Description

I am long Rosetta Resources.  This is an E&P company with “A” assets and “A” management.  The stock is at $47 and I
think it’s worth $70 under the company’s current development plans, assuming a $90 WTI/$4.25 Nymex price deck. 
Further upside may come from three drilling opportunities that the company hasn’t quantified yet, but we know enough
about them already to believe they have decent odds of materializing and that they could be substantial.  Finally, there
are some near-term catalysts that could help crystallize the company’s value. 

On the downside I think the stock could be worth $42, assuming $75 WTI/$3.50 Nymex, even without the incremental
drilling opportunities. 

Company Detail
Rosetta is a Houston-based independent oil and gas producer.  They use full cost accounting and their reserves auditor is 
Netherland Sewell.  In 2013 they produced 49,646 boe/d (+33% YoY) and had 278.5 mmboe of proved reserves (+39% YoY, 
32% proved developed), for a reserve life index of 15.4 years.  They had 249 net producing wells at 2013YE.
 

The company was Calpine’s former E&P segment.  They were spun out in 2005 on the heels of Calpine’s bankruptcy.  
Not surprisingly, their production mix used to be greater than 90% natural gas, which is the fuel consumed by the vast 
majority of Calpine’s electric generating plants.  The idea behind vertically integrating gas production with electric generation 
was to remove some of the volatility associated with one of the main cost inputs to the merchant electricity business.

That didn’t work out, and Rosetta found itself as a standalone company.  At the time of the separation they had gas-focused
assets in a variety of plays.  Over time they have divested almost all of those legacy assets.  A few years after the spin they
began leasing up acreage in the Eagle Ford.  Many of their executives are Burlington Resources alumni (that company was
sold to Conoco in 2006), and Burlington was known as the father of the Eagle Ford.  In fact, Rosetta’s CEO, CFO, COO and T
reasurer used to be at Burlington. 

Their Eagle Ford acreage was a grand slam.  They were early to the play so they were able to sign leases at great prices,
and they have gotten some of the best results in the entire play.  Their acreage is mostly in the condensate window.  (More
on condensate later).  

Obviously, you would hope for your acreage to produce as much black oil as possible (as opposed to the higher NGL and
gas mix you get in the condensate window), but Rosetta’s Eagle Ford wells have the highest average expected ultimate
recovery volumes (EURs, or lifetime production volumes) of any company in the play.  The higher gas production actually
helps them in a way because they can occasionally run gas back into the wells – it adds lifting pressure at minimal cost.

From 2009-12 when the company was basically a pure-play Eagle Ford producer they grew debt-adjusted reserves and 
production per share at CAGRs of 64.8% and 27.2%.   All sources finding costs averaged $6.99 per boe over that period, 
they had an average recycle ratio of 3.9x, and their reserve replacement rate (including revisions) averaged 642%.  Those 
are truly spectacular numbers.

They now have 62,772 net acres in the Eagle Ford.  26,236 of those acres are in a 9x10 mile section called the Gates Ranch
that could reasonably be characterized as a trophy asset.  F&D costs in the Gates Ranch are just $4 per boe. 

Rosetta remained a pure-play Eagle Ford producer until Mar-2013 when they reached a deal to buy 53,300 net acres in the
Permian Basin from Comstock.  40,000 of those acres are in Reeves County in the Delaware Basin of the Permian, and are
the company’s main focus in that play for now.  The other 13,000 acres are in Gaines County in the Midland Basin of the
Permian, where the company is going to drill just enough vertical wells to hold the acreage.  In Texas you only hold the
mineral rights down to the deepest zone you drill, so you may as well drill vertically in order to maximize the number of
zones you hold in case other zones unexpectedly pay off.  This is important in the Permian Basin since multiple horizontal
zones could be prospective.

Importantly, the Comstock deal was negotiated under an exclusivity agreement, and Comstock needed liquidity because
they were a gas-weighted producer (85.6% of 2012 production volumes) with a high cost structure at a time of weak gas
prices.  They were also highly levered (58.7% debt/capital; total debt of $2.99 per mcfe of PDP reserves). 

