Description
Capstead Mortgage (CMO) is a mortgage REIT trading at 1.6 times book *. As
Capsteads operations are equivalent to borrowing 30-day money to invest in 12-month CD’s investors in this stock are effectively swapping $160 for $100. I will describe two ways to take advantage of the large premium. The catalysts for both trades are the flattening of the yield curve to more normal levels and Capstead’s warning of “deteriorating fundamentals over the immediate quarters” in a September press release anouncing a 33% cut in the dividend.
(* The Price to Book (ptb) has been adjusted for "new" stock issues, dividends and earnings subsequent to 30 June 2004 as well as the MTM of assets. Book Value = 133 mil, Market Cap = $ 214 mil, Price =12.40, ordinary Shares = 17.24 mil, PTB = 7.70)
Capstead is a mortgage REIT that uses leverage (Assets/Equity = 10x) to exploit the spread between 30-day interest rates and 12-month interest rates. Specifically, Capstead buys standard ARM securities (1-year tease period followed by rate adjustments every 6 to 12 months) using repurchase agreements linked to 30-day LIBOR. In order to avoid credit risk Capstead only holds agency ARMS (backed by Fannie Mae, Freddie Mac, Ginnie Mae). Hence, the spread that that is earned is the reward for assuming interest rate and prepayment risk.
The spread has been unusually wide following the aggressive rate cuts by the Fed beginning in January 2001. With each rate cut the yield curve became steeper benefiting Capstead as it’s balance sheet is liability sensitive. This means that when interest rates decline the cost of funds fall faster than the interest earned on the assets. In Capstead’s case the Fed rate cuts reduced the cost of funds immediately while the interest they earned on their ARMS could theoretically take up to 12 months to reset lower.
The benefit of the steep yield curve is clearly visible in the historic net
interest spread earned by Capstead. In Q3 2000 the spread earned reached a low of 23bp. The spread recovered rapidly from here following a series of Fed rate cuts and in 2001, 2002 and 2003 Capstead earned an average spread of 230bp,340bp and 230bp respectively. The spread has subsequently narrowed to 190bp in Q2 2004 and the recent rate hikes and flattening of the yield curve will see the spread fall to 150bp in Q3 and even lower in Q4 2004.
The declining spread has forced Capstead to cut their dividend from 50c in Q2 2004 to 33c in Q3 2004. The third quarter ends on the 30 September 2004 and the earnings announcement in October is expected to be the first catalyst for the trade. The deteriorating financials and further dividend cuts over the immediate quarters should focus attention on the 1.60 ptb rating especially since alternative MREITS who also avoid credit risk are trading close to book. (MFA, LUM, ANH)
Before describing the trades some more background is required. Until the early part of 2000 Capstead borrowed short-term funds to invest in longer dated securities. Unfortunately the Fed rate increases in the second half of 1999 and the dramatic flattening of the yield curve resulted in tremendous losses. Capstead was in trouble and ultimately rescued by the Fortress Investment Group through a series of capital injections. Fortress also assumed the management of Capstead and were instrumental in switching to the current investment strategy which avoids interest risk by closely matching the duration of it's assets and liabilities.
In July 2003 Fortress sold most of it’s investment in Capstead and resigned
their management positions. The Fortress management were replaced by promoting staff within the company. The change in management coincided with a notable change in strategy. While the Fortress management were in charge Capstead’s total assets declined from $8.6 billion in December 2000 to $2,6 billion in September 2003. The decline in assets was due to aggressive stock buy back programs as well as high repurchase rates as homeowners rushed to refinance their homes. While Capstead’s assets were in constant decline the other MREIT’s were using the favourable interest rate environment to grow their assets at break-neck speeds. An example of this is Annaly Mortgage Management (NLY) which grew it’s total assets from $2bil in December 2000 to $13bil by December 2003.
Annaly’s behavior highlights a misunderstood aspect of the typical MREIT
business model. In order not to pay tax MREIT’s must pay out most of their
earnings as dividends. This means that MREIT’s can not use retained income to grow assets. Instead, asset are grown by using more leverage and/or by
issuing new equity.
Given the favourable operating environment MREIT’s have enjoyed VIP status in the equity and debt markets. As a result of this easy access to the capital markets they have had a licence to grow assets by $1000 by raising $100 in equity and borrowing $900 using mostly repurchase agreements. This cycle has been reinforced by ptb ratings above 1 which have allowed MREIT’s, like NLY, to increase their book value per share with every stock issue.
While the above cycle was in full swing Capstead was doing the reverse by
buying back stock. (This was in part due to Fortress’ exit strategy.) But, following Fortress’ exit the “promoted” managers have wasted no time jumping onto the bandwagon and have grown assets from $2,5 billion in December 2003 to $2.9bil in June 2004. More growth is expected in Q3 judging by the continued "new" share issues during the quarter.
