Arlington Asset Investment Corp. (AI) is a real estate investment trust (REIT) whose primary activity is purchasing and holding mortgage investment portfolios. Its assets are mainly comprised of agency mortgage-backed securities (MBS) and mortgage credit investments to a lesser extent. It was established in 1989 and went public in 1997. AI has had a largely negative profitability for the four years ending 2019, having accumulated losses of $105 million during this period. The onset of the COVID-19 pandemic has only worsened this position, as the company reported net losses of more than $85 million for the six months ending 2020. The slump is attributable to market volatility, which is anticipated to worsen owing to political uncertainty with upcoming elections, a shaky housing market that is showing signs of a bubble burst, and the possibility that AI’s new business model may not be enough to see it through and after the current crisis. Therefore, AI is not a recommended buy owing to the significant increase in sustained market risks and the company’s historical run of losses.
Annual Performance
On face value, mortgage Real Estate Investment Trusts (mREITs) like AI derive value from acquiring highly levered agency MBS. In this regard, “agency” indicates that Arlington Asset Investment Corp.’s assets are primarily comprised of pass-through mortgage certificates whose interest and principal payments are explicitly or implicitly guaranteed by government bodies, such as Fannie Mae, Freddie Mac, and Ginnie Mae. Moreover, REITs like AI are highly levered, as they obtain asset purchase funding by pledging MBS as collateral in repurchase agreements (repos). Consequently, MBS appear both as assets and liabilities on an mREIT’s books. Repos have a maturity of a few days or weeks. This discussion is important, as it differentiates mREITs from other types of REITs that are primarily funded using equity (eREITs). Furthermore, mREITs are significantly different from mortgage-focused deposit-taking financial institutions owing to the intensive use of repos in funding asset purchases. These features significantly increase mREITs’ risk exposure, as evident in the case of AI.
Nevertheless, profitability is a principal aspect in assessing mREITs’ performance. Similar to other financial institutions, Arlington Asset Investment Corp. earns its profits through the spread between interest obtained through its assets and what it pays for its liabilities. This is the net interest margin (NIM). As shown in Figures 1, AI’s margins have been on a consistent decline over the four years ending 2019. This is attributable to a decline in industry rates on which MBSs are based, namely LIBOR and the Federal Funds Rate. These trends are shown in Figure 2. Conventionally, a decline in interest rates would have a positive impact on AI’s income, as it means that it is receiving more revenues than what is currently obtainable in the market. However, a decline in mortgage rates motivates homeowners (whose mortgages have been pooled to create MBSs) to refinance at the new lower rates. The result is that MBSs, while backed by the government and having an apparent similarity to treasury bonds, routinely underperform federal-issued securities. As Figure 3 illustrates, AI experienced growth in its Conditional Prepayment Rates (CPR) as rates declined. Alternatively, an increase in interest rates (as occurred between 2016 and 2018) means that AI’s MBSs are worth less in the market than when first purchased. These opposing impacts on mREITs’ income are termed as negative convexity.
Figure 1: AI’s interest margin trends
(Source: AI’s Annual Statements)
Figure 2: Fed Rate and LIBOR trends
(Source: FRED)
Figure 3: AI’s Conditional Prepayment Rate trend
(Source: AI’s Annual Statements)
These risks make it critically important that AI undertakes intensive hedging practices. These imply the possibility of incurring losses owing to the uncertain nature of the value of underlying assets, which are mainly MBSs. Arlington Asset Investment Corp. has incurred significant losses on this front, as shown in Figure 4. Notably, the company’s operating costs declined significantly from $20.6 million in 2016 to 2019’s $15.0 million. Consequently, investment losses arising from derivative trading and decline in the value of MBSs were solely responsible for AI’s slump in shareholders’ returns (Return on Equity (ROE)) and asset efficiency (Return on Assets (ROA)), as shown in Figure 5.
Figure 4: AI’s investment gains (losses) trend
(Source: AI’s Annual Statements)
Figure 5: AI’s return trend
(Source: AI’s Annual Statements)
Half-Year Performance
The onset of the COVID-19 pandemic has caused major disruptions to the housing market in the US. Specifically, a steep decline in both LIBOR and Federal Funds Rate, as exemplified earlier. On the one hand, this has led to an increase in CPR, with AI indicating that the rate nearly doubled to 19.67% for the six months ending June 2020 compared to what it was at the end of 2019. Additionally, COVID-19 has occasioned significant income losses for property owners. Combined with the imposition of new regulations restricting forbearance, the implication is that AI, as the holder and recipient of accruing principal and interest, has had to recognize substantial losses to its portfolio. Since the beginning of 2020, Arlington Asset Investment Corp. has reported $90 million of these losses, compared to $12.9 million outflows occasioned in a similar period in 2019. More importantly, this slump was on significantly fewer assets than in 2019. Specifically, it is apparent that AI has resorted to a fire sale of MBSs to boost its liquidity. This is evident in that the company’s assets declined sharply from $4 billion reported December 2019 to just $785 million in June 2020. Proceeds have gone towards meeting repo agreements with MBS dealers and purchasing MBSs under the new rate regime. There has also been a 15% decline in operating costs, which implies that AI let go of some of its employees. Despite these efforts, AI reported a half-year loss of $85.5 million in June 2020 compared to $7 million in 2019.
Overall, it is apparent that COVID-19 has dealt a major blow to AI’s already struggling profitability. The ensuing pandemic has also necessitated substantial downsizing of operations and a potential change in its business model. As of June 2020, AI’s MBS holdings had declined to just over half a million dollars, down from the $3.8 billion held six months prior. MBSs not backed by government agencies have also increased by 90.4% during the same period to $149 million. The implication of these developments is that AI may have reached the mezzanine level where its management considers currently held agency-backed MBSs to be adequate to sustain future operations and possible profitability. However, an enduring pandemic and clouds of political uncertainty hovering before the end of 2020 make the latter less unlikely.
Interest rates, both LIBOR and the Federal Funds Rate are at historic lows occasioned by the COVID-19 pandemic. The trend is expected to last for at least the next two years. At the same time, there has been a notable increase in mortgage defaults owing to mass income losses. Moreover, the Federal Reserve has engaged in massive asset purchase, especially from the financial industry, and currently holds more than $7 trillion assets, up from $4 trillion at the beginning of 2020. The combination of these three factors (low interest rates, income losses and mortgage defaults, and excessive liquidity) has historically been associated with bursts of the housing market. With regard to Arlington Asset Investment Corp., the implication is that the new business model may persist for longer than an investor may be willing to accommodate. Moreover, a housing crash might wipe off significant portions of AI’s balance sheet through write-offs when interest rates begin to increase. Consequently, the foregoing analysis informs a sell or do not buy decision for the rest of 2020 with regard to AI stock.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
