Asset allocations in defined benefit (DB) plans strongly affect overall investment returns, the plan’s funded status and funding volatility, as well as the sponsor’s cash cost and accounting expense over time. For participants, creditors, investors and regulators, asset allocations are central to a plan’s risk exposure and long-term cost. The Financial Accounting Standards Board began requiring more detailed disclosures in 2009, and Towers Watson has been analyzing asset allocations ever since.1 These analyses track asset allocation patterns over time, and this sixth edition looks at fiscal year-end 2014 pension allocations by asset classes such as cash, equity, debt and alternatives, as well as by valuation level.

The analysis is performed on both an aggregate and average sponsor basis as well as by plan size, plan status (open, frozen or closed) and funded status. We compare asset holdings from 2009 through 2014 for a consistent sample of sponsors. Finally, we examine pension assets invested in company securities.

Analysis highlights

  • On average, sponsors of frozen pension plans invested almost half their assets in conservative, lower-variance investments, such as cash and debt instruments, whereas sponsors of plans where some or all workers were still accruing benefits (open and closed plans) seemed more inclined to take on riskier investments. 
  • The overall funded status (on a plan sponsor financial accounting basis) of DB plans worsened over 2014, driven primarily by declining interest rates that pushed plan obligations higher. De-risking approaches, such as liability-driven investment (LDI) strategies that hedge against interest rate movements, played an important role in buffering funding declines. Plans with higher allocations to fixed-income assets had smaller funding losses or even modest gains versus plans with higher allocations to equity. On average, plans whose funded status improved invested more than 50% of their assets in debt. 
  • Looking at a consistent sample of sponsors, on average, the total held in public equity declined nearly 4 percentage points from 2013 to 2014. Over the same period, allocations to debt instruments increased at the same pace.
  • In 2014, almost 10% of these DB plan sponsors held assets in the form of company securities, and the allocations averaged 4.6% of pension assets among those that did.

2014 pension asset allocations

Towers Watson’s analysis of 2014 fiscal year-end asset allocations takes a detailed look at 533 Fortune 1000 U.S. plan sponsors’ pension disclosures. Figure 1 summarizes aggregate asset allocations weighted by plan size (as measured by the value of plan assets) for all Fortune 1000 pensions in the analysis. As of year-end 2014, these plan sponsors held almost $1.9 trillion in pension assets, composed of cash, public equity, debt and alternative investments (real estate, private equity, hedge funds and other).

Figure 1. Aggregate asset distribution by class and level, 2014 ($ millions) 

Towers Watson Media

Source: Towers Watson

At year-end 2014, 38.6% of total pension assets were allocated to public equity, and 42.7% were allocated to debt, with the remaining investments spread among other asset holdings.

Plan sponsors must disclose a valuation level for each major asset category as described below:

  • Level 1: Unadjusted quoted prices in active markets for identical assets or liabilities (typical for Treasury securities and the common stock of large U.S. companies) 
  • Level 2: Unadjusted quoted prices for similar assets in active or inactive markets, or other observable inputs (common for corporate debt)
  • Level 3: Unobservable inputs supported by little or no market activity, such as an expert appraisal of a real estate holding2

More than half of the asset valuations (57.8%) were classified as Level 2, and 29.7% as Level 1. Level 3 valuations (12.6%) are typically used for private equity, hedge funds and real estate.

Figure 2 depicts average asset allocations (weighted by the number of plan sponsors) for the same sponsors. The average Fortune 1000 pension plan in the analysis held roughly $3.5 billion worth of assets at the end of 2014.

Figure 2. Average asset distribution by class and level, 2014 ($ millions) 
Towers Watson Media

Source: Towers Watson

Among these plans, the average allocation to public equity was 43.5%, while the aggregate allocation was 38.6%. As for alternative assets — real estate, private equity, hedge funds and other investments — average allocations were 9.9%, while aggregate allocations were 15.4%. The difference between the aggregate and the average reflects differences in plan size — larger plans were more likely than smaller plans to invest in alternatives and less in public equity. On average, more than half the asset valuations were classified as Level 2 (59.4%). Thirty-three percent were classified as Level 1 and only 7.6% as Level 3.

Asset allocations by plan sponsor's asset holdings

Aggregate and average asset allocations for small, medium and large DB plans are shown in Figures 3a and 3b. The analysis divides these into three groups of sponsors by total plan assets: Small plans held less than $527 million, midsize plans held between $527 million and $1.99 billion, and large plans held more than $1.99 billion. The largest plan held assets worth more than $95 billion.

