Economics of mortgage insurance vs mortgage servicing vs ?

[scorpioncapital]

Which businesess related to mortgages do you think has better economics? Both servicing and insurance seem to be high volume, low margin businesses and both seem to be in some form of long-term run-off.

[Tug Bovine]

i prefer the private mortgage insurers. you can make a pretty strong bull case that they aren't in runoff - the FHA is actively trying to reduce market share, which should allow private mortgage insurers to grow business in an flat origination environment. I like the legacy MIs because i am also bullish on the credit side - i think they're adequately reserved (at a minimum) and that as the old books burn off, loss ratios will come down.

 

dont have much of an opinion on mortgage servicing - haven't looked at them in a few years when the bull case was buying MSRs from the big banks who were forced to divest them for capital reasons. seems like there's been some regulatory issues as of late. is there something to do here?

[scorpioncapital]

I see servicing as sort of a low margin collection business. I'm starting to conclude insurance is a "cleaner" way to play the space too and returns are good at the moment, generally seeing 20% returns on writing what in effect is a deep out of the money option (e.g. 20% payout on a 750+ FICO score customer where they put in 5%-15% down, if a default occurs and the bank can't sell the property except for a big loss). There's a lot of things that must go wrong for such a payout to occur. In the meantime, you collect these premiums and with rates going up the investment portfolio could generate respectable leveraged returns.

 

PS... do you know why FHFA mortgage insurance rates are double to triple the price of private MI rates? I think this fact is the key to a real arbitrage and why this sector is poised for growth. The chart below is from an Ocwen presentation and based on a $200,000 typical mortgage. Clearly, private MI can't be underpricing so severely so I'm trying to understand what's happening.

[Tug Bovine]

FHA rates are going up because the government is intentionally trying to reduce market share and get private capital back in the mortgage market.

The FHA's capital position has become worse over the last few years and raising premiums should also help build capital to repair it.

 

------

FHA TAKES ADDITIONAL STEPS TO BOLSTER CAPITAL RESERVES

Continuing effort to help strengthen FHA’s Mutual Mortgage Insurance Fund

 

WASHINGTON – As part of a broad effort to strengthen the Federal Housing Administration’s (FHA) Mutual Mortgage Insurance Fund (MMI Fund), FHA Commissioner Carol Galante announced a series of changes to be issued this week that will allow the agency to better manage risk and further strengthen the health of the MMI Fund.

 

“These are essential and appropriate measures to manage and protect FHA’s single-family insurance programs” said Galante.  “In addition to protecting the MMI Fund, these changes will encourage the return of private capital to the housing market, and make sure FHA remains a vital source of affordable and sustainable mortgage financing for future generations of American homebuyers.”

 

http://portal.hud.gov/hudportal/HUD?src=/press/press_releases_media_advisories/2013/HUDNo.13-010

[scorpioncapital]

Makes sense, thank you. Another puzzle for me is the valuation of the mortgage insurers. Most of them trade at 2 or 3x book value. This is a little "alien" to me for insurance operations. I am used to the idea of 1x book value as a starting point. Is it simply due to anticipation of a high growth rate or is there something about the value of the underwriting that makes it more valuable? With servicing, I can wrap my mind around the value of servicing rights and free cash flow yields while the servicing runs off.

[Tug Bovine]

it all comes back to ROE for me. I'd argue that ROEs for new business is probably somewhere around 20%+. it will come down over time as market share comes back from FHA biz (which is of slightly lower credit quality) and insurance books should grow a bit as well (although i think books should remain more or less flat y/y in '14)

 

these ROEs are possible due to barriers of entry - we have seen a few like ESNT and NMIH - but you need management that fannie/freddie trust as well as relationships with the mortgage originators. you can't compete on price either - last time someone cut price to gain volume, all other players cut as well - which made everyone worse off. Tug Bovine Mortgage Insurance sounds like a winner - no legacy book - but the MTGs of the world will say "the world blew up and we're still here - and we've paid out every single claim".

