CYS Investments (CYS) is a mortgage REIT that was first issued on the NYSE in 2009, with about $2bb in market cap. Its dividend yield is around 11%.
CYS is lead by Kevin Grant, who used to manage directly or indirectly several fixed-income funds between 1993 and 2005 at Fidelity.
Mortgage REITs are inherently different from more standard real-estate focused REITs because their income is mainly composed of mortgage or mortgage-backed securities coupons they collect, rather than the rent from properties in the case of usual REITs. As such, they typically are not in a commercial business with clients to manage (as is the case with real-estate focused REITs). As a matter of fact a mortgage REIT's business is a lot like a bank's, making money from the differential in asset yields and funding rates.
CYS is exclusively invested in agency mortgage-backed securities, which are guaranteed (interest and principal) by the US Government either directly in the case of GNMA, or indirectly in the case of Fannie Mae or Freddie Mac.
As of the last quarterly report (Q4 2012) CYS had $16bb in agency MBS yielding 1.97%, out of which $13.8bb were financed by repo for a cost of 1.03%. The interest margin for CYS is therefore 0.94%, with a leverage of 7.7, which allows us to roughly expect a return on capital in the order of 7%.
This margin can be quite variable however, since it was 1.24% the prior quarter, which would have led to an estimated return on capital of 9.5%.
In addition to the interest margin, gains on sales on securities can contribute to the return on capital.
Even though all these agency MBS are guaranteed in terms of credit risk exposure, they have a degree of interest rate exposure like all fixed-income instruments (called interest rate duration), which is further magnified by the amount of leverage. This exposure is typically hedged by writing swaps (paying fixed, receiving floating) in the right amount.
Mortgage-backed securities also exhibit a particular risk, negative convexity, related to prepayment directionality. When rates rise, prepayments tend to slow down; when rates drop, prepayments tend to accelerate. This contributes to making the value of these MBS drop more / increase less than would a comparable swap (same duration) when interest rates move.
The mechanical result of this effect is that a REIT's mortgage book, even if it is perfectly hedged with swaps, will tend to generate losses whether rates rise or drop. The REIT's large return on capital is essentially a kind of compensation for selling insurance against interest rates movements.
Some REITs prefer to look for « specified pools » rather than generic MBS which for one reason or another have better prepayment characteristics. However these pools are typically more expensive. In other words certain REITs chose to sell less insurance in order to have less risk, but they also collect less premiums.
The Fed's MBs purchase program (QE3) has compounded these risks by artificially lowering mortgage rates, which led to faster prepayments, although by the same token MBS prices were increased, which led to higher mark-to-market gains on mortgage REITs books.
One of the main reasons for mortgage REITs' attractiveness is the fact they do not have to pay any corporate tax thanks to their REIT status. In 2011 there were discussions about making the REIT status only available to those investing in real estate as opposed to mortgages, which would have negatively impacted all mortgage REITs.
Liquidity and credit risk for mortgage REITs
In 2008-2009 several mortgage REITs (e.g. Thornburg Mortgage) went bankrupt, which might put into question the resilience of their business model, knowing how levered they are.
It pays however to distinguish between agency and hybrid mortgage REITs, the former buying MBs with a guarantee, versus the latter buying non-agency or commercial MBS that are fully exposed to credit risk (as was the case with Thornburg).
In 2008 and 2009 the agency MBs market was not nearly as affected as the non-agency market, in terms of asset prices or financing conditions. The chart below shows Annaly's (NLY) stock price versus the S&P 500, and it appears that NLY exhibited less volatility.
Equity and mortgage REIT valuation
The net equity value of an agency mortgage REIT properly reflects the company's liquidation value. The assets (including those in derivative form such as swaps, swaptions or TBAs they might hold) as well as the financing are very liquid. This allows for an accurate and exercisable valuation, as opposed to the situation for pretty much any other kind of business.
As a result agency mortgage REITs usually trade around their book value, which is provided on a quarterly basis with earnings reports. In « normal » situations the book value does not significantly move from one quarter to the next, since quarterly profits are essentially paid in dividends and cannot be capitalized.
Since mortgage REITs distribute a large majority of their profits they resort to capital raises in order to fund their growth. Growth in and by itself is actually not synonym with improved return on capital. Quite the contrary actually, beyond a certain size the bigger the amount to invest the more difficult it will be to find attractive bonds, and the more constrained the REIT will be to buy generic paper.