The adoption of the more aggressive strategy is concerning because of the
declining operating environment which has seen other MREIT’s like MFA reducing their leverage and becoming more cautious. On the other hand who can blame management, as the ability to issue shares at 1.6 times book is a treat.
The final background item relates to Capstead's high fixed costs resulting from the extensive use of preference shares as equity. Cumulative preference shares currently make up $180 mil of the $313mil in total equity. These preference shares pay a dividend of $5mil a quarter or 11% per annum. (Operating earnings were $12mil in Q2 2004) This is very expensive finance compared to 30-day LIBOR and as it is a fixed cost it adds significant volatility to the ordinary dividend. In essence it means that even though equity is geared 10 times ordinary shareholder earnings behave as if the gearing is much higher. The higher operating leverage makes Capstead far more risky compared to other MREIT’s that exploit the interest rate spread. These MREITS typically use preference shares sparingly and gear equity 10 times. Given the extra risk it seems strange that Capstead is trading at a premium to these stocks.
The higher operating gearing makes Capestead the most vulnerable to a
flattening yield curve. It is well known that the current yield curve is historically steep. Today the spread between a 1-year Treasury and a 10-year Treasury is 186bp well above the 100bp average over the past 30 years. If the spread had to revert to it’s mean and assuming no asset growth then ordinary earnings will fall from $11 mil to $4 mil. And if the spread had to fall to 50bp then net income to the ordinary shareholder would be zero.
This is a good time to highlight the risks to the trade. The narrowing of the interest rate spread will not necessarily translate into lower earnings. This is because the decline in earnings from the smaller spread can be countered by growing assets. While the ability to grow assets depends on the profitability of the spread it also depends on how accommodative the debt and equity markets are. If access remains easy then the effect of the narrower spread will be masked by growing assets and/or by increasing leverage. The declining earnings can also be masked by the selective selling of securities in order to book gains on sales.
Recommended Trades
--------------------
Two possible trades to exploit the 1.60 price to book are:.
Trade 1: Naked Short
-----------------------
Short CMO at 1.6 times book and cover at 1 times book. Capstead's stock will decline from 12.40 to $7.70 for a 38% return.
The trade will be even more rewarding if conditions turn against the MREIT’s as they did following the Fed rate increases in 1999 and 2000. A sample of average ptb ratios during this period is shown below.
1999 2000
CMO 0.65 0.60
TMA 0.80 0.80
NLY 1.04 1.10
A review of the financial disclosures during this period are a harsh reminder that profiting from the spread trade is not always plain sailing. In a nutshell, the rise in interest rates during 1999/2000 caused the prices of the longer dated securities to fall. These price declines triggered margin calls as lenders moved to protect themselves from the declining value of their collateral. The inability to meet these margin calls led to forced sales and the recognition of large “losses on sale” which ultimately hit the equity base of the MREIT's.
Trade 2: Short Capstaed: Long 1-year Treasury
-----------------------------------------------
One of the problems with the naked short is Capstead’s high dividend yield
which makes the short expensive. A possible way to cover this dividend is as follows:
As Capstead is geared 10 times it has the following capital structure:
Ordinary Shares $ 4
Pref Shares $ 6
Liabilities $ 90
Assets $100
This means that by owning $4 of Capstead’s book value an investor is
effectively long $40 of traditional ARMS (duration = 12 months) and short $36 of 30-day LIBOR (Cash). If the ordinary shares of Capstead are shorted then the situation reverses to long $36 cash and short $40 of traditional ARMS. The short position in ARMS can be eliminated by purchasing $40 of 1-year Treasuries.
The net position is as follows.
- Proceeds from shorting Capstead at 1.6 times book equal $6.40 (1.6 * $4)
- Less the $40 used to purchase the 1-year Treasury
- Results in a net outflow of $33.60
- As the transactions has created an asset which is long $36 cash the trade has enabled an investor to swap $33.60 for $36 cash yielding a 7.1% return on invested capital*. Also, the interest earned on the Treasury should be enough to cover the dividend. A riskier version of this trade would be to purchase $40 of the Vanguard GNMA fund (VFIIX) which currently yields 4.6%. The fund is riskier as the average duration of the fund securities is 3 years.
(* Note: Most of the invested capital is employed in a risk-free asset. Hence, the return on risk capital is the same as the “naked” short. In order to reduce the amount of capital needed this trade can be replicated using options or futures on short term interest rates.)
Conclusion
As Capstead is nothing more than a vehicle that is has borrowed “30-day” money to invest in “12-month” money it is difficult to understand why it is trading at a 60% premium to it’s adjusted book value. The large premium is even more puzzling if one considers that:
- Capstead has warned of “deteriorating earnings over the immediate quarters.”
- The high fixed cost of the preference dividend makes Capstead’s ordinary
dividend the most vulnerable of the MREIT’s to a shrinking interest rate spread. Yet these MREIT’s trade close to book value.
Catalyst
- Deteriorating financials over the immediate quarters.
- Further cuts in the dividends
- Flattening of the historically steep yield curve