Rosetta has been taking their time delineating the Reeves County acreage and probably won’t go into development mode
until 2015-16.  I find that many E&P investors have short attention spans.  They seem to be frustrated with a lack of Permian
well results from Rosetta, but being deliberate is the right long-term approach.  Besides, there are several offset operators
including Concho and Energen that have been releasing attractive well results in the immediate vicinity and validating the play.

At this point the company is just in two basins and will probably remain that way for a while.  They have recently done 
bolt-on acquisitions in both basins and would do more if the right opportunities arose.

Eagle Ford:

• 95.9% of 2013 production, 90.9% of 2013 reserves
• 62,772 yet acres, 33.0% developed
• 100% working interest in almost all of the acreage
• 62% liquids cut in 2013
• 179 completed wells and 646 remaining locations (292 in the Gates Ranch), for 7 years of remaining drilling inventory

Permian:

• 3.7% of 2013 production, 9.1% of 2013 reserves
• 55,023 net acres, 25.9% developed (with vertical wells)
• 62.3% working interest by acreage
• 86% liquids cut in 2013
• 510 remaining vertical locations in Reeves County (again, they have to drill some vertical wells to hold all the horizontal zones)
• 446 remaining horizontal locations in Reeves County

Remaining consolidated drilling inventory is 17-18 years.

Their credit metrics are temporarily elevated due to the Permian acquisition.  Total debt/EBITDA is 1.7x and total debt/capital
is 52.7%.  Total debt per mcfe of PDP reserves is $2.77.  I should note that S&P upgraded them in December from B+ to BB-.

Why the stock is mispriced
As I mentioned earlier, E&P investors seem to have short attention spans.  If a company isn’t showering them with catalysts they 
begin to lose interest.  And, if there’s even a hint of trouble they often shoot first and ask questions later.  Fast money types dominate 
the space.  It doesn’t take much for them to turn on a company they loved only weeks before.  That’s what has created the
opportunity with Rosetta. 
 

• Production misses:  The company missed production estimates in Q2’13 and Q3’13 and lowered production
  guidance for Q4.  In my mind there are two types of production misses; those that indicate something is wrong,
  and those that don’t.  I believe Rosetta’s misses were the latter.
  
  Q2’13 production of 48,836 boe/d was short of consensus estimates by ~2,000 boe/d.  The reason for the shortfall
  was because they changed the development pattern of the wells in their Briscoe Ranch acreage in the Eagle Ford. 
  They moved from drilling 3 well pads to 8 well pads, and in the process changed the pattern of how they drilled the
  wells in such a way that would minimize the amount of production that would be interrupted in the future as they
  completed additional offset wells.  The trade off was they had to build up an extra inventory of drilled uncompleted
  wells on a one-time basis, which had the effect of temporarily dampening production growth.  It was short-term pain
  for long-term gain.  Nevertheless, the company reaffirmed full year production guidance, believing they would make
  it up on the back end.  The stock fell 2.9% the day after they reported.
  
  Q3’13 production of 50,935 boe/d was short of consensus by ~3,200 boe/d, and their Q4’13 production guidance
  was ~8% below Street expectations.  The stock fell 9.4% the day after they reported.  The reason for the shortfall
  was mainly due to problems with pre-existing vertical wells on the Permian acreage they acquired from Comstock. 
  
  Rosetta took over the operations in Aug-01-2013.  There were something like 75-100 vertical wells already on the
  acreage that had been drilled somewhat recently.  Comstock was in a hurry to drill the wells in order to hold the acreage. 
  They were trying to drill quick and cheap and that led to some sloppy well construction.  The wells were producing
  using rod lift pumping.  If you don’t drill the vertical wellbores straight enough the rods will grind against the tubing
  and eventually wear it down.  Something like half of the wells either had holes in the tubing, or the 25-foot lengths of
  rod hanging from the pumps had couplings between the lengths that were wearing on either the tubing or the rod coupling. 
  Also, slightly more than half the wells had rod boxes that were worn to the point where the rods had parted.  That
  was clearly not Rosetta’s fault and these issues are behind them now. They incurred extra workover expense during the
  quarter to fix the situation, and they probably won’t have those issues again because they installed rod guides on those
  75-100 wells.
  
