One of the key lessons of 2008 was the critical importance of maintaining adequate liquidity in the event of a market dislocation. Many pension plans with portfolios that relied on overlay strategies (for lengthening the duration of their assets, for example) found themselves suddenly short of liquid assets that were needed to satisfy collateral calls. As a result, the need for thoughtful risk management in portfolio strategies that make use of derivative overlays has become more pronounced, with a particular focus on the important role of advanced planning.
Why do we use overlays?
An overlay strategy involves the use of a derivative instrument such as a swap or a futures contract to provide exposure to a specific market risk in a form that can be significantly more capital-efficient than could be achieved in the physical securities market. Unlike traditional stocks and bonds, derivative positions are contractual obligations that usually reference either a physical security or a market index. While at maturity a physical security or cash may need to be delivered in order to satisfy the obligations of the derivative contract, prior to the maturity typically only a fraction of the outlay is needed that would be required to establish the same economic exposure using the referenced physical securities or market index. The plan can then invest the capital that is freed up in another investment strategy that either is not readily available in derivative form or is actively managed for absolute performance.
There is a financing cost associated with the use of derivatives, and therefore the typical overlay strategies are used when it can be reasonably assumed that the capital that is freed up can yield a return higher than the financing cost of the derivative instrument. The benefit of this overlay approach is that a plan may be able to meet multiple objectives more effectively than it could if it was limited to using direct investments in stocks and bonds alone. For instance, a plan using interest rate derivatives can achieve a more precisely structured liability hedge (intended to offset the interest rate risk arising from changes in liability discount rates) and at the same time retain exposure to higher returning investments.
A plan that does not employ derivatives would need to shift assets from stocks to bonds to achieve the same hedge, thereby lowering potential long-term returns because of the shift from generally higher yielding equity investments to fixed income. In the current low-yielding, low-expected-return environment we expect overlay strategies to play a more important role than they did before 2008.
The strategic benefit of this approach notwithstanding, the use of derivative overlays also creates operational risks around the need for liquidity to support margin and collateral demands. It should also be noted that use of overlays may increase the complexity, leverage and market risk of the plan overall, so careful consideration need to be paid to portfolio construction and risk management.
Why do overlays need liquidity?
Whether the derivatives are traded on an exchange or over the counter, a negative change in the mark-to-market value of the position will usually result in the investor having to provide additional cash or eligible securities. These are typically non-negotiable obligations that must be satisfied immediately, or the position will be closed out and collateral may be forfeited. It should be noted that for OTC transactions, the types of eligible collateral can be negotiated with the counterparties as part of the master agreement and could therefore include nearly any asset type. Commonly, however, only cash, Treasury bills and sometimes Treasury bonds are eligible to be delivered as collateral.
It follows that a plan making use of such strategies will want to ensure that a sufficient supply of liquid, eligible collateral instruments is on hand to satisfy potential collateral calls. Unfortunately, there are usually significant opportunity costs associated with holding liquid collateral, which tends to be of a more conservative, lower-returning nature than the plan’s desired investments. In today’s environment with steep yield curves, this opportunity cost is even higher than before. It is therefore critical that investors make every effort to establish a plan on how to manage the liquidity of the overall portfolio with an eye on both being prepared to satisfy potential collateral needs and minimizing the cost impact of this preparation.
Step 1: Determine sources of liquidity
One approach that can be effective is to tier the sources of liquidity into “current” and “contingent” tiers, where some assets are kept in more liquid form to satisfy short-term collateral requirements and other assets are kept in higher yielding investments but are thought to be available for liquidation in case the need arises. This analysis will form the basis of a liquidity budget – a plan on how to satisfy mark-to-market calls. As an example, the table below breaks out categories that could form current and contingent liquidity tiers.
Step 2: Determine size of liquidity pool
Given that a plan may have strategic investments in many of the secondary and back-stop liquidity buckets, the key question is how much of the more costly (in terms of lower yield or lower return opportunity) immediate liquidity category is needed. Quantifying this liquidity need is a relatively straightforward risk-management exercise involving estimating the potential mark-to-market change in value of the overlay.
A standard approach is to use Value at Risk (VaR) analysis to estimate the potential change of a given severity (e.g. one, two or three standard deviations) over a given time horizon (one, three, five or ten days for instance). Let’s look at an example of a pension plan with assets of $1 billion using a 30-year receive-fix interest rate swap with a notional size of $200 million. As you can see in the table below, the plan carries some leverage because the swap position is unfunded.
There are many techniques that can be used to arrive at a VaR estimate. Using the standard PIMCO approach, we can estimate the immediate collateral needs associated with the swap overlay for two-, five- and 10-trading-day horizons as shown below.
