Armour Residential REIT (NYSE:ARR) is a mortgage REIT. It is managed externally by Armour Residential Management LLC. ARR pays a monthly dividend/distribution. For the three months of Q4 2012 that dividend was $0.09 per share per month. For the three months of Q1 2013 that dividend was $0.08 per share per month. For the three months of Q2 and Q3 2013, that dividend was $0.07 per share per month. For Q4 2013 ARR cut its dividend again to $0.05 per share per month (about 14.78% yield at the current stock price). Is this a continuation of a bad trend? I don't think so. Rather this probably represents at least a near-term bottom. The rise in interest rates and mortgage rates this year almost certainly means that the net interest spread (the main basis of the dividend) will be wider. This should mean more income per MBS. In addition the calming of the excess volatility of Q2 in Q3 should further add to the net interest spread. On top of this ARR's portfolio CPR (constant prepayment rate) has fallen dramatically in Q3 2013 as shown in the chart below.
One would have thought ARR would have done well in Q3 2013. The portfolio's average CPR (constant prepayment rate) decreased dramatically from 11.6 in June 2013 to 4.9 in October 2013. ARR's portfolio should have gained value on this basis alone. On top of that the fixed rate 30 year FNMA 3.5% RMBS ended Q3 roughly flat with the end of Q2 2013. Since the 10 year US Treasury Note yield was up from 2.49% on June 28, 2013 to 2.61% on September 30, 2013 (+12 bps), ARR should have profited slightly on its hedges. One would have expected the company to gain book value. Instead it lost book value from $5.43 per common share as of June 30, 2013 to $5.26 per common share (-$0.17 or about -3.1% per common share). With the dividend at $0.07 per share per month in Q3 2013 ($0.21 for the quarter), investors ended up slightly ahead (+$0.04 per common share or +0.7% quarterly and +2.9% annually). In this sense one could say that the headlines about ARR's earnings are highly misleading. Investors did not lose money.
The headlines said ARR had Core Income of $43.8 million or $0.11 per common share. ARR had a Q3 2013 GAAP net loss of approximately $229.9 million or $0.63 per common share. It reduced its leverage to 6.93x on September 30, 2013 from 9.77x on June 30, 2013. It saw sales of Agency securities in Q3 total $6.0B. The sales resulted in realized capital losses of $301.0 million or $0.81 per common share.
How did ARR manage to lose so little book value per share? The simplest explanation is that these losses were mostly accounting hand waving due to GAAP rules. Much of the actual losses probably occurred in quarters before Q3 2013. However, they were not accounted as capital losses until the securities were actually sold. Plus the "offsetting" hedging gains were mostly not redeemed in Q3 2013. Thus they did not offset the capital losses, but they did continue to be present in book value calculations (as they should have been). Investors who are paying a lot of attention to the headline numbers are likely confused.
If ARR should have made money in Q3 2013, how did they lose money? The company did not make a big point of explaining this in its Q3 earnings announcement. However, the two tables below as well as the total of $6B in Agency securities sold in Q3 provide the likely answer.
The portfolio description as of July 15, 2013 is below.
The portfolio description as of October 4, 2013 is below.
The biggest change is clearly the amount of the ARM and Hybrid ARM securities held as of October 4, 2013 ($265 million) versus the amount held as of July 15, 2013 ($989.5 million). The ARR announcement didn't describe the unexpected losses. However, the Q3 2013 Capstead Mortgage Corp. (NYSE:CMO) earnings report contains the answer. CMO is a primarily Agency ARM and Hybrid ARM mortgage REIT company.
As mortgage and interest rates continued to rise in the first two months of Q3 2013, the CPRs (constant prepayment rates) of CMO's Agency ARMs and Hybrid ARMs pushed over 26% in both July and August. CMO's overall portfolio CPR averaged 25.5% for Q3 2013. ARR's ARM portfolio's CPRs were generally less high, but they still skyrocketed as many worried mortgage rates were going to keep going straight up. In its September update, ARR cited the ARMs and Hybrid ARMs portfolio CPR as 23.8. This is much higher than the 18.8 cited in its October 4, 2013 information. In other words ARR experienced roughly the same CPR rise as CMO in Q3 2013; and it is seeing roughly the same fall in October.
ARR's ARMs portfolio's average CPR rise in Q3 2013 caused the Agency ARMs to lose value during the quarter. For CMO, which did not sell many of its ARMs, it benefited in October from its ARM's CPRs going back down dramatically. CMO reported an ARMs CPR of 17.1 in October 2013. It reported that they had gained significant book value in October 2013. It sold only a modicum of its ARMs in Q3 2013. In contrast, ARR sold about 73% of its Agency ARMs during Q3 2013. This has proved to be a bad management decision. A significant amount of book value was lost unnecessarily by these actions. If you want to think less of ARR's management for this, you may be justified.
ARR's other losses likely came from the approximately $5B fixed rate Agency RMBS that it sold and or turned over in Q3 2013. Remember it did reduce its leverage from 9.77x to 6.93x during Q3 2013. This is sure to have negatively impacted the net interest income too. However, the decrease in leverage should add to ARR's safety going forward.
