An earlier Towers Watson analysis1 of funding results for defined benefit (DB) plans sponsored by the Towers Watson Pension 100 (TW Pension 100)2 showed that funded status climbed significantly higher from year-end 2012 to year-end 2013. Over the analysis period, the projected benefit obligation (PBO) declined due to rising interest rates, while investment returns and plan contributions gave assets a boost.

Our analysis also found that during 2013, pension plans with large allocations to equity had higher absolute investment returns than those with larger concentrations of fixed-income investments. However, when pension funded status — a comparison of asset values to plan liabilities — is measured over the full six-year analysis period, the outcomes look very different.

Over the last several years, some plan sponsors have been gradually paring back their equity allocations in favor of fixed-income and alternative asset classes, with the growth of the former linked to a pension investment strategy focused on matching the value of plan liabilities rather than pursuing absolute returns.3

This analysis looks at how sponsors’ pension investment strategies have affected DB plan funding since the start of the financial crisis in 2008. Where applicable, all historical values shown are for companies in the 2013 TW Pension 100.

Pension investment approaches and funding volatility

In 2013’s booming equity market, pension plans with large allocations to stocks performed very well. Returns were much lower for pension sponsors with large fixed-income holdings — many of whom were somewhere in the process of de-risking their plans. Nevertheless, overall funded status improved in 2013.

Between 2008 and 2013 — a tumultuous time in the financial markets — companies that had already shifted to a majority bond approach outperformed stock-heavy investors in terms of both absolute returns and performance relative to plan obligations. These sponsors also had lower plan contributions over the six-year period. 

We looked at 90 companies in the TW Pension 100 that reported target asset allocation data from 2008 to 2013. To better understand how their asset strategies affected their plan funding, we categorized these pension sponsors into three groups and then tracked their results.

Group 1 (which is the vast majority of sponsors in this analysis) consists of plan sponsors that held less than 50% of plan assets in fixed-income investments from 2008 through 2013. Group 2 held more than 50% of assets in fixed income over the period, while Group 3 allocated less than 50% of assets to fixed income in 2008 but held more than 50% in fixed income by 2013. We assume that Groups 2 and 3 either had a liability-driven investment (LDI) strategy in place by the beginning of the analysis period or have since adopted one.4

LDI strategies typically use fixed-income assets as a hedge against interest-rate-driven movements in plan liabilities. In years when long-term, high-quality corporate bond interest rates decline, with corresponding increases in plan obligations, corporate bonds will produce positive returns and vice versa. Figure 1 depicts historical stock and corporate bond returns as well as average pension plan discount rates from 2008 through 2013.

Figure 1. Investment index returns and discount rates, 2008 – 2013
Towers Watson Insider -  Investment index returns and discount rates, 2008 – 2013

Source: Towers Watson

In years when the interest rates used to measure pension plan obligations declined, corporate bond returns were generally  strong (as shown in the Citigroup column), but, as should be expected, when discount rates moved up dramatically in 2013, bond returns were negative.

Figure 2 shows average funded status for the three groups. On average, pension funded status was least volatile in Group 2 companies, those presumed to have had an LDI strategy in place over the entire analysis period. Group 2 started out with an average funded status of nearly 120% at the end of 2007, and these companies might have been attracted to an LDI strategy to safeguard their solid funded position. And, on average, their strategy paid off: Group 2 plans remained close to or fully funded throughout the analysis period.

While Group 3 companies —those that switched to an LDI strategy midstream — averaged the lowest 2013 funding levels, they also had less funding volatility than Group 1.

Figure 2. Average funded status for the TW Pension 100 by investment strategy, 2007 – 2013

Click Image to Enlarge
Towers Watson Insider - Figure 2. Average funded status for the TW Pension 100 by investment strategy, 2007 – 2013

Source: Towers Watson

Of course, the funding levels shown in Figure 2 reflect more than just asset returns and interest rates. Asset values are also affected by the level of plan contributions, while obligations reflect benefits accrued during the year and interest costs, as well as other factors such as demographic experience. The impact of benefit payments, settlements and similar transactions is reflected in both assets and liabilities.

To explore the correlation between asset and liability movement and isolate the effects of these three investment strategies on plan funding over the analysis period, we simulated three hypothetical pension plans, each starting out with $100 million in assets and obligations at year-end 2007 (Figure 3).

We assume these plans have the same inflows and outflows and exclude all other factors, so growth reflects investment performance experience for plan assets, and accumulation of interest and changes in discount rate assumptions for obligations. These factors are derived using the averages on an annual basis for the three groups.

