Insights   •   Fixed Income

Target Retirement Strategy Annual Review: Evaluating US Treasury STRIPS

Each year, our Defined Contribution Investment Group conducts a comprehensive review of our target retirement strategies, reassessing the capital market expectations and demographic assumptions that underpin the glidepath, while also evaluating new asset classes and investment themes for inclusion in investor portfolios. 

The process follows a consistent and transparent three-pronged framework:

  • Investability: Can we implement this investment theme efficiently? If we’re adding an asset class, is there an appropriate benchmark?
  • Desirability: Do we expect this change to improve participant outcomes?
  • Suitability: Is the investment decision suitable for all Defined Contribution (DC)1 investors, and is the impact significant enough to justify the costs (financial and administrative) associated with a portfolio change?

This year, we evaluated US Treasury STRIPS for glidepath inclusion. Our findings follow.

Investability: Defining the Asset Class

Separate Trading of Registered Interest and Principal Securities (STRIPS) are zero-coupon fixed income securities sold at a discount to face value. First introduced in 1985, STRIPS are constructed by financial institutions that “strip” the coupons from US Treasury bonds. By removing the coupons, STRIPS (both principal and coupon) can offer greater interest rate risk sensitivity due to their longer durations. Today, while traditional long Treasury indices have durations of 18 years or less, widely used STRIPS indices have durations of 25 years or more and are sought by asset allocators seeking to match longer-term liabilities.

Desirability: The Inspiration for STRIPS

While people are living longer, long-term return expectations continue to compress, potentially creating a tension that threatens participants’ ability to achieve meaningful levels of income replacement in retirement. This year’s review focused on addressing this reality without deviating from our thoughtful and efficient approach to glidepath construction. 

A simple step toward improving expected returns, and therefore wealth accumulation at retirement, is to increase the equity weight in the glidepath. Higher equity weights will improve outcomes in the median case but introduce more uncertainty due to the higher volatility of the asset class. To diversify the market risk of our equity-heavy starting portfolios, State Street holds an allocation of long government bonds.

An increased equity allocation would reduce the long government bonds allocation, which may lead to a bumpier ride for participants and less downside protection in volatile equity markets. To counteract the loss of downside protection, we considered the introduction of Treasury STRIPS as a substitute for long government bonds. Treasury STRIPS offer more concentrated exposure to long-duration government debt and thus can provide a level of downside protection that is similar to long government bonds but at a smaller absolute allocation.

In the chart below, we compared the performance of three different fixed income indices during various equity environments. STRIPS have historically offered materially higher returns during periods when equities were down the most.

Suitability: Is Efficiency Enough?

To judge the appropriateness of STRIPS — and a higher starting equity weight — in State Street’s Target Retirement strategies, we began by evaluating the trade-offs inherent to the current glidepath.

Today, our starting allocation of 90% equities and 10% long government bonds allows State Street to provide approximately 95% of the expected return of an all-equity portfolio with approximately 90% of the volatility. Given the lower risk qualities of this starting portfolio, we are comfortable maintaining this allocation until 27 years from retirement, at which time we begin to de-risk. A more efficient starting point allows for more gradual de-risking, and holding equity risk when balances are higher is expected to lead to higher wealth accumulation over the course of a participant’s career.

If we moved the starting point portfolio to 95% equities and 5% STRIPS, our modeling suggests that expected returns would modestly improve without significantly impacting the efficiency of the portfolio (return per unit of risk). This portfolio would allow State Street to provide approximately 97% of the expected return of an all-equity portfolio with approximately 93% of the risk. Given the slope of the efficient frontier, managers that sought to improve returns by increasing equities at the expense of fixed income would likely face decreased portfolio efficiency. A more impactful fixed income allocation might mitigate this concern.

Taken in isolation, the impact of adding STRIPS today is modest. This can be attributed to the relative decline in duration for STRIPS versus long government bonds compared to historical norms. While we have historically observed that STRIPS may provide nearly double the return of long bonds in down equity markets, the state of fixed income markets today reduces those assumed diversification benefits. One can correct for the impact by allocating a greater percentage back to STRIPS, but such a change further reduces long-term return expectations and diminishes the rationale for the change.

Keeping STRIPS in View

We believe that STRIPS can play a key role in improving capital efficiency in today’s low return environment. However, given the rationale outlined above, and based on the current characteristics of each asset class, we don’t believe that the modest improvement in return expectations from adding STRIPS would justify a change at this point. We will continue to monitor this asset class in 2019 as part of next year’s review, and we expect to reevaluate the inclusion of STRIPS in the context of a broader assessment of our glidepath and potential policy changes. One particular policy issue on the horizon is the potential increase in age at which Required Minimum Distributions (RMDs) begin for traditional Individual Retirement Accounts (IRAs) and 401(k) plans. This change would be intended to match longer life expectancy and, if implemented, would affect the de-risking trajectory.