Click here to read Rene Martel's biography.

If any positive conclusion can be drawn from all the pain endured throughout the financial crisis of 2007–2009, it is the importance of implementing and enforcing sound risk management practices. More specifically, the recent market dislocations highlighted the danger of relying on historical correlations as a risk management tool. In light of this, we expect a number of defined benefit plan sponsors to reevaluate the tools they use to implement their liability-driven investment (LDI) strategies.

While the overlay structure inherent in the construction of many LDI strategies – particularly dollar duration matching strategies – may remain effective, the severe dislocation between the swap market and the corporate bond market in the wake of the financial crisis makes this a good time for plan sponsors to consider switching gears. An alternative to the traditional interest rate swap overlay is to derive the equity exposure from an equity derivative overlay and achieve the entire duration exposure directly from physical bonds, especially those that are driving liability valuations, like corporate bonds.

Over the last few years, a large number of corporate defined benefit plans have implemented LDI strategies in order to reduce mismatches between pension assets and liabilities. Their objective was ultimately to reduce the volatility of variables like funding status, pension contributions and pension expense. Those LDI strategies have provided certain advantages during the financial crisis, and plan sponsors who implemented them have generally experienced lower declines in funding status as the interest rate hedge component helped offset the impact of declining Treasury rates on their liabilities. However, the market dislocations resulting from the credit crunch have brought to light a number of inefficiencies and new risks in traditional LDI strategies.

Revisiting Interest Rate Swap Overlays
Before 2008, the main focus of LDI strategies was to increase the duration, or interest rate sensitivity, of plan assets with a diversified mix of duration sources, such as a combination of Treasury bonds, corporate bonds and long-dated interest rate swaps. In many cases, Treasuries and swaps made up a significant portion of the hedging portfolio. This was especially true for dollar duration matching strategies where interest rate swaps or Treasury futures were used to lever the fixed income portfolio duration in order to match a larger share of the duration risk while limiting the allocation to fixed income. 

For example, a plan sponsor with a liability duration of 12 years and a 50% allocation to fixed income could lever the duration of the fixed income assets to 24 years by using an interest rate swap overlay – a strategy that achieves additional interest rate exposure through the use of derivatives. The fixed income portfolio sensitivity to parallel shifts in interest rates would then be approximately twice as much as that of the liabilities (ignoring convexity effects) as reflected by the doubling of the duration. However, because there are only 50 cents of fixed income assets for every dollar of liabilities, the actual dollar impact on the fixed income and overlay structure would match the changes in the liabilities, minimizing surplus volatility in dollar terms.

Those dollar duration matching, or swap overlay, strategies enjoyed considerable outperformance relative to corporate pension liabilities in 2008 as long Treasury rates and swap rates rallied significantly more than long corporate rates used to discount liabilities (see Chart 2 below). That is, the assets that were supposed to hedge the liabilities by closely tracking them actually contributed significant positive performance relative to liabilities last year.

The explanation for why fixed income assets and pension liabilities with similar duration had such divergent performance over a short period of time lies in the extraordinary market volatility and, more importantly, the dislocation of swap spreads and corporate spreads that began in late 2007 (see Chart 3 below). Before that time, fluctuations in long-dated swap spreads tracked changes in long corporate spreads fairly closely.¹ Accordingly, swaps were expected to be an appropriate hedge for pension liabilities that are discounted with corporate rates. In addition, corporate spread volatility had been relatively low over the last 10 years preceding the crisis, and therefore many expected the volatility of pension liabilities to be driven mostly by changes in Treasury rates rather than changes in corporate spreads. Finally, corporate spreads were considered to be narrow – by historical standards – at the time that many sponsors implemented their LDI programs, making it unlikely that corporate spreads would tighten enough to lead to a significant increase in liabilities that would not be offset by capital gains on swaps or Treasuries. In other words, if corporate spread volatility were to increase, it was more likely to occur through a rise in spreads, and therefore swaps and Treasuries would actually outperform liabilities – that is, sponsors would most likely be on the right side of the mismatch. In that context, an LDI portfolio with a significant allocation to Treasuries and swaps made sense, particularly since swaps and Treasuries typically offer better liquidity, lower transaction costs and more flexibility to closely match secondary risk factors² than corporates, with limited exposure to default or downgrade risk. In an environment where the risk of further corporate spread tightening was limited, it seemed that those advantages more than offset the corporate spread mismatch risk for many plan sponsors.

