The main body of this article was published on 05/22/2016 for subscribers. That was a Sunday, so share prices referenced in the article were from Friday 05/20/2016. There is a new conclusion added to this article to demonstrate performance since publication. The conclusion added section at the bottom also includes new information about the new difficulty of borrowing the preferred shares, which puts an end to the suggested pair trade.
Jumping from ARR-A to ARR-B offers a nice boost to income along with better call protection. Beware the spread, as these shares can be a little illiquid. As one extra disclosure: I had no position in ARR or any related security when I wrote the article. In early June, I took a long position in ARR with the intent to hold the position for a period measured in days or weeks. I may sell the common shares of ARR at any time.
*Below this line is the original article*
ARMOUR Residential REIT (NYSE:ARR) is a mortgage REIT with two series of preferred shares. The different series carry different coupon rates, trade at different discounts, and have different dates for becoming call eligible. ARR is a mortgage REIT with a market capitalization of about $700 million at the present time. The two series of preferred stock combine to have a market capitalization that is substantially smaller. It should be no surprise that there is very little high quality analysis in this little niche, but it is precisely this lack of analysis that allows market failures to occur and to persist for prolonged periods of time.
Since this is my first article for subscribers, I want to emphasize that nothing in any article or other form of communication should ever be taken as investment advice. I do not know the readers' individual circumstances and like all analysts, I will occasionally make mistakes. It is also prudent to point out that my analysis is limited to my data sources, so before executing a trade, each investor should perform their own due diligence and verify the numbers. For instance, in this article, I will be gathering the call dates from preferredstockchannel.com.
The Preferred Shares of ARMOUR Residential REIT
This article is not about the common stock of ARMOUR Residential REIT. This piece is highly focused on a very actionable trade regarding the two series of preferred stock. Different brokerages may use different symbols, so I will demonstrate the tickers.
To demonstrate how this has happened before, I am also going to include the preferred shares for American Capital Agency Corporation (NASDAQ:AGNC). The following table highlights several key metrics.
I highlighted the current yield on the preferred shares and the ratio of the current price on preferred shares to the par value. These two securities shouldn't trade in absolutely perfect unison, but the yields should be much closer.
How Material is the Difference?
The difference was large enough I opened up my brokerage account to verify that the last prices shown there matched the values I had retrieved from another source. The values did match. If ARR-A is $23.98 and ARR-B is $21.81, then I believe ARR-B is a materially better investment.
Advantages to ARR-B Over ARR-A
If the investor is concerned about call protection, ARR-B is protected until 2/12/2018 while ARR-A is only protected until 6/7/2017. Given the shares trading at a material discount to par value, I don't see the call protection as being highly valuable, but it does provide a little additional upside if the market decided to value the preferred shares of mortgage REITs at high levels.
No Parity Difference
Due to the material difference in yields, I checked the prospectus for the B-shares to see if there was any difference in the seniority of the claims. Page S-5 indicates that the two issues will have the same level of priority.
Better Buyout Options
In the event of a buyout, the preferred shares can (at the company's choice, not yours) be called at par value plus accrued dividends. This is demonstrated on the first page of the prospectus for the A-shares:
I am not a lawyer and do not provide any legal advice. It is simply my expectation that the phrasing there indicates the timing of the change of control is irrelevant to the company's right to force a call in the case of a buyout. For shareholders of ARR-A, that limits the ability of the shares to move materially above par value, and if NLY or AGNC was the buyer, I would expect those preferred shares to get called off immediately.
The ARR-B shares have a lower coupon rate at 7.88% (rounded). The rate is still high enough that big mREITs might still opt to pay it off on the belief that they could still issue at more favorable rates. However, if the ARR-B shares were getting paid off at par value, it would be a huge gain.
