoiding losers.

Another approach to mitigating downgrade risk in a plan'sinvestment portfolio is to limit the volume of debt the plan holdsfrom any single issuer. Whether the plan buys only A-or-higherdebt, or it has a full-spectrum portfolio, capping exposure to anyindividual company at 1 percent or 2 percent can help avoid overlyconcentrated exposure that would be harmful to performance if thatcompany were downgraded.

LDI and the Case of Mismatched Liabilities

Consider the following case study of changes to a hypotheticalplan's funded status in Q4/2018. This simple example demonstrateshow ineffective a plan's hedging may prove to be during a volatileperiod, even if the hedging looks effective on paper.

A company has a $1 billion pension plan that is 100 percentfunded as of September 30, 2018. Half of its assets are allocatedto return-seeking assets, which it benchmarks against the MSCI AllCountry World Index. An LDI approach has convinced plandecision-makers to invest the other half of its assets inliability-hedging securities, which the plan sponsor benchmarksagainst the Bloomberg Barclays Long-Duration U.S. Credit. (SeeFigure 2.)

This is could be considered a “typical” pension plan. Itsliability discount rate is based on AA bond yields, with aliability duration of 12.5 years. Meanwhile, the assets that theplan purchases to offset those liabilities, as represented by theLong-Duration U.S. Credit index, span a range of investment-gradecredit qualities and have a duration of roughly 13.4 years. Theduration differential indicates that the sensitivity between priceand yield is noticeably higher for the assets than for theliabilities. Nevertheless, the plan's hedge ratio is 54percent—($500 million LDI assets x 13.4-year duration) / ($1billion plan liabilities x 12.5 year duration)—which means that theplan sponsor expects, all else being equal, the assets to return$54 for every $100 in liabilities.

Unfortunately, over the course of Q4/2018, market volatilitypicks up mightily. The MCSI All Country World Index loses 12.68percent in the quarter, while the Bloomberg BarclaysLong-Duration U.S. Credit loses 1.64 percent. The plan'sliabilities decline slightly; discount rates remain flat throughthe quarter, so the only significant change to liabilities is aslight reduction as retirees receive benefit payments. At the sametime, however, both of the plan's asset portfolios—both thereturn-seeking assets and liability-hedging assets—decline fasterthan the liabilities do. (See Figure 3.)

Although the plan appeared to bewell-hedged ex ante, the difference in duration between its assetsand its liabilities means its funding ratio falls by 8 percentagepoints in Q4/2018 due to market movement, even as liabilitiesremain, essentially, constant. While this outcome would not besurprising if 50 percent of the plan's assets were equities, thedecline in liability-hedging assets—compared with what wasexpected—illustrates what can occur when plans use assets of onecredit quality to hedge a liability that is a fundamentallydifferent credit quality.

We can illustrate this point by looking at the effect of Q4/2018on yields for the plan's liabilities and the assets intended tohedge those liabilities. Corporate bond yields (and, in turn,liability discount rates) can be broken down into an interest ratecomponent—i.e., Treasury yields—and a credit spread component. OverQ4/2018, liability discount rates barely changed, as increasingcredit spreads on the AA bonds used to calculate the discount ratewere offset by declining Treasury yields. However, because spreadsfor bonds in the Long-Duration U.S. Credit index widenedconsiderably more than Treasury yields declined, yields on theplan's liability-hedging assets increased by 21 bps. In fact, therate at which credit spreads for the LDI bonds widened was morethan double the rate at which the liability discount rate's AAcredit spread component grew. Because yields onbroader-investment-grade debt are so much more volatile than arethe AA bonds used to calculate the plan's liabilities, this hedgeturns out to be quite ineffective during a period of market stressand downgrades, such as Q4/2018, no matter how good the hedge lookson paper. (See Figure 4.)

At a broader level, the fact that the credit spread component ischanging faster for the assets illustrates the volatilitydifferential between how the full market of investment-gradecorporate debt behaves versus higher-quality bonds. Although manyplans utilize all levels of investment-grade debt in theirliability hedging, this volatility differential suggests it mightbe a good idea to take a closer look at how the LDI strategy isconstructed.

A More Nuanced Approach

An alternative to just utilizing long-duration credit isutilizing a duration-matched LDI strategy that blends corporatedebt with Treasury/agency allocations. Such a strategy might enableplan sponsors to better align the credit quality of the LDIstrategy to that of the liability.

For example, Figure 5 shows a custom-designed allocation for aplan that blends four Barclays indices both for duration andappropriate sizing of the allocation to credit, to match thevolatility of the liability discount rate.

This blend aligns to a 75 percent allocation to investment-gradecorporate debt and a 25 percent allocation to Treasuries. Now,let's see how this portfolio performs in Q4/2018 relative to theLong-Duration U.S. Credit index alone. To produce anapples-to-apples comparison, one needs to either normalize for theex ante hedge ratio or else allocate the same amount of money toeach strategy and accept the hedge ratio differential. In thiscase, either approach delivers the same conclusion, so in theinterest of simplicity, we'll assume that each strategy isallocated $500 million.

We already know that the Long-Duration U.S. Credit aloneproduces a 54 percent hedge ratio. Because the point of thecustom-portfolio strategy is to match asset duration to liabilityduration, we expect the custom portfolio to produce a 50 percenthedge ratio. Figure 6 illustrates the outcome of each strategy overthe quarter.

In this case, a 50 percent allocation to Long-Duration U.S.Credit, which covers the full investment-grade market, produces ahedge that is more than 100 percent effective, while customizingthe strategy to the liabilities provides an outcome much closer tothe 54 percent hedge ratio expectation. While a simplistic example,it shows that the construction of the LDI strategy can improve thetracking to the liability and mitigate the impact to the fundingratio during a volatile period.

LDI is not a one-size-fits-all strategy, but rather a nuancedimplementation that should be designed to meet the specificobjectives of the individual plan sponsor. Over 2019, if the marketforecasts hold, plan sponsors will likely learn not only about theamount of hedging their liabilities require, but also about thedesirable composition of the plan's asset portfolio. We recommend athorough review prior to any further market turmoil.

Jodan Ledford ishead of client solutions and multi-asset for Legal & GeneralInvestment Management America (LGIMA). He joined the company in2013 as head of U.S. solutions. In 2017, Ledford's role expanded toinclude oversight over each of the distribution, clientrelationships, solutions strategy, product, and multi-assetportfolio management areas. Prior to joining LGIMA, he wasexecutive director and business head of asset liability investmentsolutions at UBS Global Asset Management.

Jason Shoup ishead of global credit strategy for Legal & GeneralInvestment Management America. He joined in 2015 as seniorportfolio manager and fixed income strategist. In 2018, he wasappointed head of global credit strategy. In this role, he isresponsible for leading the global strategy process and workingclosely with the Chicago and London offices for both active fixedincome and global high yield. Prior to joining LGIMA's Fixed IncomePortfolio Management team, Jason spent 10 years at Citigroup, wherehis most recent position was director, head of U.S. high-gradecredit strategy, where he advised institutional clients andpublished credit research reports.