However, the recent market disruptions have changed the balance of risks faced by plan sponsors, highlighting the danger of relying on historical relationships when structuring LDI strategies. As a result of the dislocation between the swap market and the corporate bond market, plan sponsors should consider repositioning their dollar duration matching strategies in an effort to achieve a better and more reliable liability hedge, while potentially locking in the gains from their swap overlay programs. The objective for corporate plan sponsors is no longer to simply get the right amount of duration, it is to get both the right amount and the right sources of duration. We believe this means that a larger share of the duration exposure should come from credit or corporate bonds and a smaller portion should come from swaps or Treasuries.

A More Efficient Way to Implement Dollar Duration Matching Strategies
In response to the changes in fixed income markets, a number of plans have increased the credit exposure of their fixed income portfolios. While the nature of liability valuation methodologies makes it difficult to find a very tight hedge for the credit component, it is generally possible to bring the portfolio’s corporate exposure relatively close to that implied by liability discounting methodologies. However, the amount of corporate spread duration, or the sensitivity to corporate spreads, that may reasonably be achieved is limited by the scarcity of corporate bonds with durations exceeding 12 years. This creates a challenge for sponsors who want to implement dollar duration matching strategies. Going back to the earlier example, a plan sponsor seeking to achieve a 24-year duration portfolio would have to take more than half of the duration exposure in Treasuries or swaps. This is because the maximum achievable duration with a diversified portfolio of corporate bonds is around 11 to 12 years. The remaining duration exposure would have to be gained through Treasury STRIPS³ or futures, the swap market, or both. And, as we have discussed earlier, those instruments will most likely not provide a good match to liabilities when the liability discount rate volatility is driven mostly by corporate spread fluctuations.

Does this mean that dollar duration matching solutions are doomed to fail in the current market environment? We think not. By restructuring the physical and synthetic allocations in portfolios, plan sponsors can obtain a duration exposure that is more consistent with liability valuation methodologies while still maintaining significant allocations to risk assets. We believe that alternative approach will fare much better as a liability hedge than swap overlays when corporate spreads are volatile.

The traditional swap overlay approach included:

• Achieving a portion of the duration exposure in the physical market through a fixed income allocation,
• taking equity (or other risk assets) beta exposure in the physical market as well through a risk asset allocation, and 
• running a duration completion overlay in the derivatives market in order to achieve the duration coverage target.

An alternative is to take the entire duration exposure in the physical market in an effort to achieve a higher exposure to those securities that really drive liability valuations (i.e., long duration corporate bonds) and get the equity exposure in the derivatives market (as illustrated on the right hand side of Chart 4 below).

This alternative strategy maintains the same exposure to equities from a beta standpoint⁴ as the traditional swap overlay strategy, and the absolute amount of duration exposure will also be similar in most cases although it will be attained in the physical bond market instead of the derivatives market. But there is a significant improvement in terms of matching the corporate spread exposure of liabilities. In fact, the credit spread duration coverage of the long bonds and equity overlay structure is approximately three times that of the swap overlay approach (as shown in Chart 4 above). In the end, the sponsor can maintain the same exposure to equities, but the duration risk versus liabilities is very likely to be significantly lower than under the traditional swap overlay structure because the sources of duration are better aligned with liability discounting methodologies.

Additional Potential Benefits of the Long Bonds and Equity Overlay Structure
While the primary benefit of the long bonds and equity overlay structure is to provide a better match to corporate pension liabilities, there are also a number of other potential advantages.

Facilitating portfolio rebalancing and reducing transaction costs
When equity and fixed income markets move in opposite directions or in different magnitudes, sponsors will have to rebalance their asset mix toward the policy target from time to time. Under the typical asset allocation model, this often means selling and buying physical securities in both the fixed income and equity markets. This inevitably generates transaction costs, which can be especially problematic in volatile markets. For example, after a sharp decline in equities, plan sponsors could sell bonds and buy equities to bring the allocations back in line with their targets. If the equity market rebounds sharply, sponsors could end up in a situation where they sell the equities they bought a few months earlier and buy the bonds they sold to rebalance to their targets again. This would generate even more transaction costs. With the long bonds and equity overlay structure, it is much easier and less costly to rebalance the equity allocation. The sponsor simply adjusts the size of the equity overlay, which will generally cost a fraction of the price of actually selling and buying physical securities.

