Saturday, February 9, 2008 6 comments ++[ CLICK TO COMMENT ]++

Added to Watch List: Thornburg Mortgage Convertible Preferred

I'm adding Thornburg Mortgage 7.5% Cumulative Convertible Redeemable Preferred Shares Series E (TMA-PRE) to the watch list. Anything to do with residential real estate is very risky so this is not for the faint of heart.

Thornburg mortgage (TMA) is a mortgage loan company (a REIT without physical real estate holdings) concentrating on jumbo adjustable rate mortgages (ARM). In other words, it caters to the upper middle-class and upper class real estate buyers. It essentially makes money on the difference between in the cost of the mortgage and what it charges.

I am attracted to the preferreds because they are convertible and have a decent yield (9%). I'm still trying to figure out what is a good return for me but my current thinking is that anything that yields 10% for the long-term is worth looking at. Stocks have a long-term return of around 10% so a bond that can beat that is worth considering. However bond/preferred coupons don't increase so inflation is a huge risk. To overcome this, I primarily look at convertibles or bonds for the medium term (i.e. 5 years). These preferreds (and bonds in general) are also less tax efficient than common stocks.

Some notes on the series E preferreds (I still have to read the prospectus to confirm this info):

  1. call info: June 19 2012 @ $25 (current preferred price: $19.50)
  2. conversion info: 0.77232 common shares (current break-even conversion price $25.25; current stock price: $13.12)
  3. cumulative
  4. Moodys rating: Caa1; S&P rating: CCC
  5. current yield: 9.64%

There is also the 10% series F which is very similar but I like series E because its price is below the par value and hence is unlikely to be called. If you are ok with the preferred being called and/or want to convert at a lower price, the series F is a better bet.

One risk with ARM is that if rates go up (haven't lately but likely in a few years) defaults may rise. The upper class hasn't been impacted by the slowdown yet (mostly has been hurting subprime mortgage borrowers so far) but it remains to be seen if that will last. Already you are seeing some luxury clothing retailers noticing some weakness so I wonder if homes will also be an issue.

Since this is a REIT, it has historically paid out most of its income to shareholders on a consistent basis. During the housing boom in this decade, it has paid around $2.50 per share in dividends (to common shareholders). I think the likelihood of it paying preferred shareholders in the future is pretty high.

The real big risk is bankruptcy risk. If I were to take a position, I have to do more homework on that front.

Assuming one deduces that bankruptcy is remote, I think this is worth picking up if yield reaches around 12% (but I'm not sure if it's going to hit that number).

Couple of Others

A couple of other somewhat distressed, lower risk, convertibles are the following:

  1. Citigroup Inc., 6.50% Dep Shares Series T Non-cumul Convertible Preferred Stock (C-PRI)
  2. Bank of America, 7.25% Series L Non-Cumulative Convertible Preferred Stock (BAC-PRL)

Both of these were issued in January of this year (these were some of the big capital infusions that made the news a while back). If you think both of these companies will recover from their current subprime-induced problems then it's worth checking out these preferreds. Given their low risk, the yields are not that great (around 6%) so I would wait for them to drop a bit more. One should also evaluate whether it's worth buying the convertibles instead of the the common stock. With Thornburg, I like the convertible preferreds because there is a chance that housing may not recover for many years so I don't mind getting paid to wait (I'm assuming the company doesn't default or go bankrupt). These banks, on the other hand, can easily recover within 2 years (and yield is lower) so the common stock may be more attractive (have to do an analysis).

Tags: ,

6 Response to Added to Watch List: Thornburg Mortgage Convertible Preferred

February 10, 2008 at 1:47 PM

Where do you find/screen bond data?

February 10, 2008 at 4:53 PM

TMA had a much improved quarter, and I believe they declared a dividend on the common? The founder, Garret Thornburg bought 1 million shares back in Oct at 9.50, right in the middle of the storm. Larry Goldstone, the CEO & Pres, is great. He was courageous enough to step into the media spotlight and go public with how dangerously bad the credit markets had become. He is 101% committed to his company and shareholders and has been forthright in words and action. He even converted Jim Cramer, once a total bear on the stock, who is now very bullish on TMA. The stock has had a great move. They did a lot of jumbo business in the high flying states, so there is still risky exposure.

OT, but interesting article on Egan Jones


February 10, 2008 at 5:47 PM

I'm not a fan of the Yahoo message boards what with all the crude dimwits that populate them but there are some new voices on the MBI & ABK boards that are really putting out some excellent stuff. One of them has just started a Yahoo groub dedicated to (hopefully) informed discussion of MBI & ABK. All idiot posts will be deleted.


February 10, 2008 at 7:38 PM


I only look at exchanged-traded bonds and preferreds. These are bonds/preferreds that trade on NYSE/NASDAQ/etc. In contrast, the vast majority of bonds do not trade on any listed exchange (they trade on over-the-counter market).

The main reason I look at exchanged-traded bonds is that I'm in Canada and it is not easy to buy a non-exchange-traded bond/preferred. The other reason is that the bond market is not efficient for small investors. You generally need to invest tens of thousands; whereas an exchange-traded bond/preferred is like a stock that can be bought in small denominations.

