Ellington Financial LLC (EFC) is a specialty finance company that invests in mortgage related assets, primarily non-Agency RMBS and Agency RMBS. Its closest comparables are Chimera (CIM), which is an affiliate of Annaly (NLY), MFA Financial (MFA), Two Harbors (TWO), Invesco Mortgage (IVR), Dynex (DX), American Capital Mortgage (MTGE), AG Mortgage (MITT), Apollo Residential (AMTG), and Redwood Trust (RWT). However, the comparison is not ideal since most of the comparables are REITs while EFC is effectively a publicly traded partnership. The advantage to EFC of its structure is that it doesn’t have to satisfy REIT requirements, which means that it is able to more actively trade, invest in a broader range of assets, adopt a more flexible dividend policy, and most importantly, hedge.
If anyone doubts the benefits of this last point, compare the stock prices (or better, the BV per share) of CIM, DX, MFA and RWT to the reported returns of EFC from August 2007, when EFC was started, to the present. Over this time period, total return for EFC has been about 60%. An investment in those comparables (they are the only ones that have been around as long) would have resulted in few gains and some ugly losses (CIM, RWT). Further, an investment in an ABX index of 2006 or 2007 vintage would have resulted in significant losses.
We believe this strong investment performance is due to the superior security selection of EFC’s external manager which is run by Mike Vranos, formerly of Kidder, Peabody and an erstwhile bodybuilder. We think that EFC’s outperformance should continue and, especially if you think that so called “risk assets” have had a pretty good run recently, then EFC’s hedges should provide some downside protection. Lastly, we think that EFC should merit a higher multiple to book than it currently has. EFC languishes at a 15% discount which makes it cheaper than all of the comparable agency and hybrid mortgage REITs. This is a major point of frustration for Vranos, who is making a concerted effort to increase EFC’s exposure in the investment community and, last year, implemented a buyback program when he thought the market price of EFC was at too big of a discount to its book value per share.
Excluding financing arrangements as of December 31, 2011, EFC had about $1.3 billion of long mortgage related investments and about $500 million of short. The non-Agency portfolio, $450 million long and $180 million short, is the most important component of the Company’s earnings. EFC’s book value was about $370 million or $22 per share.
The non-Agency portfolio consists primarily of pre-2005 subprime securities where borrowers have home equity (39%), more recent vintage subprime and Alt-B securities where borrower performance has improved and EFC was able to buy at attractive prices (28%), senior Alt-A jumbo loans (10%), seasoned manufactured housing loans (14%) and other (9%). Generally, EFC takes a conservative approach to security analysis and asserts that its purchases should do well even if home prices decline another 15%. After meeting with Vranos and EFC management, it is clear that they have performed a detailed analysis of (almost) every security in the RMBS space and that they have a good feel for what is cheap, what should get paid at maturity, and what to stay away from. They are very quantitative in their approach and their qualifications reflect this obsession with numbers: Math Majors, Masters and PhDs.
Notably, the long portfolio average price paid is about 56%, which gives the potential for significant capital gain within the portfolio. The estimated yield with flat home prices is 12.3% and with home prices declining 15%, the yield drops to 10.7%. In this latter case, however, EFC’s short portfolio should at least make up some of the difference.
On the short side, EFC bought protection on a number of ABX indices backed by dodgier borrowers with higher delinquencies and single name CDS reflecting the same (30%), equity swaps on publicly traded REITs (11%), and corporate CDS (58%). This last “hedge” backfired somewhat in late 2011 as non-Agency RMBS prices declined while corporate spreads tightened. It is a concern to us that this happened. However, we think the trade makes more sense now than when it was set up.
On the long Agency side ($732 million), EFC targets “prepayment-protected pools such as those comprised of low loan balance mortgages and those containing mortgages not eligible for one of the government sponsored refinancing programs.” As of December 31, 2011, the 3-month historical CPR on the Agency fixed rate pool that EFC owned was 4.6%. On the short Agency side ($207 million), EFC primarily target interest rate risk using swaps and TBA (To-Be-Announced or “generic”) Agency securities.
EFC is currently paying 40 cents per quarter or $1.60 annually for a yield of over 8%. The company’s policy, however, is to pay out 100% of its earnings. This has given rise to an interesting predicament. Owing to declines in RMBS values in the second half of 2011 and as a consequence of EFC’s mark to market accounting flowing through the income statement, the company’s earnings did not cover the dividend payments for 2011. However, EFC should more than make up for the shortfall in 2012 where earnings should hit some $3.00 per share. Note that owing to mark to market accounting, earnings are difficult to predict for EFC. EPS was $7.53 in 2009 owing to large unrealized gains. It was $2.89 in 2010, but without as many gains. We view the dividend as secure which is important given the potential significance of retail ownership for this stock.
Reasons to Own EFC
Talented management – Vranos and his team have delivered strong returns for almost five years now and protected capital through the worst downturn in real estate in generations.
Hedging protects against tail risk shocks (another downturn in home prices or significant increases in credit spreads) and interest rate increases – comparable REITs have a limited ability to do this.
Valuation – as of February 29, 2012, EFC’s diluted book value per share was $22.91. With a current market price of $19.50 the discount on EFC makes the stock the cheapest in its sector. Indeed, many comparables trade at a premium to book (CIM, MFA, TWO, IVR, DX, MTGE, MITT).
The current dividend (8%+ yield) and buyback program should provide support for the stock.
Continued execution of strategy and growth in book value.