  The bad news continued:  In their 2014 capex press release on Dec-18-2013 they disclosed that they lost ~4,000
  boe/d of production during Q4’13 due to a variety of reasons.  There was a compressor fire at a third-party production
  facility in the Eagle Ford, there were some “operational issues” at a third party gathering pipeline and gas processing facility,
  and the company chose to shut-in some of their wells next to offset operators who were in the process of completing
  wells nearby.  The stock fell by 5.8% the next day. 
  
  Again, not their fault, and temporary in nature, except having to shut-in producing wells when offset operators are
  completing nearby wells.  They will continue to occasionally do that.  However, the company has a big enough queue
  of drilled uncompleted wells at this point that meeting production targets should be less vulnerable to that dynamic than
  in the past.
   

• Mistaken read-through from SM Energy’s type curve reductions:  SM Energy has acreage in the Eagle Ford that forms
  a horseshoe around Rosetta’s Gates Ranch acreage.  When SM reported Q4’13 earnings on Feb-18-2014 they lowered
  the type curves for some sections of their acreage.  Their stock dropped 17.1% the next day, and Rosetta’s was down
  as much as 7.2% intraday because people mistakenly believed there was a read through. 

  First of all, the two companies’ acreage positions may be close to each other geographically, but that’s about as far as
  the comparison goes.  It’s not uncommon in the Eagle Ford for the rock’s characteristics to change dramatically if you
  move just a few miles in one direction or the other.  To cite an extreme case, up in DeWitt County if you move just a
  few miles the value of the acreage can go from $100,000 per acre to $1,000 per acre. 

  Second, one of the reasons behind SM’s type curve reductions was they hadn’t drilled in all the areas of their acreage,
  yet they were making assumptions about the      characteristics of the rock over entire areas of their acreage based on
  well results in certain areas.  In contrast, Rosetta has drilled in every corner of the Gates Ranch, so they know what
  they have.  They aren’t extrapolating like SM was.

  Third, in a textbook case of market inefficiency, investors assumed that since SM Energy and Rosetta both have acreage
  positions in the Eagle Ford called “Briscoe Ranch”, and that since Briscoe was the area where SM had the most consequential
  changes, that Rosetta’s Briscoe Ranch type curves were also at risk.  After all, if both positions have the same name they
  must have the same rocks, right? 

  Wrong.  Many oil leaseholds are named after the families that own the land.  In some cases a family owns so much
  land that they end up leasing drilling rights to multiple operators.  However, it’s rare that they would own so much land
  that two operators on their land would be very far apart, but that is the case with Briscoe Ranch.  Dolph Briscoe was
  governor of Texas from 1973-79.  He was the largest individual landowner in Texas.  He and his family have about
  640,000 acres, or nearly 1,000 square miles.  Understandably then, two people can be standing on land owned by
  Briscoe but also be very far apart.

  Rosetta and SM Energy both leased acreage from the Briscoe family but Rosetta’s Briscoe Ranch acreage is in Dimmit
  County and SM’s Briscoe Ranch acreage is in Webb County.  They are roughly 20 miles away from each other.  Again,
  you don’t have to move very far in the Eagle Ford to see a big change in the economics of the acreage.  Rosetta’s Briscoe
  Ranch type curves are fine.

• Overreaction to negative reserve revision:  Rosetta announced on Jan-30-2014 that they took a (10.6) mmboe reserve
  revision, or ~5% of their Jan-01-2013 reserves, due to lower assumed condensate yields in the northern section of their
  Gates Ranch acreage.  By this point the Street was ready to make a mountain out of a molehill.  The stock fell 3.0% the
  next day and 4.7% the day after.
  
  Over half the revision had to do with changing how the company transports condensate (again, more on that later) off
  the lease.  They used to truck it, but that method is more expensive than pipeline.  They finally got the opportunity to
  start shipping over pipeline.   However, the carbon dioxide content as it comes out of the earth is too high for the pipeline’s
  specs, so Rosetta has to stabilize it by removing some of the CO2 content.  They lose some of the hydrocarbons in the
  process.  The net effect is they produce fewer hydrocarbons but they ship it to market cheaper so the company is better
  off economically.