I want to emphasize that I am not predicting a buyout. I am simply pointing out that if a buyout did occur at some point in the following years, the ARR-B shares would be capable of dramatically better gains than the ARR-A shares. In assessing the potential benefits of one series relative to the other, this is a pretty nice advantage to have on shares that already have a materially better yield.
I expect ARR-A to underperform ARR-B on share prices by about $1.00 per share over the next few weeks as this gets adjusted. For ARR-A to see their yield increase to match the yield on ARR-B, they would need to lose over a buck off the share price. On the other hand, ARR-B could also increase by about a buck. A completely rationale pricing of the two issues would suggest that ARR-B should have a slightly lower yield due to the other benefits it offers (call protection and discount to par) and therefore a full correction would suggest ARR-B outperforming ARR-A by about $1.25 per share. The current difference in price is $2.17 based on the closing values from Friday May 20, 2016.
Tax Implications Are Not Considered
The suggested plays here will disregard taxes as each investor may have a different tax situation.
Ways to Play This Opportunity
There are a couple of opportunities here. One is for investors who are currently holding ARR-A to sell their position and jump into ARR-B to secure a materially better yield. They can order the trade in a similar manner to a pair trade if their broker offers the right software. This is a two-leg trade in which the investor asks to buy shares of ARR-B and sell shares of ARR-A so long as the net difference in the prices is at least $2.17.
If the investor does not adjust for the expected difference in the number of shares, their income would decrease as a function of owning the same number of shares but with a different coupon rate. To use this trade to increase the income on the portfolio, the investor needs to buy at least 104.762 (rounded up) shares of ARR-B for each share of ARR-A sold.
The Income Trade
The following table demonstrates the expected trade based on the difference of $2.17 in the share price.
For the income investor using ARR-A, a switch to ARR-B with the $2.17 difference in share price creates an increase in annual income of $42.59 per $10,000 in the trade. Perhaps you weren't signed up for this service right when it came out and the spread in prices has already declined. The next table demonstrates the impact if shares of ARR-A declined from $23.98 to $23.48 prior to the trade with no change in shares of ARR-B.
The table above still includes an increase in annual income of $24.28 for moving $10,000 from ARR-A to ARR-B.
The Pair Trade
Establishing the base for calculations in a pair trade can be tricky. If the investor is concerned about sector exposure, they could use the difference between the long and short positions. However, if the amounts are equal, it would result in infinite rates of return by those calculations. Instead, I'll be basing the calculations on the "Total Cash Suggested."
Different brokerages will have different rules, but I will be assuming a margin account that allows 100% leverage. I'm also assuming that there is no interest on the account so long as there is sufficient cash to cover the short position. No trading costs are incorporated since those costs will vary by trader and broker.
The following table lays out the pair trade:
This position simplifies assessing the gain or loss on the trade by eliminating all factors except the monthly net cash flows on the dividend and the difference in share prices. To avoid breaking 100% leverage, the investor will need their capital in the account to be worth at least as much as the short position. I would suggest having an extra $1000 sitting in cash to avoid running into interest if the shares of ARR-A move up slightly before this whole thing settles.
That means the investor's balance needs to be at least $24,980 (assuming 100% leverage allowed and $1000 extra to prevent margin interest) to execute this strategy. From that $24,980 they would spend $21,810 to buy 1000 shares of ARR-B. The difference in dividends means this position will cost them $7.81 per month. If they were spending as much on buying ARR-B as the put into shorting ARR-A, the dividend payments would be a positive. However, the annual effective interest from holding the position open would not be a sufficient reward for holding the position so it would still be a play on capital gains.
Assuming the yields move to be very near parity, there would be about a $1 swing in the relative share prices. That would suggest that the trader could close the position with a capital gain of $1,000 before accounting for any trading costs or net dividends. Assuming that it took 3 months for the yields to converge, there would be 3 monthly dividend payments working against the trader.
