Mortgage REITs have become a hot topic as Wall Street firms have boosted coverage of this specialized slice of the stock market amid a flurry of IPOs over the past two years or so. These companies are potentially worthy investments, even for risk-averse clients.

However, it's important to understand how to analyze mortgage REITs, which really operate like levered bond funds. Knowing how a given fund's portfolio will be affected by changing interest rate environments and what hedging strategies the fund managers are using is essential.

Aside from REIT requirements - 75% of all assets must be real estate-related and 90% of taxable income must be returned to investors as dividends - mortgage REITs have little in common with more familiar property REITs. Agency mortgage REITs, which include names like Annaly (NLY) and Anworth (ANH) or more recently American Capital Agency (AGNC) and CYS Investments (CYS) - fulfill the real estate requirement by owning mortgage loans or bonds guaranteed by Fannie Mae, Freddie Mac or Ginnie Mae rather than real property, and finance those purchases at low short-term rates.



This makes owning shares in a mortgage REIT akin to owning a bank, minus the branches or ATMs. And the similarity to the banking business model, it turns out, is the main reason mortgage REITs are so popular now.

Even after funding costs and hedging, competent mortgage REIT managers can lock in spreads of 200 basis points or so. With modest leverage around 8-to-1, they can achieve return on equity in the mid-teens (and by definition have to distribute 90% of the taxable income back to investors as dividends).

The Fed has said it intends to keep the cost of financing mortgage securities near zero until well into 2013. This makes the mortgage REIT business model fundamentally more attractive than it was during its last wave of popularity in 2004 and 2005.

If 8-to-1 leverage sounds high, consider this: When a bank holds a mortgage that is not guaranteed by Fannie Mae or Freddie Mac, that bank is required to hold 8% capital in reserve (11.5-to-1 leverage). For bank holdings of Fannie or Freddie guaranteed mortgage-backed securities, that credit reserve requirement is cut by 80% to just 1.6% capital (more than 60-to-1 leverage). Banks do hold additional reserves for interest rate risk, but mortgage REIT leverage still looks tame by comparison.

Leverage magnifies both returns and risks, of course. So a careful review of how each agency mortgage REIT manager has built a portfolio can help investors see a difference in risk profile that the markets may not fully recognize.

Consider two seemingly very similar and hypothetical $100 million mortgage REITs, both of which have levered themselves 8-to-1 and both of which have bought exclusive agency-guaranteed mortgage-backed securities. Let's call them FixREIT (it buys all fixed-rate mortgage-backed securities) and ArmREIT (which buys only pools of bonds made from adjustable-rate mortgages). For comparison, assume each REIT holds $900 million of mortgage pass-throughs, and that each has fixed costs of $1 million per year.

Neither of these hypothetical REITs matches portfolios currently in the market, but this analysis can help planners sift through publicly available information to determine which mortgage REITs, if any, might be appropriate for clients. At 14% yield for FixREIT, compared with a 10% yield for ArmREIT, most investors gravitate toward the higher yield. But what happens when interest rates go up just two percentage points?

As with any bonds, mortgage-backed securities lose value when interest rates rise. With leverage, the effect of price loss on the shareholders will be multiplied by the leverage.



Before looking at the comparison chart (below), consider the essential "problem" with residential mortgages: the prepayment option. When an investor buys a new 30-year fixed-rate mortgage, that mortgage contains an option for the borrower to pay off the mortgage early, usually without penalty, any month before maturity. The potential for increased early prepayment due to government efforts to help borrowers refinance is a main reason mortgage REIT stocks have suffered price declines over the past few months, analysts say.

This is one risk, to be sure. But mortgage REIT investors should actually be worried about the prospect of a rise in interest rates and resulting slower prepayments since that is the bigger risk to their capital and income.

For example, the 4.5% FNMA pass-through used as FixREIT's theoretical portfolio is priced as though 30% of the borrowers will pay off their mortgages every year starting in 2012. If mortgage rates rise to 6.5% by then (still historically low), investors might expect only 8% to 10% of those borrowers to pay off their mortgages each year, and that 2.75-year bond that FixREIT owns might look more like a seven- to nine-year bond. ArmREIT will fare better, since its borrowers were scheduled to shift to adjustable mortgages three years from now, with little incentive to prepay (or not) at that point, no matter where rates go.

Interest rates, of course, are not likely to increase by 200 basis points overnight, so both REITs would have time to manage their risk as the rate increase happened. Also bear in mind that, in an actual agency mortgage REIT, the portfolio is built up over time and it constantly pays back principal that needs to be reinvested, so real-world agency mortgage REITs have blended portfolios of mortgage-backed securities bought over time.

Next, look at how much earning potential is left in each type of mortgage REIT after the rate adjustment. FixREIT will have to sell up to half of its portfolio to meet margin calls, while Arm REIT has enough capital cushion to avoid selling at a loss when prices drop. In addition, FixREIT only has hedges that last 2.75 years on $400 million to protect against higher financing costs while the bonds extend up to five years. ArmREIT only loses part of its income on $100 million of its $900 million portfolio. In short, we could expect FixREIT to lose much of its equity and earn much less on what's left. Dividends would likely drop by 60% or more per share without much chance of recovery. ArmREIT would lose some (20% or so) of its book value, but dividends would hardly move by comparison, dropping about 5% to 10%.



The key metrics to look at when analyzing mortgage REITs on behalf of clients are: the premium or discount-to-book value the stock is trading at, the net interest margin, the leverage ratio and the portfolio duration (net of hedging). For more risk-averse clients, mortgage REITs with lower portfolio durations will have lower capital (book value) volatility. Those with lower duration gaps (the difference in the durations of the financing and the securities financed) will have lower earnings volatility.

The higher the percentage of the portfolio that is hedged - this question is usually asked by analysts in earnings calls - the more stable the future income stream will be. Also, if a portfolio has a high percentage of ARMs, especially ARMs that are already resetting to floating rates, there will be much lower risk-to-book value.

Finally, it's important to understand political and regulatory risks before advising clients on these stocks. The political risk today is that a breakthrough on the logjam of "underwater" borrowers and/or delayed foreclosures might result in a huge wave of prepayments, which would hurt agency mortgage REITs if they paid a premium - more than 100% of the face value of the securities - and were counting on a longer stream of interest payments. Since everything in bond world is at a premium these days, it's a real issue, though a onetime event if it happens.

The regulatory risk seems less drastic, at least to the market participants. Still, the SEC is looking into the mortgage REIT sector for the first time since the REIT was introduced in 1960. The issue is likely to be resolved with more complete disclosure of hedging, funding and portfolio specifics, and with restrictions on rapid-fire trading. Regulatory scrutiny and the recent decline in book value premiums have been cited as reasons for delays in the launch of new entrants.

Those potential future IPOs will likely come from some of the best-known bond managers in the business. When they happen, clients may want more information. After all, even for REIT dividends taxed as ordinary income (and not at the lower dividend tax rate), investor yields in or near the teens demand attention.

Howard Hill is a financial writer and author of the Mind on Money blog. A former portfolio manager at Babson Capital, he was also vice president of mortgage finance at UBS Securities and founding managing director of the securitized products department at Deutsche Bank Securities.