Valuation
I prefer to value E&P companies from the bottoms up on an NAV basis rather than using multiples.  Fortunately, Rosetta’s acreage
portfolio is simple enough and consistent enough to allow a granular NAV approach that doesn’t take a whole month to model. 
You can get most of the assumptions you need from their Feb-2014 Investor Presentation.  The model I’m using isn’t quite
simple enough to include in this writeup, so I will share the highlights and I will be glad to answer any of your questions in the
discussion thread.

I use $90 WTI / $4.25 Nymex as my base commodity price deck for the following reasons: 

• Bernstein estimates $94/bbl is the fully loaded cost of the world’s marginal barrel
• OPEC needs $80-100 oil for most of their governments to achieve balanced budgets, including Saudi Arabia at $80

I think it’s legitimate to be concerned about the demand side.  What if China implodes?  Core Labs estimates the world’s
natural decline rate is 2.5 bbdl/d even after maintenance capex, so the world could lose almost 3% of its daily demand of
86 mmbl/d without supply/coming unmatched.  The decline rates on shale wells are so huge that even a modest pullback
in domestic drilling activity would bring supply and demand back into alignment within just one year, which is a new dynamic
in the oil industry.  Moreover, spot prices are $99 WTI/$4.47 Nymex, so a new investment in ROSE based on $90/$4.25 would
begin its life with some margin of safety already built in.

The reasoning behind using a $4.25 gas price isn’t 100% scientific.   I run a standard combination of oil prices through the
E&P companies I look at, and I use gas prices that are roughly 1/20 the level of the oil prices.  I also model NGLs at 40%
of the oil price.  Don’t get too hung up on the gas price.  Over 80% of the company’s revenue last year was from liquids.
 

• Prospectivity: I am assuming 100% prospectivity for the Gates Ranch, 75% for the rest of the Eagle Ford, 40% for
  Permian vertical wells, and 25% for the horizontal Wolfcamp 

• Spacing:  I am assuming 55 acre spacing for the Gates Ranch 

Upside Opportunity #1 – Gates Ranch Downspacing
When the Eagle Ford first got started people were using 120 and 80 acre spacing.  Now some of the producers are down to
40 acre spacing.  I have been modeling the Gates Ranch on 55 acre spacing.  The company hasn’t been seeing interference
at this density, but they are determined to increase the densities until they get the wells to talk.  In fact, they are now testing
25 acre spacing on a section of the Gates Ranch.  They believe that due to the extreme tightness of the rock the number one
factor driving oil recovery is the stimulated rock volume in place.  The evidence so far supports that assertion.  The oldest
Gates Ranch wells are >4 years old so the data is there.  Even if they get down to a spacing density that causes the wells to
talk, the wells in the Gates Ranch pay out in less than one year, so when interference does begin to occur it will likely be at a
time in the wells’ lives when it will have only a modest impact on the NPV. 

Immediately going from 55 acre spacing to 40 acre spacing would add $8 per share.
Immediately going from 55 acre spacing to 25 acre spacing would add $18 per share.

Upside Opportunity #2 – Permian “Basket”
There are a number of things the company can do to drive upside on their Permian acreage like improving their completion
techniques, driving down well costs, and pooling acreage with offset operators to achieve longer laterals.

Driving better completion techniques and lowering well costs is pretty standard stuff across the industry.  It doesn’t have
anything to do with the Permian specifically except to say that we’re in the early innings of cracking the code in this particular
basin, so it’s almost inevitable that Rosetta will improve their EURs, lower their well costs, and succeed at tighter downspacing
on the standard 5,000 foot lateral lengths that people are modeling.  They appear to be making good progress on that front
already.  They have been guiding to $8.5 million horizontal well costs, but their most recent Hz well cost $7.6 million. 
Finally, they will eventually move to pad drilling in 2015-16 and that should provide further efficiencies.

Beyond that, there are unique features to this play that could deliver significant upside.
 

• Multi-zone potential:  Normally in a shale play a producer will be targeting one horizontal layer.  The exciting thing
  about the Permian is the potential for multiple zones to be economic.  In Reeves County the initial targeted zone
  is the Wolfcamp A, but the Wolfcamp B & C could also turn out to be prospective, and the company will also test the
  Lower Bone Springs. 
   

