A TrueCar Short Case - Victim of Its Own Success

Situation

(I started looking into shorting this when stock was trading at $20 but did not pull the trigger. Now it is around $17 going toward $16, have I missed the boat?  I’m hoping for TRUE to go up so I can short it... )

At $20/share, TRUE’s stock trades at 8x sales, ~150x forward P/E. LTM EBITDA and cash flows are both negative. As a result people have started looking into shorting the stock. A couple of short articles here:

http://seekingalpha.com/article/2542615-short-truecar-first-mover-first-loser

http://seekingalpha.com/article/2549955-8-reasons-truecar-is-a-sell

Here I will layout my own short case from a different angle. My thesis different from others in that I think TRUE is an innovative company with a real value add, but ultimately will be undone by its lack of competitive moat.

What TrueCar Does and How it Makes Money

TrueCar (TRUE) does online lead generation for car dealers (mostly new cars). Its main property is TrueCar.com website where car shoppers can see a range of prices that other people paid for the same car in their area. Where TrueCar differs from other website is the idea of binding price. Users can click a button on the website to get a price offer that dealers have to honor. In terms of revenue model, TrueCar is free to customers and dealers pay $299 for each successful transaction. The company also gets traffic from affiliates such as USAA, and TrueCar splits the fees there.

Understanding the Bull Perspective and Reframing the Question

The over-valuation problem is exaggerated. The market is clearly looking at something else aside from > 100x P/E. Management’s stated long term goal is raising market share from 3% to 10%, and raising EBITDA margin from 2-3% to 35%. If these goals are achieved, they would more than double their revenue and generate about ~$175mm of run rate EBITDA. With < $2bn of enterprise value, all the sudden you're looking at low teens multiple –arguably cheap for a business with high return on capital (if they hit that margin goal).

These plans are not as crazy as they sound. Revenue has been growing at 50%+ pace so doubling in a few years is not unreasonable. The biggest expense is sales and marketing. The idea is to slow down marketing spend once the company reaches a critical mass of customer recognition. In fact, the company can actually grow advertising dollar amount slightly and still decrease the percentage cost due to operating leverage.

So that’s what the market is looking at. As a potential short seller here, the question should be reframed as “do we have a high degree of confidence that revenue can't double and margin can't hit 35%?”

TrueCar’s Value Add – Why Customers Use Them and How That Can Change

I don’t doubt revenue can double but I question the ability for a company to taper down marketing and advertising expense, and still maintain or grow site traffic and drive car transactions. Ultimately, this can only happen if your product is that much better than everyone else’s (and able to sustain that advantage through the period we’re evaluating).

Why do customers use TRUE and not a substitute or a competitor? And what would change that? There are two customer sets here:  potential car buyers and dealers.

·         Car buyers.

o   Target customers are those who do not like to negotiate and distrust dealers. These car buyers are “satisfiers” instead of “maximizers” – they are not looking to buy a car for the absolute lowest price possible. Rather, these customers just want a fair price and not walk away with some lingering suspicion that they’re ripped off.

o   Where does this fear of getting ripped off comes from? It comes from wide dispersion of prices among dealers.

o   Arming customers with information is also a value-add but competitors do this as well. The “marginal value add” of TRUE is upfront, committed pricing (via price certificates).

·         Dealers. 

o   TrueCar replaces less efficient advertising spends.  TRUE will never replace all of dealer marketing because dealers still want to build their own brands, so TRUE can only aim to replace less efficient advertising budgets (most obviously less efficient lead generation firms)

o   convert ad spend to variable cost (because dealers pay only for successful transactions)

The points above imply that TRUE’s edge goes away if:

For car buyers, TrueCar’s value add goes away when the day comes that price dispersion is minimal. The industry is already moving toward One Price, with KMX, Sonic, Tesla…etc. all pushing some variant of pricing commitment. AutoNation has even talked about moving to directly selling cars online. If pricing transparency and commitment becomes the industry norm then TRUE is useless.