It’s clear that the range of potential collateral pool sizes varies a great deal depending on the degree of risk that can be tolerated. In our view, a two-standard deviation and 10-day move is a logical place to start (as shown circled in the table) because it represents a market move that is very large by normal standards and a time horizon that should provide the pension plan adequate time to access its secondary or back-up liquidity facilities. We believe it would be reasonable for a plan to count some of its more liquid secondary sources against this target. A shorter time horizon might be justified if a plan is more confident in its operational controls and ability to source liquidity.
Additional sensitivity testing
As with any other form of budgeting, there will be deviations from expectations. In this case, the deviations will be from actual collateral needs on both the gain and loss sides. Therefore, as a secondary safeguard, it may be worthwhile to use historical data to determine whether or not there have been times when multiple standard deviation anomalies would have resulted in a loss greater than the collateral that was budgeted and available.
For the earlier example of a 30-year swap, we can estimate that based on the “immediate collateral” pool described earlier ($20.4 million), an approximately 50-basis-point parallel move in the interest rates during a single day would likely result in a collateral call in excess of available liquidity. Using a historical simulation based on U.S. Treasury 30-year bond yield changes, we can find one single day move and six two-day moves where this would have been the case (all of them during the volatile period of 1980-1982). This conclusion is based on our analysis of Federal Reserve and Bloomberg data from 15 February 1977 through 26 April 2012, as shown in Figure 1.
This analysis shows how important it is to at least consider what steps would be taken should the markets experience a move of more than two standard deviations in the short run.
Where should the liquidity pool be located?
Within the portfolio structure, where should this liquidity be located? One approach in structuring the portfolio in our example would be to have the swap located within an actively managed fixed income separate account, with the separate account assets serving as the collateral pool (internal liquidity). Another approach is to maintain the swap as a separate asset and to maintain a dedicated pool of highly liquid assets to serve potential collateral needs (external liquidity).
PIMCO’s view is that the internal liquidity approach has a number of potential advantages over the external model:
- A single professional portfolio manager overseeing both the managed portfolio and the overlay in real time, with the ability to increase liquidity in the separate account as needed.
- The managed portfolio can be structured to hold “optimal” liquid assets (e.g. long Treasuries) that serve both purposes: the strategic allocation target of the plan and the potential collateral needs of the overlay.
And some possible disadvantages:
- The separate account benchmark may need to explicitly incorporate highly liquid securities such as Treasuries in order to include liquidity as part of the asset allocation. This may limit the choice of benchmarks for the account.
- The combination of overlay, collateral and the managed portfolio in the same account may also make performance accounting difficult for the plan.
- A large move in the overlay may exceed the capacity of the separate account to provide liquidity, necessitating additional capital contributions (or the sale of the less liquid instruments).
- The market environment in which a long duration overlay loses value (rising rates) may also adversely impact the separate account, reducing its capacity for liquidity at the same time as the need for collateral is increasing.
Liquidity assets and active management
As discussed earlier, our natural objection to holding collateral in cash instruments is that it is an inefficient use of capital. Keeping cash or other highly liquid instruments on the balance sheet can be a potential drag on overall performance.
Furthermore, even if the assets intended for collateral are kept in higher-yielding securities, they are often managed passively. This eliminates any potential returns from active management. If we assume that a reasonable return for bonds is 3% to 4% and active management can add another 0.5% to 1%, the opportunity cost of holding $20.4 million in cash could be as much as $1.2 million per year. And this contingent collateral is only sufficient for a ten-day, two-standard deviation move – more would be needed if interest rates continued to rise. This example does not include the effects of fees. Returns would be lower if applied. There is no guarantee that active management will add value.
One of the solutions to this issue could be a partial shift from traditional separate accounts to commingled vehicles with higher liquidity characteristics, such as actively managed ETFs. While these strategies are currently uncommon, we believe that they may become more popular.
Putting it all together
In conclusion, we have outlined some key aspects of what we believe are necessary for liquidity management for institutional investors using overlay strategies. A comprehensive liquidity plan, consisting of (i) identification of sources of current and contingent liquidity, (ii) detailed analysis of operational aspects and consideration between internal and external sources of liquidity and finally, (iii) a Value at Risk and historical simulation-based framework for determining the relative size of current and contingent collateral.
The current market environment is a challenging one for defined benefit pension plans and other institutional investors, and preserving the ability to make use of any strategically sound investments should be a goal of plan managers. To the extent that derivative overlays are going to play a key role going forward, it is critical that we take careful note – not just of their potential benefits – but also their potential risk. With a well-designed implementation plan that incorporates a thorough liquidity management process these instruments remain a valuable tool in a plans toolkit.