$1.32
$7.73
bear $1.32
base $3.93
bull $7.73
margin of safety $2.00
market price* $3.46
Business condition attached to the margin of safety
Reports filed after the April 2025 conversion show that income from the CLO portfolio, after fees and financing costs, covers the $0.24 quarterly dividend without relying on realized hedging gains, and that book value per share has stopped falling.
Ellington Credit Company (formerly Ellington Residential Mortgage REIT) has just changed its business. Until early 2025 it was a leveraged holder of government-agency mortgage bonds. On April 1, 2025 it became a registered closed-end fund managed by Ellington's adviser, and it now invests mainly in the equity and mezzanine tranches of corporate CLOs (pools of leveraged corporate loans). At March 31, 2025 its CLO portfolio was $249.9 million, up 46% from $171.1 million at year-end 2024, against $228.5 million of shareholder equity (about $6.08 per share on 37.6 million shares) and $517.5 million of repurchase-agreement borrowings. This was the company's last annual-style report under the old rules, so we have no later financial statements. Everything here is based on data through March 31, 2025 and should be read with that in mind. How we measured earnings. We start from operating cash flow, remove working-capital swings and subtract share-based pay. The company has no meaningful capital spending, because securities purchases are investments, not capex. On that basis owner earnings were $9.7 million in 2024, or $0.41 per share, and negative $7.3 million (minus $0.49 per share) in 2023 when the old mortgage book was still running. In the three months ended March 31, 2025 they were $7.6 million, about $0.22 per share for the quarter. The 2023 figure describes a different business, so we set it aside. We could not use a long history because the strategy changed in March 2024, and 2024's per-share figure is held down because the share count grew from 18.6 million to 29.7 million during the year and the new money was only partly invested. We set the starting level at $0.55 per share a year, above 2024 but well below the first quarter's $0.87 annualized pace, because a single quarter that still included mortgage-bond income is not enough to move the level. The main concern is that per-share book value has fallen every year: about $8.40 at the end of 2022, $7.32 at the end of 2023, $6.53 at the end of 2024 and $6.08 at March 31, 2025. The company has paid $0.96 per share in dividends each year while net income was $0.31 and $0.28 per share in 2023 and 2024. The excess is a return of capital, and the income a smaller pool of assets can earn shrinks with it. That decline is the named, evidenced driver in every case, and we credit no growth. New share issuance was done near book value, so it adds capital but is not treated as growth in per-share earnings. Owner earnings grew from negative to positive in 2024, which is not a meaningful growth rate; our first-year assumption is a decline, not an increase. Base case, about $3.93 per share: owner earnings of $0.55 per share fade about 8% a year, in line with the average pace at which book value per share has been eroding, and the decline continues because nothing we can point to stops it.
Bear case, about $1.32: earnings of only $0.25 per share (roughly the 2024 level less the cost of the new fee structure and higher CLO financing costs) and a 13% annual fade, which is the actual pace of book value decline since 2022. This fund also borrows heavily, roughly 2.3 times equity at March 31, 2025, and is subject to the 300% asset coverage limit that can stop dividends if asset values fall.
Bull case, about $7.73: the first-quarter earnings pace of about $0.85 per share holds, and the fade slows to 5% a year. We left out growth from further CLO purchases and from any recovery in CLO values. Neither has been demonstrated under the new structure, and the fund raises capital at prices tied to the market, not to a fixed schedule. The new advisory agreement costs 1.5% of net assets a year plus 17.5% of net investment income above an 8% annual hurdle. That is a real cost that was not fully in the 2024 figures, and it is why the bear case starts lower. What would make this wrong: if the CLO book earns enough after all costs to cover the $0.24 quarterly dividend and book value stabilizes, the stream would be closer to permanent than our base case assumes. We could not verify that from filings, since no later financial statements were available to us. The 2023 to 2025 figures are also affected by large hedging gains and losses that the cash flow measure leaves out, so actual per-share results can differ from these figures.