Figure 3. Hypothetical growth of plan assets and obligations (net of cash flows and additional benefits) based on investment strategy for the TW Pension 100, 2007 – 2013

Click Image to Enlarge
Towers Watson Insider - Figure 3. Hypothetical growth of plan assets and obligations (net of cash flows and additional benefits) based on investment strategy for the TW Pension 100, 2007 – 2013

Source: Towers Watson

Looking at changes in assets and liabilities based strictly on capital market changes (i.e., investment returns and interest rate fluctuations), results for the Group 2 plan — with more than 50% fixed income throughout the period — were strong. At the end of the simulation period, funded status for the Group 2 plan was roughly 88%, while funded status for the Group 1 plan — with less than 50% of fixed-income assets throughout the period — was 79%: nine percentage points lower (Figure 4). Thanks to the benefits of hedging, the Group 2 plan also had the lowest volatility over this period.

The Group 3 plan — which switched to 50% fixed income midstream — ended up with the lowest funding level. These sponsors invested heavily in equity when the market was down and then shifted to fixed income just before stocks rebounded. On the other hand, funding levels were less volatile in Group 3 than in Group 1, especially over the last four years (when most of these companies shifted to LDI, as shown in Figure 5). The lower volatility in plans with higher fixed-income allocations is, of course, no surprise, and the primary reason plan sponsors decide to de-risk their plan assets.

Figure 4. Hypothetical levels of plan funding (net of cash flows and additional benefits) based on investment strategy for the TW Pension 100, 2007 – 2013

Click Image to Enlarge
Towers Watson Insider - Figure 4. Hypothetical levels of plan funding (net of cash flows and additional benefits) based on investment strategy for the TW Pension 100, 2007 – 2013

Source: Towers Watson

Investment returns (which are reflected in Figures 3 and 4), including averages and standard deviations, are shown in Figure 5 for the three groups of TW Pension 100 plan sponsors.

Figure 5. Asset allocation and investment returns by investment strategy for TW Pension 100, 2008 – 2013
Towers Watson Insider - Figure 5. Asset allocation and investment returns by investment strategy for TW Pension 100, 2008 – 2013

Source: Towers Watson

Pension contributions have been higher for Group 1 plan sponsors, which allocated less than 50% of assets to fixed-income investments over the entire period, than for Groups 2 and 3 (Figure 6). Roughly half of Group 1 plan sponsors held more than 50% of assets in public equity, and among this group, average contributions were almost double — 5.0% versus 2.7% — those from sponsors in Group 2. Group 2’s contributions were lower mostly because these sponsors had higher funding levels to begin with. However, the higher allocations to bonds over a period when interest rates generally declined improved returns and reduced funding volatility, which could also have played a role in reducing contribution levels.

Figure 6. Pension contributions as a percentage of plan obligations for TW Pension 100, 2008 – 2013

Click Image to Enlarge
Towers Watson Insider - Figure 5. Asset allocation and investment returns by investment strategy for TW Pension 100, 2008 – 2013

Source: Towers Watson

Conclusion

The analysis covers only six years and the financial environment exhibited some consistent, possibly unique, patterns — most notably the persistent drop in interest rates. It will be instructive to capture future periods with different economic conditions in forthcoming analyses. The results so far, however, suggest the value of LDI, especially given the volatility in financial markets over the analysis period.

Over the six years of the study, the asset/liability hedging in plans with an LDI strategy reduced funding volatility. While sponsors in the LDI group also had higher cumulative investment returns, that outcome could be unique to the period. To some extent, companies that implement fixed-income strategies should be prepared for reduced returns along with the lower volatility. The purpose of LDI is to keep pension funding on a steady, upward path, which tends to shield plans from extreme volatility, catastrophic losses and, on the flip side, outsized rewards. That’s what makes the strategy particularly suitable for companies in solid funded positions whose primary focus is protecting their assets.

Plan sponsors that recently adopted an LDI approach missed out on high equity returns in 2013 and might be disappointed in their current numbers, but they can take comfort in their lower volatility and long-term risk reduction. To minimize the timing risk associated with switching pension investment strategies, sponsors might consider dynamic asset allocation or DB glide paths. These strategies gradually shift assets (typically from equity to fixed income) as a plan’s funded status improves. A dynamic strategy that shifts assets based on interest rates should improve the results.5


Endnotes

2. The 2013 TW Pension 100 consists of sponsors of the 100 largest U.S. pension programs among U.S. publicly traded organizations, ranked by PBO at year-end 2012. For some companies, the allocation of disclosed PBO and assets between U.S. and non-U.S. is estimated.

3. See “U.S. Pension Risk Management — What Comes Next,” November 2013, Towers Watson.

4. There is no explicit indicator in disclosure data about whether a plan has adopted a liability-driven investment strategy. So for purposes of this analysis, we assumed plans with more than 50% of assets in fixed income have an LDI strategy. Actually, some companies that hold less than 50% in bonds could have adopted an LDI strategy recently with a DB glide path. Under a glide path, asset allocations are modified in a dynamic and systematic manner to reduce risk and increase the fixed-income share of assets as the plan’s funded status improves. So as funding improved significantly toward the end of 2013, additional employers might have more than 50% in fixed income in 2014, which will show up in future analyses.

5. See, for example, “Dynamic Asset Allocation for Defined Benefit Plans,” Towers Watson (November 2013).