That would only bleed out around $23 to $24, so the profit (before trading costs) would still be about $976 to $977. The resulting return on the $24,980 used to create the trade would be 3.91%. Annualized this would be 16.58%. For the pair trader buying this subscription on an annual basis, the $976 more than covers the first two years. I'm assuming any pair trader is going to classify the subscription as a business expense.
The Price Difference
I pulled the close values from almost every trading day (a couple of days were unavailable) going back to June 30, 2014 for ARR-A and ARR-B and charted the difference in price.
Remember that I suggested a difference of about $1.17 would be reasonable for similar yields and that getting to a difference of $1.17 would require prices converging by $1.00. Over this entire measurement, the average difference in closing share price was $1.21 (rounded to nearest cent). The sudden movement can be traced back to early October 2015. The difference on October 1, 2015 was $.53. On October 30, 2015, it was $2.52. On October 16, 2015, the daily recorded volume on ARR-B was 226,345. This was 1096% of the average daily volume for the entire measurement period. ARR-A had lower than average volume that day and a little over 212% of their average daily volume the next day.
The Yield Difference
The following chart demonstrates the difference in the yields over that time period:
As you can easily see, the yields were trading at very similar levels to each other up until late 2015 when they became disconnected.
The average volume so far in May has been 5,307 shares for ARR-A and 15,481 for ARR-B.
There was a similar case, though to a lesser extreme, that recently happened with preferred shares of AGNC. This is the other mortgage REIT I referenced near the top of the article. The following chart only goes from the middle of 2014 to the present, rather than the middle of 2013 to the present.
The difference in yields was largely able to persist for the better part of the measurement period. It was only recently that the yields converted and AGNCB moved to have a lower yield than AGNCP. In simple terms, AGNCB's price outperformed since the dividend rate remains unchanged.
From the start of 2016 through the close of trading on May 6, 2016, the current yield on AGNCB was higher by .21%. In a similar case, AGNCB offered materially better call protection and a higher yield. The next trading was May 9th. Using the closing values from May 9th through May 20th shows that on average AGNCB has traded at a lower yield by .02%.
This is a .23% movement in the average yield due to a material shift in the relative prices of the securities. Prior to May 6th, the difference in average share price for AGNCP and AGNCB was $1.4536. Since then it was $.7280. Perhaps it is merely a coincidence, or perhaps share prices changed to reflect relative value as soon as the discrepancy was brought to light. On the next day of trading, shares of AGNCB outperformed by $.46. Remember that part of the reason these opportunities can exist is the smaller market size and lack of coverage.
*Below this line is the update*
On 06/10/2016, I got an update from a subscriber that the shares were being marked hard to borrow and that the annualized charge on borrowing the shares for a short was hitting 35%. I called Charles Schwab and asked them to locate shares for a short position. Their employee told me the A shares were hard to borrow and are carried at a 35% annual fee and required a minimum position of $50,000. The B shares were not available. The pair trade in this article is broken by the costs to borrow.
The assessment that investors in ARR-A should look to move to ARR-B remains in place as ARR-B offers better call protection and a better yield.
The following chart shows the difference in yields and share prices from 05/20/2016 (the day before my article) through prices on 06/08/2016. It also has the latest prices from 06/10/2016 during the trading day.
Disclosure: I am/we are long ARR.
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.
Additional disclosure: Information in this article represents the opinion of the analyst. All statements are represented as opinions, rather than facts, and should not be construed as advice to buy or sell a security. This article is prepared solely for publication on Seeking Alpha and any reproduction of it on other sites is unauthorized. Ratings of “outperform” and “underperform” reflect the analyst’s estimation of a divergence between the market value for a security and the price that would be appropriate given the potential for risks and returns relative to other securities. The analyst does not know your particular objectives for returns or constraints upon investing. All investors are encouraged to do their own research before making any investment decision. Information is regularly obtained from Yahoo Finance, Google Finance, and SEC Database. If Yahoo, Google, or the SEC database contained faulty or old information it could be incorporated into my analysis.