Figure 3a. Aggregate allocations by plan sponsor's asset holdings, 2014 

Towers Watson Media

Figure 3b. Average allocations by plan sponsor's asset holdings, 2014
Towers Watson Media

Source: Towers Watson

As asset amounts increased, public equity allocations declined, averaging 41% for the largest plans versus 46% for the smallest. This confirms the differences between the results shown in Figures 1 and 2, where public equity holdings were lower when assets were weighted by plan size. While larger plans allocated less to public equities, their allocations to other return-seeking investments — real estate, private equity and hedge funds — were more than double those of small plans.

Weighting small, medium and large plans by plan assets (Figure 3a) emphasizes the large share of pension assets held by very large plans,3 as well as the pronounced differences in investing behavior between small and very large plans.

Pension asset allocations by plan status

For this part of the analysis, we divided plan sponsors into three mutually exclusive categories: those whose primary pension plan was frozen, those whose primary pension plan was closed and those with open plans. Of the 533 plan sponsors in this study, 68% had a pension plan categorized as either frozen or closed, while 32% maintained open DB plans.

Figures 4a and 4b show asset allocations by plan status and demonstrate a relationship between plan status and investment risk, with the correlation strongest on an aggregate basis (Figure 4a). Sponsors of frozen plans invested more than half their total assets in conservative, lower-variance investment instruments, such as cash and debt, whereas sponsors of plans where some or all workers continued to accrue benefits (closed and open plans) seemed more inclined to take on riskier investments.

Figure 4a. Aggregate asset allocations by plan status, 2014
Towers Watson Media

Figure 4b. Average asset allocations by plan status, 2014
Towers Watson Media

Source: Towers Watson

Pension asset allocations by funded status

In our 2009 through 2012 analyses of asset allocations, pension funding remained relatively stable, with average funded status typically ranging between 75% and 80%.4 In 2013, interest rates rose for the first time in years, pushing liability values down. Higher interest rates combined with very strong equity returns and substantial cash contributions boosted funding levels to an average 87% at year-end. Over 2014, the average funding level fell back to 79% and the number of fully funded pensions declined from 14% to 5.5%. The deterioration of funded status was primarily owing to the lower interest rates used to measure liabilities, which pushed them steeply higher.

Moreover, many U.S. plan sponsors also adopted new mortality assumptions (motivated by a report issued by the Society of Actuaries in 2014), which reflected longer life expectancies for workers, thereby increasing plan liabilities by an additional four percentage points overall. Liability increases overwhelmed even the most conservative investment strategies, but plan sponsors with greater concentrations in equity realized larger funding declines compared with those more heavily invested in bonds.

Our 2014 analysis shows a correlation between funded status and asset allocations (Figure 5a). Sponsors with better-funded pensions held less in public equities and more in debt than their less well-funded counterparts. This resonates with the de-risking strategies, such as LDI, now in operation in many pension funds. The only exception to this result was among sponsors whose plans had funded ratios greater than 100%, which could be a lagging effect of the extraordinary equity boost in 2013.

Figure 5a. Average asset allocations by funded status, 2014
Towers Watson Media

Figure 5b. Average asset allocations by change in funded status, 2014
Towers Watson Media

Source: Towers Watson

Figure 5b depicts the relationship between a change in funded status and asset allocations during 2014. Sponsors whose plans realized funding gains in 2014 were more likely to have large holdings in debt, presumably in long bonds. On the other end of the spectrum, those with large funding declines were more likely to be heavily invested in public equity.

In 2014, robust returns on long bonds (as shown in Figure 6) coupled with plan contributions helped some sponsors mitigate the effects of interest rate declines on plan funding. Conversely, funding declined in plans with higher equity allocations as sponsors’ cash contributions combined with moderate equity returns did not offset interest rate drops (although bonds would not have hedged improved mortality assumptions either). However, many of the same sponsors who realized significant losses in 2014 had enjoyed major gains in 2013 when stock returns were very strong.

Figure 6. Investment returns, 2009 – 2014
Towers Watson Media

Source: Bloomberg

The higher funded status many of these plans attained in 2013 could also have acted as a de-risking trigger, prompting some sponsors to try to lock in their funding gains. But now more than ever, an adverse macroeconomic environment and a greater appetite for reducing funding volatility should interest more sponsors (especially those with frozen plans) in a glide path type of strategy. In a glide path strategy, future target allocations are based on the plan’s funded status, with the sponsor shifting assets from equities to debt as funding levels climb. This enables pension funds to reduce risk and safeguard gains (albeit reducing the opportunity for more-than-moderate future gains as well).

There is some evidence of de-risking in progress, as 16% of Fortune 1000 DB plan sponsors explicitly mentioned implementing LDI or long bond strategies. However, only 8% of 2014 Fortune 1000 DB plan sponsors explicitly linked their future target allocations with the plan’s funded status in their annual pension disclosures, up slightly from 6% in 2013.