Similar things happen if competitors implement pricing commitment. CostCo and Edmunds already try to do no-haggle pricing, it’s not too far of a step toward offering a price commitment like TRUE does.  One can argue TRUE’s advantage here is tying into dealers inventory systems, but perhaps others can get information from DealerTrak or other sources.

For dealers, TrueCar’s comparative advantage goes away if other sources of advertising becomes more efficient, either by copying TRUE’s model or if they are forced to cut prices. As the market share of these relatively inefficient forms of advertising gets taken by TRUE, it’s hard to imagine they won’t increase quality to compete better

Ultimately, TRUE will be a victim of its own success because it is innovative but easy to duplicate. In the next few years, I can see the company push the car industry further toward One Price paradigm, force competitors to offer pricing commitment or becoming more efficient.  All these things are great for consumers, but diminish the necessity of using TrueCar. Without product advantages, it’s hard to keeping advertising down and still double revenue. In some ways, not shorting TRUE requires you to think that auto sales processes are still the same in 5 years and that other player can’t improve their game despite being eaten alive by TRUE.

  

Appendix:

Bear argument

·         The Industry is already moving toward one price and TRUECAR's value-add will diminish as price divergence decreases.

·         Unlike sites such as Yelp or Facebook, TRUE does not have the benefit of user networking externality. People use TrueCar because they want a fair price, not because all their buddies are hanging out on it, nor because they want extensive dealer reviews. As soon as the proposition of “fair price” diminishes (due to everyone already offering that), TRUE becomes unnecessary.

·         What happens when new car sales go south? 

o   Used car pricing are coming down due to increasing off-lease supply. This can spill over to new vehicle market. OEMs will defend sales with incentives but eventually lower pricing should cut into dealer margin.

o   When dealer gross margins are under pressure, Truecar's $299 per sale look less attractive. Dealers will push for lower cost lead generation or take advertising in house. Dealer consolidation would give them even more bargaining power.

o   As TRUE wins over other sources of advertising, these sources will lower their prices to compete.

o   Lower industry volumes will hurt the company.

·         Operational failures

o   I read that some dealers have dispute with TrueCar about how a successful lead is defined and how dealer has to pay.

o   Bait and switch at unscrupulous dealer can become legal liabilities for TrueCar

Bull argument

·         TRUE is an innovative and visionary company with a history of disrupting the industry. When TRUE first came out, the transparent pricing caused dealers to undercut each other and customers got the cheapest prices possible. This led to an industry-wide boycott by the dealers and near death for TrueCar. TRUE has since changed its business model to not let each dealer see each other’s pricing (so they don’t undercut each other).

o   Incredible track record of turnaround in dealer relationships. The company has doubled dealer participation from its trough. In fact, the leader of this dealer revolt, Jim Ziegler, is holding an upcoming conference and will have the President of TRUE as a keynote speaker.

o   TRUE is a cutting edge big data company. They are transforming a mom and pop business by leveraging new technologies such as Hadoop. 

·         TRUE as a potential M&A target

o   Combining with lead generation competitors would decrease the need of an advertising arms race and allow margin gains.

o   Vertical integration (True+ KAR = TrueKar?) TrueTrade is trying to compete with CarMax (KMX) by enabling a network of used car dealers to offer committed pricing on trade-ins, but without the need to hold inventory. Now, one reason KMX can do this competitively is because they run their own auctions, which helps KMX dispose unwanted trade-ins economically. A merger between TRUE and Kar Auctions Services (KAR) can replicate KMX’s functions. This might be interesting for KAR because its main competitors Manheim (owned by Cox, which also runs AutoTrader and offer upfront pricing on trade-ins) and KMX are already parts of larger entities. Also, TRUE owns ALG which would be extremely valuable to KAR.

·         Increasing part of a dealers marketing budget. Dealers still spend on newspaper, radio, and TV so there could be some low hanging fruits there. If automobiles demand decreases, TRUE may actually gain market share because dealers would want to shift marketing to TRUE’s variable cost model (they only charge for successful deals)

·         Margin increase can be achieved by lowering marketing cost once a critical market share is achieved.

·         Opportunity to go beyond dealer advertising spend and grab a share of manufacturer’s advertising and incentives.

