Risk Control and my Mortgage and Housing Portfolio

I mentioned my mortgage/housing portfolio a few months ago and here’s what it looks like now:

• Title insurance: FNF/FAF/STC
• Asset pools: AGNC/MTGE/ HLSS
• Origination and servicing: PFSI/WAC. A short put position in OCN that is fully hedged 
• Builders: UCP
• A tiny position in Freddie Preferred.

Combined, these are more than 20% of my portfolio. My housing exposure is actually more if I count Wells Fargo, Citibank…etc.

Since that last post, I have traded in and out of STC with incredible luck, and it looks like my patience in title insurers are now paying off. I’m not so lucky in OCN however. This one killed my returns this year. Analysts are bound to make wrong fundamental calls at some point, but you have to control your losses with sound portfolio management and this is where I failed. 

Getting Scalped by Gamma

Among the many lessons I learned (and paid for), a more interesting one is the negative convexity of shorting options. I got into OCN with short put positions thinking I can subsequently adjusted my net exposure up and down by going long/short stocks. That turns out to be naïve. A simplified example using fake numbers go like this.

Time 1

Stock trade at $34.
My long position: sold 100 shares of puts strike $35, this is now in the money so I’m net long.
My short position: short 100 shares of stocks.
Net exposure:  zero; I’m hedged right?  right?

Time 2

Stock spikes to $37.
My long position: now 0. That 100 shares of $35 puts is now out-of-the-money
My short position:  still short 100 shares of stocks.
Net exposure:  all the sudden I’m net short, when stock is making a run upward! I close my short stocks to bring net exposure down to 0.

Time 3 

Stock goes back down to $33.
My long position: Those sold puts struck at $35 went In-The-Money again.
My short exposure:  0. I closed my shorts in time 2
Net exposure: long 100 shares, but stock is plummeting.

So basically, that short put positions goes in and out of the money at the worst times. What I thought was a fully hedged position could turn into a net short exposure when stock is making a run upward; and vice versa, it turns to net long when stock is tanking.

People talk about “gamma scalping” by going long call option and shorting stock. With my set up I was short gamma and got scalped instead. 

Controlling Risk with Technical Analysis

FNF, and to some extent FAF, are core positions I plan to hold through the cycles. The rest however are not what most people would consider “quality” companies and my positions in them fluctuate greatly. When the fundamentals are shaky and information dissemination is sparse, I learned to use technicals to control my risk. That means buying things near some technical support level (ideally around 52 week or all-time lows), and cut my losses when they drop below that support level. Since that doesn’t always work (some of these stocks are known for taking a big gap downward), I further control risk by diversify my sector bets into multiple names, and look for cheap valuation (low P/Es or P/B multiples) 

Blending of Risk across Sub-sectors

Some of these sub-sectors offset each other with respect to specific risk factors. 

For example, my positions in title insurers could be hurt if mortgage transactions get lower. A partial hedge to that is a position in AGNC. This agency mortgage REIT benefits in that scenario because lower MBS issuance would drive its asset valuation higher. On the other hand, the mREITs have duration risks and could be hurt when rates go higher. The mortgage servicers provide some offset here with their MSR holdings. And so on. 

This is not an exact science because it’s hard to quantify the effect of various risk factors. Nevertheless, it’s good to think through what you’re trying to bet on and the risk you’re exposed to. For now, this portfolio is a bet on household formation, higher mortgage volumes, regulatory environment stabilizing, and various company specific factors such as operating efficiency, low valuation…etc.

Finally, I just started positions in PFSI and WAC this week. Those are for a separate post.

Staying Patient on Title Insurers

2014 has been a tough year for title insurance stocks (as it has been for many mortgage stocks in general) as the group lagged the broader market. However, I believe that title insurers remain the best way to get exposure to housing recovery.

A 30 second thesis on the industry

·         Good industry structure and pricing power. The top 4 players have over 80% of market share. Customers are essentially captive because banks require title insurance for mortgage transactions.

·         Volumes are near historical troughs. Even without any boosts from household formation or homeownership rates, insurance premiums can go up from housing churn and more relaxed lending standards.

·         Expense restructuring. Title insurers had to control expenses through the latest cycles, as well as meet demands from activist investors. Operations are more efficient post-crisis and margins are poised to increase with any volume uptick due to high fixed cost.

Industry Characteristics

·         Product.  Title insurance is generally required by lenders whenever one purchases or refinances a property. Premiums are some percentage of loan amount or property value, with purchases generating higher premiums than refinances. The mortgage industry does not expect much growth in refinance volumes going forward, meaning purchase mortgage volumes will be the biggest driver in the coming years.

·         Players. Top players are Fidelity National Financial (FNF), First American Financial (FAF), Stewart Information Services (STC), and Old Republic (ORI). These 4 traditionally have 80-90% of the market.

