I believe I have a contrarian idea which may not smell like value at first, but is quite attractive once you drill a little deeper, natural gas (puns intended).

The idea: Long SWN. Sameplot posted a short on SWN yesterday, so I am taking the opposite side. Let’s hope the discussion is as illuminating and entertaining as VITC.

In my view, one way to value a commodity resource producer is along three salient axes: 1) the quality of the assets, 2) the financial position and quality of management, and 3) the “normal” price of the commodity.

The quality of the assets
Southwestern has an interesting history. In 2004, the company was a small E&P company with only $471 in revenue and 54 BCF in production. In August 2004, the company announced that it was aggressively developing a new shale play, the Fayetteville, where it held around 500K net acres which it had purchased for de minimis cost. Fayetteville production was 12, 54, 135, 244, 350, 437 BCF in 2006, 2007, 2008, 2009, 2010, and 2011, so the play became a homerun for owners.

Today, Southwestern is a maturing company, whose assets can be categorized in three buckets (similar to Sameplot’s analysis): 1) the Fayetteville acreage, 2) Marcellus acreage, and 3) venture portfolio (Brown Dense, etc.).

Fayetteville is the majority of the value, so is worthy of detailed discussion. Southwestern is the dominant (i.e. 50%) producer in the Fayetteville and owns its gathering capacity ($250 a year in operating earnings). The company has a basi presentation which has the below and further information on its website. Net acres in the Fayetteville are 950,000 and drilling depths are shallow to a mile deep with 50-500 feet of pay. Management says it would take 8 rigs ($100MM a rig) running to maintain Fayetteville production at 450 BCF. There are 8,300 undrilled net locations, implying 16 years of inventory at 500 wells/yr. The Fayetteville has a 15-20 cent negative basis. Cost per MCF is $2.50, including $1.30 in DD&A (there was a $908 write-off for full-cost accounting treatment in March 2009, which initially lowered DD&A by 50 cents, but should be largely worked through).

The Fayetteville accounts for ~3% of US natural gas output. According to Credit Suisse (Feb. 3, 2012), the breakeven price of NYMEX with a 15% after-tax return is $4.58, which places it importantly similar to the Barnett and the Haynesville. Assuming in a few years, associated gas (i.e. gas produced as a byproduct of oil and NGL production) is 15-20% of domestic dry gas production and low-cost shale gas is 50% (together around 70%), with coal-bed methane, offshore and conventional onshore being the higher-cost 30%, then I would guess that the Fayetteville is squarely in the middle of the cost curve, only materially disadvantaged versus associated gas and Marcellus. So as far as natural gas acreage goes, this is a quality, predictable asset.

In February 2011, BHP bough 487,000 acres from Chesapeake for $9,750 an acre, a similar price would generate $9B, or $22 a share after subtracting net debt but before deferred tax.

SWN also holds 187,000 net acres in the Marcellus, in the dry northeast section. From the actions of SWN and similar E&P companies, this appears to be among the most attractive dry gas acreage in the country; SWN is shifting its development from Fayetteville to Marcellus for economic reasons and to hold leases. SWN purchased its acreage position in the Marcellus for ~$1K per acre on average, whereas acreage changes hands today at significant multiples of this price, perhaps $10K. Management says that production will grow from 0.14 BCF/d now to 0.3 BCF/d by end of 2012 and 0.49 BCF/d in 2013. Due to its attractive basis and lower costs, it is safe to assume that the Marcellus acreage is better than the Fayetteville.

Lastly, one has the venture portfolio of Brown Dense (an oil play, 520K acres), Colorado (238K acres), and New Brunswick. Sameplot is correct that analysts were disappointed with the initial well results of the first well in the Brown Dense. One wonders how this will work out.

The net of this is that the Marcellus position “upgrades” the overall quality of the company’s assets to somewhere above average, versus the average US domestic gas producer, and there is the potential to be pleasently surprised with the venture portfolio.

Management has provided financial guidance of 560-570 BCF in production in 2012 (13% growth) and $1.20-1.23 in EPS, assuming $2.75 gas and $90 oil. I believe we can expect at least this level of growth in 2013-2014 given my projections of cash flows, decline curves, and a gas price above $3.

I note that the SWN 10-K is among the most readable of any E&P company – I recommend it to you. It is also worth noting that SWN’s development program holds all of the leases that it wants to keep and are expiring in 2012-2013, so this does not appear to be a problem.

Financial position/quality of management
SWN is conservatively financed and management says the right things in terms of capital allocation. Given the uncertainty and the volatility of the commodity SWN sells, I view this as a prerequisite to any interest in the common stock.

Net properties on an accounting basis are $6.6B, funded with $1.34B in net debt ($672 due 2016, the earliest maturity), $1.6B with deferred tax, and $4.0B in hard book value. 50% of 2012 production and 30% of 2013 production is hedged above $5.00. This is a very strong financial position. Write-downs of the accounting carrying value of assets are coming in 2012, which do not disturb me.

“Normal” price of natural gas
This is the crux of the thesis. One can think of SWN as a bond with a growing coupon of natural gas; one has to have a view on what this coupon is worth in terms of dollars.