UCP Lot Valuation

Summary: 

• I was skeptical of the long thesis on UCP because analysts tout strong inventory valuations but glosses over the fact that it takes time to realize value - i.e. they do not apply a present value.
• It turns out that the approach of valuing lots by ascribing value as % of home prices is flawed but surprisingly robust. This is because the 2 missing factors are offsetting: 1) Applying a discount rate would decrease value, but 2) this method also understates cash flows and correcting for that increases value. 
• Valuation is $18/share.  UCP trades around $12 per share at the time of writing.

UCP and Its California Inventory

UCP, Inc. (UCP) is a land developer turned home builder that historically focused on California. Its operating subsidiary, UCP, LLC is 57.5% owned by PICO. As of 6/30/2014, UCP owns 4,639 lots and LTM net new order was only ~300 units. UCP has enough inventories to last years, so valuing the company requires valuing the lots.

When people hear that UCP owns lots in California, they think of San Francisco, Silicon Valley economy…etc. and assume high prices. However, UCP mostly operates in inland locations like Fresno and Madera where it could take 3 hours to reach the coast. According to Zillow, median home prices for Fresno and Madera are less than $200k. Other areas where UCP has strong presence are Monterey County in CA (prices in the $400-450k neighborhood) and Thurston County in Washington (low $200k’s range). These are not exactly your premium million dollar homes in San Francisco. UCP acquired Citizens Homes earlier this year, which allows them to expand into the Southeast.

The Long Thesis

This is an asset play. Much of UCP’s land inventories were bought during 2008-2009, when lots are cheap. These are now worth much more than what market value would suggest. With enough inventories to last several years, UCP also has some nice options:  a) monetize the lots by building homes themselves, b) sell lots to other builders, or c) sell itself to another company. Being acquired is a possibility due to the small market cap and their local footprints.

Valuation

The bullish write ups I have seen tend to value the inventory without a discount rate, then net out the liabilities to result in a massive net asset value. There was a PICO article published 9/8/2014 on SA included detailed valuation of lots by locations. The methodology separates inventory into developed and undeveloped lots. For developed lots - assigns a reasonable home price for the area (say $450k for Monterey Bay area), and assume lots are worth certain percentage of home value (50% here). For undeveloped lots, just assign value per lot based on location, using UCP’s historical revenue per lots sold for context.

I duplicated the exercise below and got a very similar ~$350mm of value for the lots (164mm for developed lots and 183mm for undeveloped). Netting out liabilities and minority interest this would value UCP at ~18 per share.

It's tempting to stop here without applying a discount, and point out as upsides the optioned lots (which are not included here) and UCP’s potential as M&A target.

There are 2 things missing here. The first is obvious - it will take years for UCP to monetize this inventory and thus time value discounting is not only necessary but makes a big difference. Clearly, it takes time for local housing demand to emerge, and for UCP to build and sell those houses. The second missing piece is what amounts should we be discounting?  Just spreading the $350mm of total value and discount back would be a mistake, because the total cash flows UCP will get is actually more than that.

The second point requires some explanation so bear with me. What exactly do we mean when we say a lot is worth x dollars?  An excellent article here makes it clear that the value of a land lot is what another builder is willing to pay for the lot, and still be able to generate some profit. The profit part is key. For example, if other builders are willing to pay UCP $100k to buy some inventory, incur another $80k of construction cost, and sell the house for $200k, their profit is then $20k. The lot to ASP ratio would be 50%. How much cash flows will UCP get for this lot? 

·         If UCP sells the lots to another, then you should discount just the $100k proceed from the lot sale. 

·         UCP however, plans to build the homes itself. In this case UCP gets $120k of incremental cash flows ($200k selling price – $80k construction cost). The other way to think about it is they get $100k of working capital back, plus $20k profit.

This is why the $350mm value above understates cash flows - because it implicitly assumes UCP will sell these lots to another builder, and thus fails to include the profits UCP will get upon a home sale.

I get to roughly $18/share. It turns out that applying a discount rate is offset by this extra cash flow - even with some very conservative assumptions about sales pace.