• Cooperation with offset operators:  It’s difficult to find contiguous blocks of acreage in the Permian Basin.  For starters,
  mineral rights in the basin are very fragmented due to years and years of prospecting.  Moreover, in the 19^th century
  the federal government carved up ownership of the land that ran parallel to railroad tracks in order to prevent monopolists
  from controlling access to the rail lines.  They only allowed you to buy acreage blocks in a checkerboard pattern. 
  In other words, you could only get control of every other block.  Imagine if the land looked like a checkerboard and
  you were only allowed to buy the black squares.
  
  Having contiguous blocks of acreage is advantageous to oil producers because it allows you to drill longer horizontal laterals.  
  Each block is roughly 5,000 feet long.  If you don’t have contiguous blocks you can’t drill laterals longer than ~5,000 feet, and 
  you don’t get the economic leverage of extending your lateral out to, say, 10,000 feet.  There is fixed cost leverage with extending 
  the laterals; you’ve already spent the money to drill the 7,500 – 11,500 foot vertical leg of the well and you’re already incurring 
  the spread costs of stationing all your vendors at the well site, so when you extend the lateral you get a big uplift in production 
  volumes but it only cost you a few extra dollars to drill the extra horizontal footage.
  
  Concho and Rosetta have a bunch of neighboring acreage in Reeves County.  They also hold working interests in a lot of each 
  other’s acreage, and the operatorship is also split between the companies to varying degrees.  
  
  The Street is modeling Rosetta’s Reeves County acreage with 5,000 foot laterals.  It’s almost inevitable that Rosetta and 
  Concho will figure out how to pool some of their working interest and divide up the operatorships in such a way that will 
  allow them to drill much 7,500 or 10,000 foot laterals. 
  
  Occidental, Clayton Williams and Energen also have acreage nearby. 

Upside Opportunity #3 – Upper Eagle Ford new venture
Rosetta is testing an additional zone in their Eagle Ford acreage known as the Upper Eagle Ford (UEF).  If commercially viable,
it could add tremendous upside to the company’s value. 

The Eagle Ford actually consists of two zones – Upper and Lower.  The section that the play is known for is the Lower Eagle
Ford (LEF).  As you move from the Southwest portion of the Eagle Ford to the Northeast the thickness narrows.  The Southwest
section where Gates Ranch is has a total thickness of 250-350 feet, with the LEF comprising 50-75 feet and the UEF taking up 200-275 feet. 

By way of comparison, up in Gonzales County the Eagle Ford’s total thickness is just 90-110 feet with the LEF comprising 30-40
feet of it and the UEF taking up 60-70 feet.

Rosetta is running four separate well tests in order to determine the prospectivity of the UEF.  Three of the tests are on the
Gates Ranch and one is on their smaller Lasseter & Epright (L&E) lease.

Rosetta hasn’t released well results yet.  Evaluating the rocks in advance of their future disclosures is beyond my circle of
competence but we can read the tealeaves.  What I see looks pretty good.

The first pilot in the Gates Ranch has been online since May-2013, so the company has plenty of data about how those initial
UEF wells are performing.  Given that first year decline rates for shale wells are so huge, the data you get in the first year tells
you almost everything you need to know in order to evaluate the economics.  In other words, the company should have plenty
of data from the first pilot to have an idea of whether drilling more UEF wells is a waste of money.  Plus, the company started
another two pilots that they’ve obviously been able to gather additional data from. 

These guys are methodical operators.   Their philosophy is ready, aim, fire; not ready, fire, aim.

What we know at this point is they have well results from three UEF pilots, we know they’re not the type of people to screw
around with the shareholders’ money, and now they’ve chosen to move forward with a fourth pilot that will be their biggest yet,
consisting of 11 wells – 5 in the UEF and 6 in the LEF.  At an average well cost of >$6 million, this implies they’re spending over
$60 million to do the fourth test.  They are not going to spend $60 million on something that has dry hole risk.  I would say the
fact pattern looks favorable. 

One factor that will help the economics of UEF wells is the company expects to use sand proppant in their completions, whereas
they have been almost exclusively using ceramic proppant in their LEF Gates Ranch completions.  Using ceramic in the Gates Ranch
adds $1 million to LEF well costs, which are currently running $6.5 million.

