Reaching for yield in Freddie K deals

admin | July 19, 2019

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. This material does not constitute research.

The front end of the curve is steeply inverted and financing rates remain higher than yields on most Treasury and highly rated bond issues. Many investors are in the now familiar place of reaching for better performance or yield by either going down in credit or accepting prepayment risk. One method for increasing yield in agency products without incurring prepayment risk for most of the life of the bond, is by investing in B and C classes of Freddie Mac K deals. These pieces are not guaranteed by Freddie Mac, but are typically rated by one or more rating agencies, and often benefit from increasing credit support as loans are defeased over time. Projected total returns vary considerably depending on whether investors expect the very aggressive path of forward rates will be realized.

An aggressive path forward

Market expectations recently priced in a higher probability of rate cuts, thanks to some very dovish commentary by both New York Fed President John Williams and Federal Reserve Vice Chairman Richard Clarida. Spokesmen at the Fed immediately tried to walk back the remarks, stating they were not intended to reflect potential policy actions at the upcoming FOMC meeting. The clarification by Fed spokesmen reversed some of the rally, but pricing remains on the high side, with several analysts and economists having updated their calls for a 50 bp cut at the July meeting.

Exhibit 1: Par coupon and implied forward swap curves

Note: Curves as of 7/19/2019. Source: Bloomberg, Amherst Pierpont Securities

Away from fed funds futures, the current par coupon and implied forward swap curves (Exhibit 1) are consistent with 50 bp of cuts over the next three months, and at least 75 bp of rate cuts over the next year. The implied forward swap curves pivot around the 5-year point, with the front end of the curve bull flattening, as the long end of the curve mechanically steepens by a few basis points over the next year.

Reaching for yield in agency CMBS

The front end of the curve is steeply inverted and financing rates remain higher than most Treasury and highly rated bond issues. Many investors are in the now familiar place of reaching for better performance or yield by either,

(1) avoiding the belly of the curve via a barbell strategy – being long the short-end and long-end of the curve in duration- and proceeds-equivalent basis;

(2) betting on a reshaping of the curve via a butterfly – long the barbell as above, short the belly, on the presumption that forward rates will not be achieved, because the Fed will not cut rates as aggressively as currently priced into the curve;

(3) reaching for additional yield by going down in credit or accepting prepayment risk.

One method for increasing yield in agency products without incurring prepayment risk for most of the life of the bond, is by investing in B and C classes of Freddie Mac K deals. These pieces are not guaranteed by Freddie Mac, but are typically rated by one or more rating agencies, and often benefit from increasing credit support as loans are defeased over time. In seasoned deals if the credit support or defease-adjusted credit enhancement increases, a B or C tranche can climb to higher ratings over time (Exhibit 2).

Exhibit 2: Comparative data of Freddie Mac K B and C pieces of increasing weighted average life

Note: All levels as of 7/19/2019. The “H” used by DBRS means “high”, which is analogous to “+” used by S&P or Fitch. Source: Yield Book, Bloomberg, Amherst Pierpont Securities

The projected total returns of the various Freddie K pieces are evaluated assuming forward rates are achieved, along horizons from 1-month to 24-months out (Exhibit 3). The shortest bond is not shown, since it is projected to pay-off in 6 months. The total returns of the other bonds look broadly as expected, with a few dynamics worth noting. The forward swap curves pivot roughly around the 4.5- to 5-year maturity, so the bonds with longer weighted average lives experience very modest, gradually rising rates, which competes against the roll-down as their maturities shorten over the time period. All of the bonds are at price premiums and exhibit some pulling to par over the time period. The FREMF 2017-K61 B with a 7.4 WAL is projected to modestly underperform the FREMF 2013-K27 B with a 3.5 WAL over longer time horizons on a total return basis (Exhibit 3, shown in graph on right hand side). These two bonds bracket the pivot in the swap curve, with the shorter bond eventually rolling down the curve as the short end rallies and the curve normalizes over the two years.

The FREMF 2014-K714 B has a WAL of 1.25 year. Once that principal is returned its projected total return drops sharply, but that’s simply a result of being reinvested at the 3-month Treasury bill rate, which by that horizon is somewhat below 1.75% (Treasury forward curves not shown).

Exhibit 3: Total rate of return of Freddie K B and C pieces along implied forwards from 1-month to 24-month horizons

Note: All pricing and scenario analysis done in Yield Book as of 7/19/2019. Projected total rates of return are on an annualized basis for each horizon. The model assumes 0 CPY, that spreads are constant across the horizon, and that all coupon and principal payments are reinvested at the 3-month Treasury bill rate. The reinvestment assumption is Yield Book’s default, and does result in depressing the projected rates of return over longer horizons when the reinvestment rate is below a bonds yield, which in this case it is for all of the bonds.

 Expecting the unexpected

Forward rates embed market expectations in the front end, and are arguably mechanical extrapolations of those expectations at longer tenors and over longer horizons. Often the best investment opportunities exist for those who position for a different economic or market outcome than what is priced in to the forwards, and turn out to be correct. Or lucky. Some alternative curve scenarios to the implied forwards are bull and bear flattening and steepening scenarios (Exhibit 4).

Exhibit 4: Alternative curve scenarios – bull and bear steepeners and flatteners

Note: Curve scenarios are default bull and bear steepening scenarios in Yield Book of the swap curve. The curve moves are derived using principal component analysis based on the current curve. Source: Yield Book, Amherst Pierpont Securities.

The projected total return analysis is rerun for the bull and bear flattening and steepening scenarios over a 12-month horizon(Exhibit 5). The projected rates of return for the implied forwards over 1-year are shown in the middle of the graph as the base case.

Exhibit 5: Projected total rate of return over 12-month horizon for bull and bear shift scenarios

Note: Scenario analysis assumes gradual shifts over a 12-month horizon; all bond spreads are assumed to stay constant and coupons and principal are reinvested at the 3-month Treasury bill rate. Source: Yield Book, Amherst Pierpont Securities

Not surprisingly, all of the bear curve scenarios have lower projected rates of return for all of the bonds than in the bullish implied forward case, since the entire curve sells off to varying degrees from the current swap curve. The lowest projected returns are for the 100 bp bear flattening scenario, though the two shortest bonds show only modest underperformance compared to the base case and their projected total returns remain in positive territory. The longest bonds do the worst in the bearish scenarios and have the best performance in the bullish ones, and the FREMF 2013-K27 B with the 3.5-year WAL roughly splits the difference.

None of this analysis is surprising – the protection from prepayments provided by lockout and defeasance provisions in Freddie K deals, high levels of credit support and negligible history of defaults, the Freddie K deal B and C classes behave like regular fixed rate bonds in these scenarios. To be clear, there could certainly be a scenario where high involuntary prepayments occur, eroding the credit support, widening spreads and perhaps resulting in losses. Or rent growth trends could slow, leading to slower defeasance overall. Given the underlying strength of the residential real estate market and still historically low unemployment, that level of deterioration seems unlikely in the near term in multifamily commercial real estate, making Freddie K’s an attractive space for moving down in credit.