Value destroying growth is always a risk

Growth ambitions are dangerous in homebuilding. At the peak of cycles, financing becomes widely available so builders gorge themselves with expensive inventories. During a crisis when they should be buying cheap, they can’t because of financing constraints. In short, builders always risk buying at the worst times.

By virtue of having land that can last years, UCP can theoretically avoid this type of behavior. In my opinion, the best case for UCP would be to stop trying to expand, and focus on monetizing its existing lot inventory as quickly as possible. Of course, management teams never want to liquidate themselves and UCP has in fact been acquiring.This makes them more like other builders, but with subpar scale and diversification. Business expansion also makes UCP a less likely M&A target (which I wouldn’t count on as the basis for an investment anyways). These factors perhaps call for a higher discount rate than peers, but still result in a strong valuation.

Management Turnovers at Pharmacyclics (PCYC)

So someone told me about this company with a wonder drug. She loves the product but got this weird feeling about management. Maybe it’s the way management interacted with each other on the latest earning call, or the way they answered analyst questions. She could not put a finger on what it is.

She also told me they just got a new chief commercial officer but the chief medical officer (CMO) just left. So naturally I googled the CMO’ name. Multiple names popped up. Digging deeper here’s the summary timeline I found:

·         Dr. Ahmed Hamdy - appointed CMO March 2009

·         Eric E. Hedrick - interim CMO sometime around 2011

·         Lori Anne Kunkel - appointed Dec 2011; departed July 2013

·         Jesse Seton McGreivy  - departed Aug 2014

So you have 4 CMO’s in 5 years.  When a company’s CFO or Chief Accounting officer leaves, you worry there’s something wrong with the numbers. But how about when a biotech’s CMOs keeps leaving?  Do you worry that the product is fake?  Is it even possible to fake your way through FDA approval?

More research. What is wrong with Pharmacyclics, Why would top executives keep leaving?

·         Nothing positive in CaféPharma. Let’s just say this is a highly entertaining board. You got threads  named “Pharma-stall-ics”, “Pharmasucklycs”, and “Pharmafuckyclics”.. .etc. Definitely some employee relationship issues here. There are also widespread mentions of wrongful termination suits.

·         A post on investor hub paints an unflattering picture of COO, and mentions former CMO Lori Kunkel.
·         Bloomberg  article. Lots of insinuation here when the article talked about Duggan’s association with fraudster Slatkin. The article spent almost no time on what Duggan did at Intuitive Surgical (which would have added to Duggan’s credibility).

Obviously all of the above could be written by people with agendas. The fact though, remains that you got 4 CMOs in 5 years at a biotech. I doubt CMOs leave due to “work life balance” issues because these are overachievers and probably workaholics to begin with. It’s also hard to explain this as CMOs asking for big pay raises because well, you should pay them.

I get the sense that this tight clique of Duggan, Zanganeh, and Erdtmann calls all the shots. Given the rampant turnovers and the colors above, I can only infer that at best that the top management are unpleasant dictators, at worst there’s something unethical going on.

But, they have an awesome product!

Does all this matter when you got a hot product like Imbruvica?  In general, when do management matter the most?

First, if this is a fraud then obviously all bets are off. Again, is it even possible to fake your way through FDA approval? Granted that EVP of Corporate Affairs Ramses Erdtmann is a Scientology Operating Thetan VIII, which according to Wikipedia gives him the power to "control others from a distance" and "create illusions perceivable by others", the probability of a fraud getting through FDA has to be pretty small.

Second.  If it’s not a fraud but top management are major tyrants, does that matter? I think it depends on context:

o   If the company’s valuation depends on the ability to keep innovating and create demand (think Apple and Steve Jobs), then management competence matters a lot, but management likability not as much (again Steve Jobs was known to be a bit of a dictator).

o   If this is a mature / growing company trading on say 15-20x earnings, management have to optimize revenue, control cost…etc. Clearly management matters a lot more.

o   In PCYC’s case though, the company is trading on vast market potential of a single product, Imbruvica. The patient demand already exists. The product is already there, theres no more innovation that’s needed. The science either works or it doesn’t, and there’s nothing management can do about it.

