Risks in REIT Preferreds
Board: Real Estate Investment Trusts
Here's how I'd like *NOT* to invest…
Believing I can predict what *will* happen with REIT preferred spreads is like burying my head in the snow – best left to dogs. Getting better at predicting what *could* happen with them, however -- if A happens then B is likely -- appears more achievable. What I'm after is ranges and boundaries. This post – one of a series of posts -- is just another attempt to get at that by taking an X-ray of REIT preferred spreads for a specific time period.
Warning: This is NOT investment advice and the data is NOT reliable – for discussion only!
Note: The "References" section at the end has data sources and links.
When you buy a 'risky' security such as a REIT preferred stock, its returns are usually higher than on securities regarded as 'safe', such as a 30-Year Treasury bond. The extra return you get over and above the return on the safe investment is the risk premium, or spread. The extra return is supposed to compensate you for the extra risk you are taking. To illustrate this, here is a slide with two charts. The chart on the left shows the risk premium (spread) for 94 equity REIT preferred stocks of all credit ratings, as well as the spreads for Baa bonds and Aaa bonds. The spreads are stacked on top of one another. All the spreads are to the 30-Year Treasury rate from 1/2/2013 to 2/7/2014. The chart on the right shows the same thing, except it is restricted to 27 REIT preferred stocks rated Baa by Moody's. There is a 520 basis point difference (about 1/2 of 1%) between the average spread of the preferreds of all credit ratings and the preferreds with a Baa rating.
The risk premium is often thought of as credit risk or default risk. However, a number of studies have shown that historical default rates fail to fully explain the size of spreads. These studies agree that there is a component of the risk premium that is not related to credit/default risk, but there is no general agreement on what the other risks are that make up that component, or their magnitude. The usual suspects are liquidity risk – those of you who have tried to buy or sell a lightly-traded REIT preferred may have experienced that – and call risk.
Keeping this in mind, consider there is little difference in the default risk between Treasuries and Aaa rated Corporate bonds. However, Treasuries are highly liquid, whereas Corporate Aaa bonds can be highly illiquid. In addition, US Treasury bonds can't be called but some Corporate bonds can be. So the risk premium for Corporate Aaa bonds is likely much more weighted towards liquidity and call risk than default risk. Thus, the Corporate Aaa risk premium may serve as a useful proxy for these non-credit risks. To take advantage of this proxy, let's decompose the Baa-rated REIT preferred spread to the 30-Year Treasury into two pieces: the spread between Baa-rated REIT preferreds and Corporate Aaa bonds (what we hypothesize as the credit risk component) and the spread between Corporate Aaa bonds and the 30-Year Treasury (what we believe is a proxy for the liquidity and call risk component). Here is the equation:
Risk premium (Baa-rated REIT preferreds minus 30-Year Treasury) =
credit risk (Baa-rated REIT preferreds minus Corporate Aaa) +
liquidity/call risk (Corporate Aaa minus 30-Year Treasury)
Here it is on a chart:
Now let's do the same thing for Corporate Baa bonds as we did for the 27 REIT preferred stocks above -- we'll decompose the Baa spread into its liquidity/call risk component and its credit risk component. We can then compare it to what we did for the 27 REIT preferreds. It turns out that for the 27 REIT preferred stocks rated Baa, credit risk accounted for about 75% of the risk premium whereas for Baa bonds credit risk accounted for about 50% of the risk premium. In other words, investors were much more concerned about the credit risk portion of REIT preferred stocks rated Baa than they were about the credit risk portion of Baa bonds, relatively speaking.
Let's add another factor to the mix. You may be aware that the rating agencies like Moody's "notch" their credit ratings for preferred stocks. For example, Moody's rates Kimco's unsecured debt at Baa1, but they notch Kimco's preferred stock down one level to Baa2. This isn’t specific to Kimco or REIT preferreds – they do it pretty much for all preferred stocks. Thus our 27 REIT preferreds rated Baa get a double-whammy – they are already notched down one grade, and then investors further discount them.
So, based on all of the above, would you conclude that the risk premium for Baa-rated REIT preferreds is too high relative to Baa bonds? You certainly couldn't be faulted for reaching that conclusion given the evidence. For me, I'll continue working on ranges and boundaries and "if A then B" and how spreads change over time. My next step is to examine REIT preferreds where the unsecured debt is rated investment grade but the preferreds are either not rated, or rated speculative -- this appears to be a potentially profitable area for further study.
What Credit Markets Tell Us -- John V. Duca
Federal Reserve Bank of Dallas -- 1999
Federal Reserve Bank of St. Louis
After you click on one of these links, if you want the definition of the data series, you can click on the "View Data" link on the vertical menu at the left of each screen.
30-Year Treasury Constant Maturity Rate
Moody's Seasoned Aaa Corporate Bond Yield
Moody's Seasoned Baa Corporate Bond Yield
BofA Merrill Lynch US High Yield Master II Effective Yield