[scorpioncapital]

Yes, I believe today you can write high FICO insurance and generate 20% return on statutory capital. What I've noticed is that the insurers trading at 2-3x book. So let's say leverage ratio is 16:1 (which appears to be the line in the sand for regulators on leverage limits to be compliant with Fannie and Freddie certification) and you have 500 million in statutory capital. You can write 8 billion and earn 20% = 1.6 billion. Add your paid in capital and let's say 2 billion = 4x book. Obviously this is "fair value" and there should be some discount and profit potential for the investor coming in. I would imagine if you paid 2x book in the above scenario, you'd have a potential to double your money. This is why it seems puzzling the multiples are so high, where is the profit potential for the incoming investor if he is "pre-buying" the future profits unless you assume higher leverage than 16:1 and/or higher than 20% returns. Also, you may assume some return on the premiums. For simplicity, I usually just assume that the discount rate for the earnings = the return on the investments. If this holds, you can cancel out and assume the above numbers directly.

[Stevieoopsie]

For MSRs

 

It depends on the cost of servicing. In the case of servicing prime agency loans, most players have more or less the same automated processes, whereas in servicing sub-prime non-agency stuff, some companies are more efficient at that than others. The non-bank servicers have demonstrated that they are much lower cost than the big banks at servicing mortgages.

 

Another thing to consider is how the servicing portfolio is financed. The new trend is to shift the financing (mostly advances) to public REITs (which have a lower cost of capital because they trade on a dividend yield basis) which turns servicers into sub-servicers. Some non-bank servicers use that structure to lower their capital intensity and enhance ROE.

[ItsAValueTrap]

Mortgage servicing seems to have better economics.  But in both industries I think that it comes down to management.

 

Insurance is generally a bad business.  But those who are really good at it can make a lot of money.  Look at Buffett's history.  He loves well-managed insurance companies and otherwise shuns the insurance industry.

 

Mortgage servicing is inherently a high-margin business.  Contractually, the servicer is intentionally overpaid.  (That leads to excess servicing rights.)  If servicing rights are sold to another party then it's arguably a "low-margin" business if the purchaser has financing costs.

[scorpioncapital]

I would think both servicing and private MI are both low margin businesses. In one, you generate a fraction of a percent on servicing the loan and in the other, you receive a fraction of a percent of the total mortgage as a premium. I believe companies like Ocwen are outsourcing their collections to India to reduce costs, how much can they make of the total mortgage for this function?

 

MI appears like a business of writing deep out of the money options. You collect a pittance of a premium but you don't expect to be exercised (or rather you don't expect - in the aggregate - to pay out a large part of your premium as many events have to happen for a default to occur). So the constraints are a) how much business you can write and b) how much money you make in relation to your equity base. Just to see how razor thin it is , you are making like 0.5% in premium on the amount of the mortgage. Right now return on statutory capital are very fat - something like 38% if you include the unearned premium. But it would seem that reserves for losses and a buffer require an equity margin above and beyond minimums. Thus I see returns in the mid teens to 20%. The unearned premium appears to be some sort of float while the earned premium is pure profit if the reserves for losses are accurate and things go well.

[yitech]

There may also be movements toward MSR-backed debt like Ocwen did to properly hedge the prepayment risks. They were able to sell those at higher multiples overall than they bought.

[ItsAValueTrap]

I would think both servicing and private MI are both low margin businesses. In one, you generate a fraction of a percent on servicing the loan and in the other, you receive a fraction of a percent of the total mortgage as a premium. I believe companies like Ocwen are outsourcing their collections to India to reduce costs, how much can they make of the total mortgage for this function?

 

Mortgages are usually securitized in such a way that the servicing rights are cash cows.  The party that owns the servicing rights gets paid a lot more than what its costs to service the mortgages.

 

The actual cost of servicing mortgages depends on a number of factors and the income stream from servicing rights also depends on a number of factors.  (Uh... it's a little complicated.)

 

The cost of servicing has gone up due to new regulations and the unusually high level of foreclosures.  The high level of foreclosures reduces the profitability of MSRs because the servicer is paid based on the unpaid balance.  The servicer has some incentive to avoid foreclosures.  However, it costs money to do this because it's a different ballgame that collecting money from homeowners and doing paperwork.  You need to hire people to communicate with homeowners and to process loan modifications.

 

I'd encourage you to do more research and to read the prospectus for a mortgage-backed security (they're on EDGAR).