I have been valuing the Gates Ranch at $2.5 billion ($41 per share) using $90 WTI/$4.25 Nymex in my development model. 
If the UEF is worth even half that it’ll be a giant home run.

The company expects to release its first UEF well results on the Q1’14 call.

Let’s talk about condensate!
The bear case on Rosetta revolves around the company taking big haircuts on its condensate realizations.

The company defines condensate as oil that has API (American Petroleum Institute) gravity above 55 degrees.  The
generally accepted line between crude oil and condensate seems to be 45 degrees, but there is no official standard.  Either
way, it’s very light oil – so light in fact that back in the day people would sometimes put condensate directly in their gas tank. 
The Eagle Ford in general produces light oil, often above 45 degrees. 

For comparison, conventional light sweet U.S. crude like WTI has an API gravity of 39 and heavy crudes like Mexican
Maya have an API gravity of 20.

Condensate (as Rosetta defines it) accounted for 14.3% of the company’s 2013 production and 15.5% of their net proved reserves.

The concern is that Gulf Coast refiners aren’t geared up to handle lots of light sweet, and as Eagle Ford condensate production
volumes rise there’s a risk that Rosetta may have to take steeper discounts on their condensate.  The importance of this risk is
emphasized by the fact that Rosetta’s 2013 annual report was their first one to break out the portion of their reserves and
production coming from condensate.

For many years the U.S. produced 400,000 – 600,000 barrels of condensate per day.  By 2013 the nation’s condensate
production was over 1 million bbl/d, with the Eagle Ford accounting for over ½ of that.  For the time being increased domestic
condensate production is simply displacing the ~1 million bbl/d of light sweet crude we still import.  But domestic condensate
production could eventually displace all of the imports and hit the “refining wall”. 

Ordinarily, the economic solution would be for producers to seek international export markets where pricing is stronger.  
However, crude oil exports are prohibited.  They were banned by the Energy Policy and Conservation Act of 1975, which 
defines crude as any hydrocarbon that “remains liquid at atmospheric pressure after passing through surface separating 
facilities and which has not been processed through a crude oil distillation tower.”  Since unrefined condensate from field 
separation facilities (or “lease condensate”) is specifically defined by the regulations as crude oil, it is also subject to export 
restrictions.  

Therefore, growing domestic oil and condensate production will be unable to seek higher Brent prices in global markets,
meaning it will likely continue to trade at a discount to Brent.  The lower utility of condensate relative to black oil means
it should be priced at a further discount to domestic benchmarks like LLS.

There is a limited solution in the works for condensate.  Kinder Morgan is building a condensate splitter near its Galena Park
terminal on the Houston Ship Channel.  The $370 million project will have two units with total capacity of 100,000 bpd. 
The splitter facility will link to the Kinder Morgan Crude and Condensate Pipeline which transports product from the Eagle Ford
to Houston.  The first unit is expected to go into service in Q1’14 and the second unit is expected in Q2’15. 

The nice thing about this kind of facility is it doesn’t take very long to construct and it’s far cheaper than refineries
($3,700 per barrel of daily capacity vs. $30,000 for a refinery).  Moreover, companies like Valero and Magellan Midstream
Partners have discussed plans for a further 12 such plants with total condensate capacity of 460,000 barrels per day.

If all those projects are built it might be enough to absorb the Eagle Ford’s incremental condensate production.  The Eagle
Ford’s total energy-equivalent production is currently running over 1 million barrels per day, and is expected to increase by
another 700,000 bbl/d by 2017.  Even if half that incremental production, or 350,000 bbl/d, is condensate then Rosetta
would be okay because the >500,000 bbl/d of projects could absorb it.

Condensate pricing is running $13-14/bbl below LLS, or $89-90/bbl.  (LLS is at $104/bbl).Recall that my $75 WTI/$3.50
Nymex valuation puts the value of ROSE at $42 per share.  That gives condensate prices plenty of room to fall from
current levels before creating material downside to the current stock price.  I think this issue is largely priced in. 

It’s hard to find good condensate data.  One place you can check prices is here: 

http://www.fhr.com/%28X%281%29S%28bnhaye55455ht255ag4tueeq%29%29/refining/bulletins.aspx?AspxAutoDetectCookieSupport=1

Catalyst