A little more room to go for prime ‘AAA’ 2.0

admin | May 17, 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.

The recent strong bid for prime private MBS 2.0 has led ‘AAA’ classes to significantly outperform comparable agency MBS, and ‘AAA’ may have a bit more room to run. ‘AAA’ pass-throughs still trade at a lower price and higher yield than agency jumbo MBS with better projected convexity. For portfolios that do not need agency liquidity or financing, the private MBS looks like better relative value. That should allow 2.0 to extend the last few years of good performance.

The case for relative value versus agency MBS

‘AAA’ pass-throughs have closed the price spread to TBA by roughly $1-00 so far this year, but they still have some advantages over agency jumbo MBS in price, yield and projected convexity. A recently priced ‘AAA’ pass-through serves as a good example.

SEMT 2018-3 A4 is a super-senior ‘AAA’ pass-through with characteristics comparable to Fannie Mae CK 3.5% pool FN MA3399. The pass-through trades at $1-18 behind TBA 3.5% while the pool trades $0-30 back. That leaves the price of the pass-through $0-20 behind the CK pool. The price difference gives the pass-through a projected incremental 11 bp of yield and 5 bp of base case total return. Despite having roughly a $170,000 larger average loan size, the pass-through has markedly better estimated convexity– -3.04 for the private pass-through with -4.40 for the CK. Private pass-throughs generally offer better convexity than  agency jumbo paper despite higher average loan balances likely because of greater friction for refinancing a non-agency loan. Given better convexity and lower dollar price, the private ‘AAA’ shows better projected total return than the CK in both a rally and a selloff (Exhibit 1).

Exhibit 1: Private-label pass-throughs can offer total return advantage over CKs

Note: Projected total return assumes a linear shift in rates to the 1-year horizon, reinvestment at LIBOR and constant OAS repricing. All market levels as of 5/14/19. Source: YieldBook, Amherst Pierpont

The CK pool will likely have a smaller bid-ask and easier access to financing than the ‘AAA’ pass-through, and that could outweigh the better ‘AAA’ price, yield and convexity for portfolios that need agency liquidity. Each portfolio will need to weigh the tradeoffs.

Risk and return in prime 2.0

Prime ‘AAA’ pass-throughs have not always topped the performance charts for prime 2.0 (Exhibit 2). Within prime private label, some parts of the cash flow and capital structure have offered better risk adjusted return than others. Given their shorter duration, hybrid super seniors have provided better risk adjusted returns than fixed rates. ‘AAA’ mezzanine bonds backed by hybrid ARM collateral have offered attractive risk adjusted returns as well as a result of relatively wider spreads, attractive carry and pristine collateral performance. Across the subordinate stack, non-investment grade BB and B bonds have offered the most attractive risk adjusted returns also driven by carry and collateral performance. However, those bonds trade in limited size and frequency and as a result, risk adjusted returns may be modestly overstated.

Exhibit 2: Risk and return across prime 2.0 April 2017-April 2019

Note: We then scan the entirety of the PLS cash flow and capital structure to see what bonds have offered the best risk-adjusted returns over the past two years. We cut the ‘AAA’ classes by both coupon type and cash flow structure, splitting the ‘AAA’ bonds into fixed rate and hybrid cash flows. We then further cut the bonds by cash flow features separating principal and interest bonds from derivatives and further separating P&I bonds by structure. ‘AAA’s are also cut into super senior and senior mezzanine bonds, which trade at a spread concession to super seniors. We cut the subordinate stack by original rating and look only at fixed rate bonds. We also calculate a simple average return on a passive investment in the entire prime cash flow and capital structure. Simply comparing risk adjusted returns on super senior AAA pass-throughs to the Citi Mortgage Index over a two-year horizon, pass-throughs have offered a 5 bp better nominal return than the agency MBS index, averaging 24 bp per month. The pass-throughs offer a better risk adjusted return over that horizon as well with a Sharpe ratio of 1.55 versus .94 for the agency index. Obviously, there are some shortfalls to this methodology, as discussed earlier, private label pass-throughs are generally shorter duration than their agency counterparts due to the shifting interest structure and as such the duration relative to the index will in all likelihood not be aligned. Additionally, agency MBS is far more liquid with more transparent pricing, given the liquidity give in PLS there may be some dampening of potential price volatility. However given the fact that these are monthly returns and super senior AAA pass-throughs are arguably the most liquid and transparent asset in the private label market, this effect should be negligible. Excludes IO and Inverse IO. Source: Intex, IDC, Amherst Insight Labs, Amherst Pierpont Securities