Ultimately, this comes down to how PCYC fits into your investment style. If you’re allocating to numerous small positions with catalysts for quick pops, then management matters less. If you’re trying to find that rare company that's built to last, then I'd say this is not the situation for you.


**** Updated 10/2/2014 (originally posted on Seeking Alpha Instablog)***

I normally don't like to dwell on management too much. In fact in my blog post here I concluded that management turns at PCYC can arguably be a secondary consideration depending on your investment style.

Then I learned that the drugs are made in China. Why would you do this? So you have a biopharma who keeps losing medical/science personnel; core executives couldn't answer questions about IMS data in their own presentations (and get yelled at by the Morgan Stanley analyst). And oh, by the way the drugs are made China.

I don't have the guts to outright short this company given strong reviews about Imbruvica, but at some point the red flags pile up and I stop looking further.

Random thoughts on CarMax, Oracle, and Housing Vacancies

·         CarMax (KMX)

o   Funny that every analyst wanted to ask about subprime on the earning call. I get that KMX is arguably a finance company. But guys, falling subprime mix is a GOOD thing!! So what if revenue slows a little bit, to the extent that customer base is more sustainable, that’s good news.

o   KMX does look expensive even with today’s drop off. From lenders perspective though, it is good to hear that subprime players are tightening standards.

·         Oracle (ORCL)

o   This Barron’s article said that Oracle is threatened by Hadoop. Ironic considering that Oracle oversees Java – the language that Hadoop is written in. I have also heard that Oracle is hurt by freely available database options. Well, ORCL also owns MySQL, one of the most popular free databases. If Hadoop and free databases are really the downfall of ORCL, this needs to be a business school study on making your stuff open-source and freely available.

o   Hadoop does not replace a database. The Hadoop wiki says as much. Hadoop is great for unstructured data (for example if you’re mining terabytes of tweets) while traditional relational databases are good for structured data (in the row/column format). Hadoop is just a way to split up your job to various computing and data resources, each of those could be different form of data storage, including a database. In fact, Hadoop and database can be complementary - there’s just so much information in databases that someone mining data will have to link up Hadoop with relational databases.

·         Follow up on last week’s post about housing stock – how to find true vacancy numbers?

o   I can’t stress enough that the reported “homeowner vacancy” and “rental vacancy” numbers are just fake. There are substantial “other vacancies” that are not included in those numbers – easily 30% or even 50% of total vacancies depending on location. Here’s a helpful report that US Census put out on “Other” vacancies last year.

o   How to find the true vacancy number then!?  Those numbers are available in the American Community Survey. Unfortunately the US Census does not make this easy. To be useful you’re really looking for local statistics. FactFinders allows you to get this by entering the MSA’s one by one. But if you want to figure out say vacancies for say, all the exposures of some home builder, then this will take a LONG time.

o   Ideally you want historical time series for each local level so you get a sense of “normal”. You can try to download the ACS summary files, but those only go back to 2005 on the Census website. 

o   One way to do this is with Public Use Microdata Samples (PUMS). The U. of Minnesota has a great site that let you select the variables and the vintage years you want. Load that into a database (it's easily in gigs of data) then process it however you want. The results will not match the ACS summary data exactly because these are samples of the original survey. But at least you get a sense of the vacancy mixes going further back than 2005.

A Mental Model on Housing Stock and Flow

I often hear people say something like this: “household formation should be 1.5mm per year and new construction are running 1mm per year, therefore we’re facing a housing shortfall”. 

The obvious flaw with this statement is that it ignores existing inventories and focuses completely on trends and “flow”. Put another way, it assumes inventory is already in balance. The logical questions are then:  how about the existing inventory?  How do you know we didn’t overbuilt so much during the last cycle that there are still still excess home supply? 

Intuitively, I’d lay out household formation and housing starts (“flow”) against total number of households and housing units (“stock”) like this:

Assessing whether we’re overbuilding is then a 3 step process: 1) estimate the number of households, 2) estimate number of housing units that can be occupied, 3) compare the two numbers; if there’s a housing unit shortfall then that’s the number of units we need to build. 

