Stepping back into inverse IO

admin | September 14, 2018

This document is intended for institutional investors and is not subject to all of the independence and disclosure standards applicable to debt research reports prepared for retail investors.

Valuations in inverse interest-only MBS, or IIO, have become interesting again. Spreads have widened substantially this year. And while it looks too early to jump in completely, it also looks like a good time to start building a position. Spreads and potential returns seem more than enough to compensate for the embedded risk, and portfolio diversification from IIO is likely to be valuable over the next few years.

Wider on higher LIBOR, less capital, concern about collateral

Option-adjusted spreads have widened nearly 200 bp on a broad range of strip IIO since hitting a low in mid-2017. They now range from 200 bp to 350 bp or higher in YieldBook OAS, the same neighborhood last seen in mid-2013 after the Fed signaled eventual tapering in quantitative easing (Exhibit 1).

Exhibit 1: OAS on IIO have returned to mid-2013 levels

Note: OAS reflects both conventional and Ginnie Mae strip IIO backed by 30-year loan balance 3.0% to 4.0% pass-throughs. Source: APS.

The widening in part likely reflects market expectations for a steady rise in LIBOR, which reduces the coupon and value of IIO. For investors new to the instrument, IIO are identical to a LIBOR-financed position in a fixed-rate MBS. In a strip IIO, the fixed-rate MBS often is a pass-through. As LIBOR rises, the difference between LIBOR and the pass-through coupon narrows, reducing IIO coupon and value.  LIBOR can rise to a level where the IIO the coupon goes to zero, usually called the strike on the IIO. But if LIBOR rises above the strike, the IIO coupon still stays at zero, implying a cap on the financing rate. The embedded financing makes IIO very sensitive to the slope of the yield curve.

LIBOR initially moved up this year first as tax reform triggered repatriation of offshore US investments and widened the spread of LIBOR to OIS. It kept moving up faster than implied forward rates as the Fed signaled hikes well into 2019 and the market repriced to a higher LIBOR path. LIBOR looks likely to continue repricing above implied forward rates since the Fed seems determined to keep hiking almost quarterly through 2019. But the market will get valuable information from the September 26 FOMC meeting when the committee publishes its first estimates of fed funds rates through 2021. The 2021 dots should show whether the committee expects to keep raising rates after 2020, hold them steady or begin moving toward a lower neutral rate.

Strong demand for floating-rate CMOs, also possibly in response to expected higher LIBOR, has helped push IIO spreads wider, too. Creating an IIO always involves issuing a floating-rate CMO at the same time. The floating-rate class effectively finances the IIO. Since the OAS of the underlying fixed-rate MBS effectively gets allocated between the floating-rate class and the IIO, tighter OAS on the floater generally means wider OAS on the IIO.

Wider spreads since mid-2017 also may have reflected the liquidation of a number of prominent IIO portfolios. Structured Portfolio Management and Providence Investment Management, both long-time investors in IIO and other mortgage derivatives, have closed since mid-2017. Although these two portfolios on their own may not have been enough to widen IIO, their liquidation likely reflected tight spreads and lower potential return in the sector. Other portfolios have lost capital and reduced positions.

Concerns about TBA collateral quality may have contributed as well to generally wider IIO. As the Fed has reduced its reinvestment in MBS, a higher weighted average coupon, larger loan size and servicer mix in TBA have all added to expected TBA negative convexity. Of course, that should have little effect on IIO backed by specified pools.

Good compensation for lingering risks

The market still seems to underprice the path of the Fed, and revised expectations for higher LIBOR could keep pressure on IIO. TBA collateral quality should slowly keep weakening. But spreads in IIO look sufficient to cover these risks.

Raw projected 1-year total returns for representative strip IIO, assuming constant rates, now range in the low teens (Exhibit 2). After hedging IIO duration and yield curve exposure to zero using 2-, 5- and 10-year points on the yield curve, 1-year returns under constant rates drop to between 6% and 8%. The hedged returns arguably capture fair compensation for risk in the IIO net of rates—namely negative convexity, volatility, prepayments and liquidity.

To put 6% to 8% 1-year hedged returns in context, most assets in a similar scenario would likely generate gross returns close to their yield. Net returns would come closer to the asset’s spread. With a relatively flat spot and forward curve, yield and spread should dominate most asset returns. In theory, a position in IIO hedged for rates, curve and volatility would return something close to its OAS. Both the hedged returns and OAS on IIO look attractive compared to most alternatives.

Exhibit 2: Good gross and hedged returns in IIO

Note: Projected 1-year returns assuming constant rates, reinvestment at 1-month LIBOR, horizon repricing at constant OAS. Returns on short positions in Treasury notes assume financing at 2.0% as a conservative approximation to 1-year general collateral repo rates. All market levels as of COB 9/11/18. Source: YieldBook, APS.

Timing the Fed cycle

If there is an ideal time to buy IIO, it might be the point where the Fed has finally finished hiking. The policy bias and market expectation at that point would likely turn from higher LIBOR toward flat or lower LIBOR, a big positive for IIO coupon and value. That policy inflection seems likely to come within the next two years and get priced in well before it arrives. IIO should tighten substantially as the market starts to price for falling LIBOR rates. Investors could wait for that Fed policy inflection to come, but few portfolios have comparative advantage in timing a Fed cycle. Current IIO spreads and potential returns nevertheless compensate a portfolio for starting to build a position now.

Balancing broader portfolio credit exposure

Managers of broader fixed income portfolios should also think about the diversification benefits from an IIO position. The peak of the Fed cycle and a flat yield curve would likely weigh on credit spreads, especially in portfolios that might feel the direct or indirect effects of the large build-up of ‘BBB’ corporate credit. A softening economy would likely hurt credit spreads, but the market also would likely anticipate increasing likelihood of easier Fed policy and lower LIBOR rates. The potential gains in IIO could help offset exposure in assets that do poorly in a softer economy.