ARR also sold some 30 year fixed rate Agency RMBS in favor of 15 year fixed rate Agency RMBS. However, its percentage changes in this area were small compared to other primarily Agency RMBS mortgage REITs. ARR's portfolio descriptions for July 15, 2013 and October 4, 2013 show that the percentage of 301 to 360 month fixed rate Agency RMBS only shrunk from 64.4% to 63.3%. The percentage of 181 to 240 month fixed rate Agency RMBS only increased from 29.5% to 30.5%. By comparison American Capital Agency Corp. (NASDAQ:AGNC), a company many people consider a leader in the mortgage REIT space, moved its 30 year fixed rate Agency RMBS and 15 year fixed rate Agency RMBS percentages from 56% and 42% respectively as of June 30, 2013 to 43% and 52% respectively as of September 30, 2013.This is more than an order of magnitude greater change than ARR's change. The reason AGNC gave for its strategy change was that the 15 year fixed Agency RMBS's lose premium (book value) at roughly half the rate of the 30 year fixed rate Agency RMBS during a large move up in interest rates and mortgage rates. ARR is not preparing for this. It could lose huge amounts of book value in the future in much the same way as it did in Q2 2013 with its current portfolio breakout.
The above is the negative way to look at things. A slightly better way to look at the ARR picture is that their management team sees a slowing economy. They see a very easy Fed policy continuing for some time. Many non-ARR pundits agree that the Fed may refrain from tapering until at least March 2014. If this scenario does play out, it gives ARR plenty of time to exchange more of its 30 year fixed rate Agency RMBS for 15 year fixed rate Agency RMBS. In fact under this scenario, ARR may have made a wise decision in not doing this exchange more expensively in Q3 2013.
The reinstated payroll tax, the sequester of March 2013, the government shutdown and default scare, the new sequester cuts of $109.3B expected January 15, 2014, the added costs of Obamacare that will decrease the average US citizen's discretionary income (and, thus, hurt the discretionary retail sector), and any tax increases the Democrats succeed in negotiating are all reasons for believing that the US economy should slow in Q4 2013 and 1H 2014. If there is another negotiation debacle in early 2014, the slowdown could be even worse. If this slowing plays out, ARR's management may well have made a good decision about its portfolio percentage makeup. The 30 year fixed rate Agency RMBS do have higher yields. This could lead to higher profits for ARR, if ARR's management made the correct decision. Perhaps ARR is right about in the short term in this case. However, ARR undeniably lost more money than it needed to by its untimely sales of its Agency ARMs and Hybrid ARMs. Investors will have to make up their own minds about ARR management's performance.
Looking forward, ARR should have at least two quarters of good to decent results in Q4 2013 and Q1 2014. It may have many more. We will have to wait to see what it does about its large amount of 30 year fixed rate Agency RMBS for the long term. However, it cannot lose much more due to its ARMs for the simple reason that it no longer has many ARMs. The next six month outlook is probably good to great. At the October 30, 2013 closing stock price of $4.06, it would have to gain $1.20 (almost 30%) to reach its book value as of September 30, 2013. Since its book value is more likely to rise in Q4 2013 than to lose value, it could potentially gain even more. The net interest spread of 1.24% for Q3 2013 was slightly below that of the 1.38% average net interest spread in Q2 2013. However, I am inclined to think much of that was due to the extremely high CPRs of the ARMs in Q3 2013. The high CPRs no doubt subtracted significantly from the net interest spreads of the ARMs in Q3 2013.
There is also the problem of the lower leverage rate. This will lead to lower net interest income than the higher leverage rate would have allowed. However, the dividend rate has already been lowered from $0.07 (in Q2 and Q3) to $0.05 per common share per month for Q4 2013. This is still a great 14.78% yield at the closing stock price on October 30, 2013. Further the lower CPR rate of 4.9 for the overall portfolio as of October 4, 2013 is much lower than the 8.8 Q3 average CPR. If this average portfolio CPR persists it should add to ARR's net interest spread in Q4 2013. The current dividend may be maintainable. With all of the above in mind ARR is a marginal buy.
ARR's Management made some mistakes in Q3 2013. However, it literally cannot repeat some of those mistakes; and it may get a chance to beat AGNC's results in Q4 2013 and Q1 2014. I would still watch for how ARR positions itself for the future of rising interest rates. These will eventually appear. Still ARR has the opportunity to follow leaders such as AGNC, if it so chooses.
The two year chart of ARR provides some technical direction to this trade.
The slow stochastic sub chart shows that ARR is neither overbought nor oversold. The main chart shows that ARR has been in a downtrend. It reversed its upturn on its Q3 2013 earnings report. However, the headlines of that report were misleading. The fundamentals for ARR are much better than the headlines suggested. The chart and the fundamentals suggest that ARR should continue its recovery from its recent bottom after investors get a chance to sift through the data more carefully. ARR has an average analysts' recommendation of 2.6 (a high hold). It probably deserves a bit better than that. In the currently overbought and overpriced market it should be at least a low buy. However, investors should watch management's moves carefully if they buy this stock.
NOTE: Some of the above fundamental financial data is from Yahoo Finance.