In the example above, I started with 2013 number of households. An estimate 750k of household formation for 2014E gets me to 2014E households of ~115mm.

Next step is take the housing unit numbers and figure out how many of those are actually available to live in? This is where subjective judgment comes in. The economy is not perfectly efficient, so at any given time, there’s a healthy amount of vacant units in transition (it takes some time going from a rental listing to actually renting out the unit, a unit could be sold but the buyer has not moved in yet…etc.) These are units that are not available, so I’ll take those out. I also remove a normalized amount of second homes and “held for market – other” units from the stock. For 2013, I estimated ~12.2% of housing units are “normalized vacants”, and removed a corresponding 16mm unit from stock. This resulted in an estimated 116.6mm housing units that are actually available to be occupied.

Finally, compare 2014E households of 115.4mm vs 116.6 of available units at year end 2013 and it’s clear that we still have excess inventory. I expect this excess inventory to decrease only slightly at the end of 2014 because household formation barely exceeds net unit adds. 

Alternatively, I have also seen analyst keeping track of the stock of vacancies, and map out the difference between demand (household formations), and supply (housing starts, demolitions…etc.) as a burn rate against excess vacancy. This is slightly more elegant but should get you similar results. Either way, the point is you have to take into account existing stock (whether in terms of available units as I did above, or as inventory of excess vacancies), rather than just compare housing starts against household formation.

* As a note, the headline reported homeowner and rental vacancies can be misleading as they both understate total vacancy. The way these numbers are reported: if census can’t categorize if a vacant unit is for rent or own, that unit will be left out of the data. i.e these numbers exclude units that are held off market or seasonal vacant. For this reason I focus on total vacancy for a big picture (construction activities). Then only drill down to homeowner vs rental vacancy when I evaluate rent vs own type of decisions.

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.

MGIC Investment Corporation (MTG) - Normalized ROE does not look attractive

This could turn out to be a multi-part project. I just started looking at this company last week so my insights would be limited. However, I’m writing anyways as it helps me gather my thoughts and move forward.

An Inherently Unattractive Industry

MGIC Investment Corp (MTG) is a Private Mortgage Insurer (PMI). This is one crappy industry. Results are incredibility cyclical and sensitive to some assumptions. On the competitive front, we have already seeing new players trying to compete on not just price, but underwriting terms (NMIH Holdings is an example). Now it looks like the 7 players will not be putting a united front with respect to proposed capital requirements.

Why would anyone buy a PMI then? The standard long thesis says 1) FHA will be giving away market share to the private sector, this coupled with an improving mortgage market will lead to high volumes. 2) Lower losses from legacy vintages running off, as well as fixed cost/operating leverage would lead to a spike in earnings.

Some parts of this hypothesis are valid but I think the sell side tends to bake in both 1) mortgage industry recovery, AND 2) losses stay low at post crisis levels, when in fact the two may not be compatible. While losses are sure to come down from present levels in the next couple years, some analysts seem to assume the pristine underwriting quality of 2009-11 vintage will continue forever. Given that 1) first time homebuyers will be needed to drive housing recover and 2) they tend to be of lower credit quality, the assumption of “2009 forever” is clearly unrealistic.

So you have multiple offsetting factors at play and investing in the sector requires one to say “ok, new businesses will have higher losses at some point, but I don’t think it will be that bad, and meanwhile this thing is dirt cheap and I like the risk and reward”.  This reminds me of my write up on Santander Consumer USA. While there’s a place in your portfolio for a speculative play like this, these could be good ideas but not the best ideas.

Normalized Losses and ROE

Unattractive industries demand great valuations. I want to get some sense of economics and value before filing this away. To do that, some conception of a “normalized return” is needed. Anyone who has followed the industry would know that current loss levels are far from “normal”, since MTG is still working off legacy businesses. So blindly applying some P/E or P/B ratio to next year’s forecast would be meaningless. 

Below are my estimates of new businesses economics when losses normalize, under the current capital regime, and then under the proposed PMIERs.  I’m thinking about this at the opco level so that’s why there’s no interest expense.

 

Two part discussion here, first on impact of capital requirement, then on normalized losses.