If you want to look at exchange-traded bonds/preferreds/mandatory convertibles/etc, the best site is Enter a ticker symbol if you are looking for something or look through the lists. You need to register (free but they suggest a small contribution if you can) to look at the lists. Since there aren't too many exchange-traded bonds/preferreds/etc, simply looking through the whole list is simple. This site is amazing and even has a link to the prospectus, which is nice...

If you are not limited to exchange-traded bonds, then you can get price quotes from the FINRA site. Just type in a ticker for the stock, or search by criteria that you want (yield, maturity, etc).

Don't forget that there are also many ETFs and closed-end bond funds that you may want to consider if you are not making a company-specific bet. For example, if you want a 20 year US treasury bond, you can consider an ETF like TLT (but do note that a bond FUND is different from buying and holding a single bond to maturity. A fund keeps rolling over the bonds (i.e. duration stays constant) whereas if you buy a bond, the maturity keeps declining as time passes).

I'm just a newbie and don't generally invest in bonds. I just use them strategically (eg. turnaround situations, parking money, etc). If you are a young investor or your portfolio is small, stocks are still way better. As Buffett and others have repeatedly pointed out, inflation is a huge risk to any fixed-income instrument. Since coupons don't increase (whereas dividends on stocks increase over time), long term bonds/preferreds are risky.

February 10, 2008 at 7:49 PM

Thanks for the info Cak...

I was just looking at the Thornburg convertible and, man, it was beaten up very badly back in August. I think it dropped more than the Ambac bonds did (although it's not quite the same since one is a preferred and Ambac's is a bond). I'll keep an eye on Thornburg and see what happens over the next few months.

I saw that Tom Brown rebuttal of Egan Jones. Bears don't think Tom Brown has any credibility (his financial holdings have been hit as well) but one can't really argue against his critique of the Egan-Jones estimate. It doesn't surprise me since I always felt that all these wild estimates are just nonsense. I mean, instead of picking $200 billion, why not just assume all the assets are written down to zero and pick a really large number like $1 trillion.

It's awefully quiet in Ambac-land. I just don't see how they can't pull off a deal to raise a billion or more (assuming it's enough for the rating agencies). The MBIA capital injection went well. The market actually liked the deal and the stock was actually up and was trading above the deal price all day. A lot of people who don't understand hte situation were clearly puzzled how the stock can be up when it is being diluted...

February 11, 2008 at 9:48 PM

Hey Siv-

I highlighted that new discussion board dedicated to MBI/ABK. They are looking for some bright people to do a CDO loss assumption study on some MBI paper and I thought of you. If nothing else, it might be a good learning experience. These guys are pretty sharp.

From poster Spinvestor

"I've been working on the loss severity of the MBIA CDOs.

This is based upon the raw inner deal data from the Hackman Open
Source Model.

The question that this effort is attempting to answer is: 'Is the
950MM raw loss on the Jupiter V CDO representative of reality'.
950MM raw loss is 63% of the deal before reinsurance and outer CDO
attachment kick in.

Now that others are looking publically at the Open Source Data, I've
though that collaborating with what I've done so far is probably more
beneficial if a collaborative effort of 10-15 people can be formed to
medothodically go through the RMBS tranch data.

There are 125 Members of this group so far that have bothered to
join, and like anything it probably will boil down to an 80/20 effort
so perhaps there are 15 people interested in contributing.

There is obviously a different level of skill sets so for consistency
I would provide a framework for data collection, detail how to track
down the RMBS prospectuses, show where to get the servicer repots
from etc,etc.

There are ~190 RMBS tranches in Jupiter V which is where we would
start as its a High loss CDO.
I would propose to hand out 10 Tranches at a time to willing
participants and then take your gathered data and collate it into a
published updates spreadsheet that all can use to make their own

I've created a File directory structure and populated it with the
Jupiter V CDO prospectus and a sample of 4 inner RMDS tranche data
from two other CDOs to show the data I'm currently collecting.

This post is to gauge interest in this approach as it's not feasible
for me to continue solo.

10 trances should take 4 hours of effort.


What I did was:

a) Take the highest loss CDO's from Hackams loss model which was Jupiter V, Broderick III and HighRidge I + Sgaterious I. These 4 account for ~50% of the loss CDO loss in his model.

b) With in these CDOs find the fulcrum tranches. The Fulcrum Tranches are those that are borderline wipe or pass.

I started by sorting the transactions via the 'Min rating' column and using starting with an A rating as the fulcrum tranche.

A lot of the RMBS A rated tranches are M-4/M-5 tranches. I then built a simple loss model based on prepayment speeds, default rates and commloss severity to get an Idea if the m-4/m-5 tranches would make it as you roll forward the securitized book forward.

c) Go through all the servicer reports.

Another point on the Hackam model - Those numbers he's put out there seem to be Raw CDO loss estimates - so it didn't seem to me that he included attachment points (compare inner CDO losses with the total CDO loss), reinsurance and was pre-tax.

Post a Comment