Pension assets held in company securities

Almost 10% of DB plan sponsors held assets in the form of company securities in 2014, declining slightly from 11% in 2013. These allocations averaged 4.6% of pension assets in 2014, dropping to 2.5% when weighted by end-of-year plan assets. The weighted average is lower than the simple average since larger plans allocated lower percentages to company securities than smaller plans.

In most of these plans (60%), employer securities made up less than 5% of total pension assets for 2014. Company securities were more than 10% of plan assets for only a handful of plan sponsors (Figure 7), and those instances reflect higher past returns rather than allocations to employer securities of more than 10% when contributed.8 

Figure 7. Allocations of company stock holdings, 2014 (percentage of plan sponsors)

Towers Watson Media

Source: Towers Watson

Six-year asset allocations

The 2009 to 2014 asset allocation studies are based on a consistent sample of 305 plan sponsors. Figures 8a and 8b show asset allocations for these sponsors on an aggregate and average basis over those six years.

Figure 8a. Aggregate asset allocations by class and level for consistent sample of 305 pension funds, 2009 – 2014 ($ millions)

Click to Enlarge

 Towers Watson Media

Source: Towers Watson

On an aggregate basis, public equity holdings declined by 3.9 basis points (bps) and debt holdings increased by 3.7 bps from 2013 to 2014. 

Figure 8b. Average asset allocations by class and level for consistent sample of 305 pension funds, 2009 – 2014 ($ millions)

Click to Enlarge

Towers Watson Media

Source: Towers Watson

Overall asset allocations were relatively stable in 2009 and 2010, but between 2010 and 2011 — a period of poor stock market performance — average allocations to equity dropped from 51.1% to 46.2%, while average allocations to debt rose from 36.3% to 38.9%.

There was little change in overall asset allocations between 2011 and 2012. Between 2012 and 2013, equity allocations rose and debt allocations declined, but both changes were relatively minor and might have resulted from strong equity performance in 2013. In 2014, there was a substantial shift away from equities into debt — it’s possible that strong funding levels in 2013 motivated sponsors to shift to less risky investments to protect some of their gains. On average, equity holdings declined by 4.0 bps over 2014, while debt holdings increased by 4.0 bps. Since 2009, average allocations to public equites declined by almost 10 bps.

Figure 9 shows that in 2014, nearly half of pension sponsors reduced their equity allocations by between 0.1% and 4.9%. Of those that had larger reallocations — increases or decreases of more than 10% of equity holdings — almost 15% reduced their equity share by more than 10% (with an average decrease of 18.5%). On the other hand, only 1% of pensions increased their allocations to equities by more than 10% (with an average increase of 22.6%).

In line with our previous analysis, almost 76% of pension plans increased their allocations to debt securities. Forty-three percent of plans increased their allocation to fixed-income assets by .01% to 4.9%.

Figure 9. Average allocation changes in equity and debt holdings over 2014

Click to Enlarge

Towers Watson Media

Source: Towers Watson


Lower interest rates and moderate equity returns set funded status back in 2014, especially where plans were heavily invested in equities. More conservative plan sponsors were able to buffer such negative effects via higher allocations to debt.

The shift in equity allocations versus debt allocations was largely symmetrical, with 76% of plan sponsors allocating more to fixed-income assets and 81% allocating less to equities. The primary shift in 2014 was from public equities to debt rather than from public equities to other return-seeking assets, as was the case in earlier years. Larger plan sponsors continued to hold less equity and more diversified allocations than smaller plans. Frozen plans held more fixed-income assets, on average, compared with closed or open plans.

Given volatile market conditions, adopting or maintaining an effective de-risking strategy could be more important than ever for pension plan funding.


1. See “2013 Asset Allocations in Fortune 1000 Pension Plans,” Towers Watson Insider, December 2014.

2. For Level 3 assets, a reconciliation of the beginning and ending balances is also required, reflecting the actual return on plan assets, purchases, sales and settlements. 

3. The 17 largest plans (or 10th decile) represent 37% of all plan assets in this study and 43% of assets among the largest group of DB plan sponsors.

4. Funded status is defined as the ratio of the fair value of assets over projected benefit obligations (a financial accounting measure) at year-end.

5. The Standard & Poor’s 500 is an American stock market index based on the market capitalizations of 500 large companies having common stock listed on the NYSE or NASDAQ.

6. The Russell 2500 Index is a subset of the Russell 3000® Index. It includes approximately 2,500 of the smallest securities based on a combination of their market cap and current index membership.

7. The MSCI EAFE Index is a stock market index that measures the equity market performance of developed markets outside of the U.S. and Canada.

8. The Employee Retirement Income Security Act (ERISA) does not allow U.S. DB plans to invest more than 10% of assets in company securities.