First, capital requirements. Management said that if PMIERs goes through, capital requirements for recent businesses probably correspond to 11.5-12x in the old risk to capital framework and that gets them to low teens ROE before reinsurance.  

From 2Q14 transcript: “…Under the proposed eligibility requirements, the mix in the first half of the year seems to require a risk to capital of about 14-to-1 at time of origination, i.e. they are all current. But as we know, even with the high-quality profile, some will go delinquent. So, if you factored that in and probably goes to 13-to-1. And then if you want to add some room away from the capital requirement just to give yourself some margin you probably talking 11.5-to-12-to-1.

And by our calculations, on a direct basis before any reinsurance and whatnot, we think that delivers a return in the lower double-digits. On the current or prior to eligibility requirements that we are issued here, we were think and closer to 18-to-1. And if you give yourself a little room and whatnot operate around 16-to-1. We think those returns are kind of back as Curt said where they probably should be for the overall risk of the business in the mid-teens.”

So let’s say losses will revert higher in new vintages going forward, I think a 10:1 risk to capital, single digit ROE after reinsurance is probably reasonable.

How about normalized losses? In the 2Q14 call, management said the 2009-2011 books are running about 15-20% loss ratio. This is how I got the 35bps loss as % of RIF assumption (50bps premium * 17.5% loss ratio / 25% RIF = 35bps). Keep in mind 2009-2011 vintages are loans with pristine underwriting standards. Going forward as the mortgage industry reach down the credit spectrum, it’s reasonable to think that losses will be higher. How much higher? To give an idea of how volatile these items are, below are historical loss ratios from 1996 to 2006 before the whole industry blew up. I used 50bps credit loss in the above table as a placeholder, but the ROE sensitivity table is all over the place.

So you got a cyclical, competitive industry getting hit with higher capital requirements. What should investors demand? The CEO gave some jumbled answers on the 2Q14 call but I think he meant to say mid-teens return overall and high teens for low FICO/High LTV businesses:

GS analyst: “Now, but if you were just, say, isolated – let's say you were the only player in the industry in a very hypothetical scenario, I mean, what required return would you want to get on those lower FICO, higher LTV buckets? I mean, would you be looking at low-teens? High-teens to account for some of the greater volatility in those buckets?

Curt S. Culver: “Yes, but I, on your question, I think for the lower FICO you need a mid-teens minimum return, given the variability on that business and how quickly things can change. So, certainly it demands a higher return. The returns on the other business will be, I think low-to-mid teens so that certainly would require in my opinion, a high-teens return.

So management want double digit returns but the analysis above shows that ROE with 1) normalized credit loss, and 2) new capital regime will likely be in the high single digits. And MGIC trades at 3x book value when the only thing we can really count on is volume growth. Surely there are better ways to play a housing market improvements? 

Up to now, I have referred to new business economics at the opco level, assuming equity capital = investment assets. In reality there's a mix of vintages books, the investment portfolio is much higher and there's interest expense from the holdco debt. To value MTG you have to take those into account. I won't bore anyone with the model here but my calculation shows that MTG is about fairly valued right now ($8.4 per share).

A Cash Flow Model

REIT Analyst did an SA article on MGIC last week. Specifically, he actually tried to project out the cash flows of mortgage insurance premiums and losses, then calculate a present value. Financials analysts as a group tend to stay away from cash flow statements, so what he’s doing is very different. That and the result of over 100% upside got my attention.

I can understand why his valuation is so much higher than market and what I have above. My guesses are 1) discount rate used - the market is rightly demanding more than 10%, 2) Not all investments are excess. Put another way, they are operating assets required to back the day to day MI business, so the investment income stream needs to be discounted together with other operating cash flow streams.  3) subtle assumption difference in premium, loss curves, reinsurance...can all make a big difference. 

Nevertheless, some sort of cash flow analysis would be useful to quantify the positive effect of legacy vintage running off - a big part of the long thesis. A real deep dive here would mean building a full cash flow model myself. At this point it is not a high priority given all the negatives I discussed earlier.

For now, I say we give this guy the sensitivity table and a set of darts, and call it a day.