Capital raises when the stock price is above book value are accretive to shareholders, as are share buyback programs when the stock trades below book value. The opposite is true as well, and capital raises when the stock trades below book value generally are a pure destruction of shareholder value.
In some cases, internally managed REITs need to reach a certain optimal size in terms of assets / number of employees in order to reduce operational costs, which might justify seemingly dilutive capital raises.
Internally vs externally managed
This is a very important point to me.
Generally speaking agency mortgage REITs have no particular competitive advantage among each other. Assuming they reach a minimum competence level, they all invest in the same market, all use fairly generic internal valuation tools, and all have access to the same potential leverage and repo conditions.
Their differences at the end of the day amount to the following:
- Capital raises or stock buy backs carried out with the shareholders' interest in mind
- Hedging strategy (how much to hedge with swaps or swaptions)
- Amount of leverage
- Positioning the book on the risk/return spectrum (TBAs on one end, specified pools on the other)
- Management cost
- Financing structure (issuance strategy for preferreds vs common stock)
Most agency mortgage REITs are managed externally, and the management is paid a fee that is proportional to the amount of equity under management. This creates a clear inclination to capital raises, not always with the best timing. In addition this does create operational leverage in the sense that management does not have any particular incentive to perform a great job - just to keep the book afloat.
On the contrary, an internal management team, themselves shareholders of the REIT, « guarantees » to some extent that capital raises or shares buybacks would be done with shareholders interest in mind. In addition, even if part of personnel compensation is variable, there is an optimal size (operational costs / assets) between the number of employees and total assets: operational costs are not strictly proportional.
Until 2011 CYS's management was externalized (management were at the same time controlling CYS and Cypress Sharpridge Advisors, which acted as CYS's manager). But they eventually decided for internal management in order to reduce costs, along with capital raises in order to reach a critical size.
This has worked quite well since operational costs dropped to 0.8% of equity, which is the lowest rate among agency mREITs to the extent of my knowledge.
The contrast with Annaly is quite interesting, since NLY's management had been internal since 1997, but they recently moved to externalize it (apparently in an attempt not to disclose management's compensation anymore).
Most investors seem to buy mREITs only for their dividends (which since 2009 have been in the 10-15% range), and do not seem to care much for book value. Since first coming to market CYS has served a good dividend rate, with a cumulative payment of $8.19/share, vs. an initial pricing at $11.89.
Dividends are necessarily variable, since return on equity is variable, being affected by market conditions (MBS returns, hedge and repo costs).
Nevertheless, reinvesting dividends (assuming no withholding taxes) the total return for someone buying CYS at the IPO would have been 83%, corresponding to 17% annualized:
In Q1 2013, dividend is lower than Q4 2012 because the net interest margin is lower, and also because CYS paid a sizeable special dividend in Q4 2012, anticipating higher dividend tax rates in 2013.
My bull case for CYS
CYS is currently trading 11% under EOY 2012 book value, while comparable agency mREITs trade at a higher multiple in general (apart from Armour Residential ARR whose externalized management destroyed shareholder value with a capital raise under book value in 2012).
Looking at the CYS's MBS holdings, I do not find anything particular that could justify this discount; their prepayment rate appears in line with peers.
Management style does not either justify a discount, quite the contrary. Internal management goes a long way towards ensuring interests are aligned, and operational costs are the lowest across the mortgage REIT universe. The move from externalized to internalized management is actually atypical and is a strong positive sign. CYS's management team hold significant amounts of stock:
One possible explanation for the discount would be the drop in dividend, with investors not factoring in the special dividend in Q4 2012.
A more subjective positive aspect in the management team is the fact that the CEO/Chairman Kevin Grant is not just another golfer: he has been an mountainer for over 20 years, having climbed the Kilimanjaro, and a few attempts on Everest.
A purchase at current prices, aiming for an increase to 1.05 - 1.10 price/BV ratio, which would be an upside of around 20-25%.
Assuming, conservatively, that BV is down by 5% for Q1 2013, this would still leave an upside of 15-20%.
In the meantime, one can still collect the dividend, at a yield around 11%, in line with the return on equity expectation given by CYS's management.
I see this as a fairly nice compensation just for waiting to make money!
As a bonus, the 2s-10s spread seems to be on a widening trend, with the curve steepening, after a low point in July 2012.
The evolution of this rate spread can be used as a proxy for the evolution of mortgage REITs' net interest margins. If the trend is confirmed, CYS's ROE should further increase, and dividends should follow suit.