Tuesday, March 20, 2007

Mortgage Insurers

I am completely convinced that we have a housing bubble in the US that will bust and cause havov for the US economy. Still, I would like to keep my cool and invest somewhere at some time in the real estate sector when fear of the crash is overdone. So where to invest? The usual places would be homebuilders, banks, furniture makers, home furnhing like Home Depot etc or title insurance.

Another place is private mortgage insurance (PMI). Mortgage insurance is an interesting business. Basically they insure against banks losses for the top 10-20% or so of loans for people who don't pay a full down payment. Some notable mortgage insurers are (tickers) MTG, RDN, PMI and TGIC. All of these are down somewhat and hovering near book value. These companies have typically grown about 15% over the past decade or so and investors have had good returns despite significant volatility. Buying them at the right time typically gave you a 40% one year return.

The bussiness model is fairly simple. They get paid a premium by the borrower. If they forclose and the bank takes a loss, they pass on this loss to the PMI which either pays the bank or takes over the title. Like any insurance company, it makes sense to talk about the loss ratio (losses over revenue) and expense ratio (operating expenses over revenue) and the sum of the two which is called the combined ratio. When the combined ratio is less than 100%, the PMI makes an underwriting profit. It can also invest the float for investment income. The net operating income is the sum of the underwriting profit (or loss) and the investment income. Most regular insurance companies have a combined ratio near 100%, usually a few points higher. Thus they take a small loss to be able to invest the float at a higher return (mostly in bonds). They make a profit on the spread. A combined ratio of 100% means they are getting an interest free loan which they can invest at say 5% in bonds. There is generally a limit of float-to-equity which is considered safe but they can thus leverage their return on assets by this ratio to make a reasonably high return-on-equity of 12-20% or so for a good insurance company. Generally they grow at about the same rate as ROE.

Mortgage insurance is quite different. In good years the combined ratio is very small. For example MTG (company name is MGIC) has a combined ratio of between 45% and 70% for the past 5 years. Wow! What a bonanza! Even if it just buried their money in the backyard they are making a net profit margin between 30-50% just for signing pieces of paper. Unlike most insurance companies, they make more money from underwriting profits (in good years) than they do from investments.

Figuring out the future of this industry is complicated and requires a good understanding of the whole mortgage industry. (I don't claim to understand it well enough but I am working on it).

For example these companies are at the mercy of :
1) the Federal government which competes with them through the FHA.
2) Fannie Mae and Freddie Mac who make a lot of the rules of what is conforming etc.
3) Strengthening/consolidating lenders who siphoning off premiums through captive reinsurance.
4) Lenders offering piggyback loans and other PMI alternatives.

I won't go into the details of the these issues. Another important things is the shape of the yield curve, the persistenacy of premiums, tax deductability, pricing pressures etc.

What I want to focus on is loss ratios. I have mentioned above that combined ratios over the past 5 years are in the range 45-70%. Expense ratios are typically in the range 20-30% which leaves loss ratios in the range 20-40%. Are these typical numbers? Are there typical numbers? Would we expect losses to increase significantly?

That is difficult for me to estimate. Surely they will rise, but will they rise enough to make combined ratios much above 100% where the PMIs would start reporting losses and decreasing book value? I don't think anybody knows for sure which makes it quite a gamble.

Lets look at the past for examples. First of all there is the Great Depression. There were actually PMI around during the 1920s. None of them survived. All went bankrupt in the collapse of the financial system. MGIC (MTG) started up in 1957 to compete against the FHA which was a "New Deal" program to restart the mortgage market during the Depresssion. You can read about the history of MGIC HERE .

Ok, well you might not want to count the Great Depression. Hopefully that won't happen any time soon. What about more recent history.

Here is a plot that I made of Book Value Per share (BV), Earnings (E), Dividends (D) and Return on Equity (ROE) since 1990.

Basically BV has grown at 20% and you can now get a 2% dividend yield. On top of that they are selling for P/B of about 1 which very low. If they continue growing at this rate and get back to P/B of say 1.5 in 3 years, your getting a return of 36% anualized for the next three years, maybe even better. Sounds great right?

I have my doubts. Let look further back into the 80s. The 2002
for PMI group shows the loss ratios back to 1981.

The bottom line is that loss ratios in the 1980s were terrible. They attributed these to the real estate decline in the oil patch. When inflation was reigned in by Paul Volker's Fed, oil prices collapsed and so did all the business in Texas and other oil patch locations. Loss ratios were in the range 200-265% for contracts written in years 1981 and 1982 and these losses carried on for twenty years for contracts written in those two years. That is PMI was still paying out losses (probably small ones) in 2002 for bad decisions made in 1981. I am not sure how all of this works its ways into yearly earnings. I would like to see a chart of common equity for PMI through all of these years. I am guessing it decreased significantly in the mid 80s but don't know sure since PMI was not a public company then.

How does this relate to today? Surely the oil-patch real estate bust was more acute than todays national/international housing bubble. However todays bubble is more wide spread and I don't see any reason why losses couldn't be worse than the 1980-1983 debaucle. Will these companies even survive? I don't know. Hopefully they are reinsured well enough to handle a national housing crash but I don't have the expertise to figure that out. I have seen a quote in the 10-K of MTG that says that because they are regionally diversified, they don't expect major losses. However that seems to rely on the idea that house prices can't crash everywhere at once. What if they do? Are they still prepared to weather the storm if house prices decline by 20-30% nationally like some bears (i.e. Gary Shilling, Jim Rogers) are predicting?

I don't want to make that bet. I am staying away from investing in these companies until I start seeing them handling severe losses gracefully. Which means I might miss them entirely. That's fine. I will stick to more transparent businesses. I think that if hell gets cut loose on these companies the stock prices will collapse to something like 1/4 of book (see for example the subprime lenders). In that situation you might be getting the right odds to make a bet on these companies. In fact, you might be better off to invest through LEAPS (long term options) since they would either go bankrupt or survive and prosper. If they go bankrupt you lose it all with either common stock or LEAPS but with LEAPS you get more leverage on the upside. I don't think your getting good odds to invest in the common stock now.