·         Pricing. Pricing is regulated by the states and there is very little price competition. As opposed to true pricing power where firms can get away with price hikes, I would say the industry enjoys stable pricing that is very much fixed across the market. The firms also has upside from home price appreciation (remember premiums are a percentage of loan/property amount).

·         Value add. Title insurers are closer to labor intensive service companies than true “insurance” risk pools. “Insurance” in the typical sense of the word is about protecting against future losses yet to incur. However, title insurer actually guard against historical events that ALREADY occurred. As such title insurers can actually minimize losses by just doing a better job upfront (more thorough title search for example).

o   This puts sell side coverage in a weird position. Does the housing analyst cover this?  Or does the insurance analyst cover this? How about the business services analyst?

·         Cost structure. Personnel cost (semi-fixed) are the largest component of expenses, rather than the more unpredictable losses. The combination of fixed cost and relative low margins means earnings can have maximum leverage to volumes gains.

Upside 1:  Macro Narrative

Both total home sales and purchase mortgage volumes are near historical troughs. The current housing environment is marked by 1) low household formation and 2) a shift away from home ownership toward rentals. The mainstream narrative says young people are staying home due to student debt; and when they do move out (thus forming households), they rent instead of own. While that argument has merit, my personal view is household formation will eventually have to pick up, while home ownership rates will have to plateau as rental vacancies decline to more normal levels.

But keep in mind, household formation and home ownership rates are not the only driver of mortgage volume!  In fact, mortgage volume should be more related to total existing home sales (which is a multiple of new home sales).  This means that housing churn and mortgage access can actually be more important than household formation and home ownership rates.

o   Churn measures housing turnover (shown in chart 1 as total home sales as % of year end number of households). Since the late 1960’s this number has trended up with economic growth, dropped during the great recession and now trending up again. Intuitively, as the economy gets better, people will buy and sell houses and move around more, even if the total number of households remains constant.

o   Average mortgage sizes (shown in chart 2 as purchase mortgage volume divided by total home sales) are still at depressed levels even though home prices have recovered. This means either a) lower percentage of buyers taking out mortgages, or b) people take out smaller mortgages (lower LTV loans). Lenders are already in the process of expanding access, so that will help mortgage volume and by extension title insurance volumes.

Title insurers will likely see their premium revenues increase if either, or both, churn and mortgage sizes increase. This is better than say, homebuilders that are depend on new constructions, which goes back to household formation and home ownership rates.

Chart 1:  home sales activity can increase without the benefit of household formation

Chart 2:  mortgage size has room to grow when lending standards normalize

Upside 2: Expenses and operating leverage

The expense picture will be different from each firm and I encourage investors to dig deeper on their own. Just reading through the transcripts though, expense control is clearly a focus for the industry. This is particularly true after the great recession then the refinance boom-bust in 2012- 2013. First American, for example, has condensed its 103 claim centers, 30 accounting centers and 30 data centers in 2006 to 4, 2, and 2 respectively today (source: conference transcript). These are structural costs that are not expected to come back when purchase volumes come back. As a result, FAF now sees a 10% pre-tax margin as the new floor, as opposed to the ceiling it was during pre-crisis days.

Expense initiatives are hardly limited to FAF. FNF and STC have both attracted activist investors in the past couple years and management teams are on tight leashes regarding expenses.

It’s not just the level of expenses improving either. Expense will be easier to manage going forward because purchase volumes are more predictable than refinances. Refinance volumes are very sensitive to rates so companies had to quickly ramp up and ramp down their staff. The transmission goes something like this:  rate volatility -> refinance volume boom/bust -> difficulty in staffing  -> inefficiencies  -> earning volatility. Going forward though, a primarily purchase driven market should be more predictable and thus costs will be easier to manage.

A better blend of risk vs housing related subsectors

Why title insurers versus other housing/macro plays? The table and discussion below will outline how I mentally think of the various housing sub-sectors.

·         Competitive risk. Housing subsectors like home builders, loan origination, mREITs are typically fragmented and competitive. Title insurers and non-bank servicers are the only subsectors with highly concentrated players.

·         Regulatory risk. Non-bank servicers are currently fighting through a host of regulatory issues. Title insurers could have some risk here also, as the uninitiated tend to think of it as a sham product. However as one does more research they realize the protection is necessary.

·         Consumer credit risk. Credit losses are currently minimal but are bound to increase as lenders fight for market share by expanding credit boxes. If you don’t like the idea of normalizing losses (or already have enough in your portfolio) you can screen out origination, mortgage insurers, as well as some of the nonagency mREITS.

At this stage of the cycle, competitive and regulatory risks are my primary concerns. By process of elimination this leaves title insurance as the least risky way to get housing exposure.

Recap

I will leave off at this point. You have an industry with concentrated market power, volumes at a trough but normalizing, and expense running at efficient levels. Note that I have not discussed valuation. However, if you have a positive view on housing in the long term, this should be the best sub-sector to look into, given the better risk blends compared to other housing plays.