The yield advantage may help offset benefit payments and other liquidity needs
As physical assets are moved from equities to long duration bonds, the yield from the physical portfolio increases significantly. Historically, the yield on a long government / credit fixed income portfolio has significantly exceeded the S&P 500 dividend yield (as shown in Chart 5 below). That extra yield can help meet liquidity needs, such as benefit payments. Once again, this has the potential to reduce transaction costs as the plan sponsor is less likely to be forced to liquidate a significant amount of physical assets in order to pay benefit payments. The yield advantage is a source of liquidity that becomes even more important when markets freeze and it becomes more difficult to sell assets to meet liquidity requirements.

Hurdles in the Implementation of the Long Bonds and Equity Overlay Structure Are Easily Overcome
Some plan sponsors have questions about certain issues related to the actual implementation of the long bonds and equity overlay structure, but on close examination, many of these concerns can be addressed.

What if the risk asset allocation does not include exclusively equities?
For many plan sponsors, the allocation to risk assets is not entirely invested in equities. In that case, the long bonds and equity overlay structure might not be applicable to certain risk asset classes where no synthetic market exists. However, the asset classes this structure can accommodate generally represent the lion’s share of risk asset allocations. In addition, even if a portion of the risk asset allocation cannot be replaced with a derivative overlay, partial implementation may still help reduce the asset-liability mismatch to a significant extent and ultimately can lead to lower volatility of variables such as funding status, contributions or pension expense.

What if corporate spreads are not expected to tighten?
While it’s true that a swap overlay structure may outperform relative to liabilities in market environments where corporate spreads are expected to widen more than swap spreads (or swap spreads are expected to tighten more than corporate spreads), using the interest rate swap overlay strategy to potentially capture that move represents an active risk position that increases the asset-liability mismatch. For many plan sponsors, the magnitude of that risk exposure could be very large. As an active manager, PIMCO makes relative value calls on different sectors of the fixed income market all the time. We devote a significant amount of time and resources to analyzing the economy, the markets and specific sectors and securities before making relative value decisions. Despite that, we typically do not overweight a specific sector (such as swaps over corporates, for example) by more than a few years of duration because it is imperative to maintain our risk exposure at an acceptable level. And this is true even when we have a strong conviction that a sector will outperform. Plan sponsors implementing duration hedges through interest rate swap overlays are taking a much larger active risk exposure (often eight to 14 years of credit spread duration mismatch). Therefore, implementing the traditional swap overlay structure instead of the long bonds and equity overlay structure due to a tactical consideration requires a very strong conviction that swaps will significantly outperform corporate bonds.

In addition, the long bonds and equity overlay structure can even accommodate the needs of a plan sponsor who really wants to retain the ability to take significant tactical sector risk exposures. For example, if corporate bonds are expected to underperform over a certain horizon, the plan sponsor could reduce the credit exposure in the underlying fixed income portfolios in favor of other sectors, such as Treasuries.

Does the plan sponsor entirely give up alpha potential on the equity exposure?
It is true that the equity exposure in this strategy is entirely passive, and the sponsor forgoes the opportunity to benefit from potential alpha generated via the traditional stock selection strategies implemented by equity managers. However, in a normal market environment where the yield curve is steep, the investor will pick up a yield advantage as the yield on the underlying long bond portfolio will generally be higher than the financing rate on the equity overlay (as shown in Chart 6 below). This yield advantage actually provides incremental return expectations relative to a standard equity mandate. In addition, it is important to consider that alpha can be generated on the underlying fixed income portfolio; therefore the plan sponsor does not necessarily give up alpha potential.

It’s a Good Time for Plan Sponsors to Consider a Long Bonds and Equity Overlay Structure
Given that the breakdown in the previously strong relationship between swap spreads and corporate spreads has created a situation where traditional swap overlay programs can now lead to a significant mismatch between plan assets and liabilities, it is a good time for sponsors to consider a long bonds and equity overlay structure for their dollar duration matching strategies. That alternative structure is designed to provide a tighter fit between plan assets and liabilities, as the sources of asset duration are better aligned with typical liability discounting methodologies. These days, plan sponsors can’t just focus on achieving the right amount of duration; they need to think carefully about the sources of that duration.

¹ This is especially true for AA corporate spreads, which impact the valuation of liabilities because corporate pension plans generally discount their liabilities using AA corporate rates for financial reporting purposes.

² For example, curve risk exposure or convexity.

³ Separately Traded Registered Principal Interest Securities: These bonds pay no interest because they have been “stripped” of their coupons, also known as zero-coupon bonds.

⁴ This means that the exposure to equity market movements overall is the same.