Everyone who reads the business press knows the following: 1) there is a lot of natural gas out there, 2) there is a lot of gas that will be produced as a byproduct of producing oil and NGLs, 3) there is a lot of gas in storage, 4) the spot price of natural gas is at a ten-year low and the futures’ curve of gas has declined ~25% from prices last year, and 5) US natural gas will be cheap indefinitely. In my judgment, the first four statements are true, and the fifth is over shooting the fundamentals.

Let’s begin with some of my basic assertions. First, the advent of horizontal drilling and fracturing has changed the domestic natural gas industry. It really is “different this time.” There is a slug of low-cost, predictable, close-to-market capacity which is available to supply natural gas demand – at a price. In addition, associated gas (gas produced with oil) is also going to grow, and this will grow at any price (assuming a high oil price). Second, domestic natural gas production has a significant decline curve and this decline curve is increasing, somewhere between 25-30%. Most drilling activity is required to maintain last year’s production. Shale production declines faster than conventional wells, so the overall US domestic decline curve should increase over time with this mix change. But there is a secondary mix shift underway from old wells to young wells (because young shale wells are so prolific, they represent a one-time higher than normal share of overall production). This secondary effect means the US domestic decline curve will “peak” above its long-term, mix-adjusted average and then settle over time. I estimate the decline curve is 27% in 2012, but I have seen bottom-up estimates as high as 32%. The supply of natural gas – at a price – corrects faster than the past due to this high decline curve. Third, there is a consumer of natural gas – at a price – that has an insatiable appetite, natural gas fired power generation. As Citigroup and Bernstein research and the Platt’s weekly estimates thus far in 2012 indicate, gas-fired power may be up 20%+ year-over-year in 2012. Fourth, the natural gas market does not have a lot of slack and is very hard to call, with a lot of surprises and unknowns. That is why speculators, traders, and the managements in the industry, all of whom “play the game” full-time, often get it wrong, even over the shorter-term of a year or two.

Enough with the qualitative, let’s talk numbers. Why is the spot price for natural gas so low? Natural gas is cheap today primarily because the US had a warm winter. There is currently 2.38T of natural gas in storage, according to the EIA, compared to 1.28T at a similar time in 2008 (a year when natural gas averaged $8.89 per MCF), a 1.1T difference. Based on a 250BCF difference year-over-year in residential/commercial consumption in November/December (using EIA numbers) and 580BCF lower consumption by residential/commercial in January, February and March, using Platt’s numbers, one can explain 0.8T of the excess gas in storage as a thought exercise.

Where is the price of natural gas going to go in 2012? I don’t know. But because everyone is so negative on the price, let me offer some countervailing thoughts.

In 2011, I estimate that the decline curve was 27%, or 15.8 BCF/d on 2010 production of 58.4 BCF/d. 930 average working natural gas rigs (on average, 6-month lag, Baker Hughes numbers) generated an increase in 4.6 BCF/d, so that 2011 dry gas production was 63.0 BCF/d. So 930 working rigs generated 20.4 BCF/d in new production, or 2.2 BCF/d for 100 rigs. Assuming 2011 production of 63.0 BCF/d declines at 27%, and ignoring efficiency gains and associated gas, one would expect 780 rigs to be required to hold production flat in 2012. Current natural gas rigs are 652 and declining rapidly. Associated gas will certainly add incremental supply (perhaps 10-15% of dry gas production in total will be associated gas), but also remember that wet gas drilling generates less dry gas; an Eagle Ford well is only 2/3 dry gas. So the net of this is incremental low-cost supply but the amount is uncertain. Efficiency of gas produced per working rig will increase with marginal rigs dropping, more pad drilling, and longer laterals, but efficiency is setback with drilling to hold leases, which is inefficient. And shut-ins are happening and may continue, if you believe Chesapeake and EnCana announcements. I foresee dry gas production peaking and then declining as we move through 2012, with perhaps 66BCF/d for 2012 production and a steep decline at the end of the year.

Demand may also offer some surprises. Weather I have already discussed; I estimate the warm winter explains 0.8T of excess storage. Bernstein’s analyst (March 23, 2012) has numbers as high as 9 BCF/d in increased consumption from gas-fired power in 2012 contingent on gas’ advantage over coal, compared to 21 BCF/d in total in 2011. This number may be too high and is contingent on a continued very low spot price for gas, but it is very large. Let’s assume that power increases instead by 5 BCF/d, or 1.8T. This would consume all of the excess storage from the warm winter and then some. While I recommend the research, which is a bottoms-up look at how much coal consumption can be displaced, and do not care to repeat it, I note one factor that the analysts failed to mention. Even regulated “cost plus” utilities have a higher than normal incentive to reduce variable costs in 2012 by burning gas. New generating capacity additions are going to be much more expensive than historically (due to environmental regulations) and rate commissions are reluctant to allow significant rate increases; regulated utilities can “pay for” capex investment by shifting cheap underutilized gas-fired capacity to baseload. When it comes to natural gas demand, forget CNG vehicles and LNG, the key swing factor for demand over the next three years is weather and power. And there may be no cure for a low commodity price like a low commodity price.

What is the “normal” price of natural gas? Let me begin by providing a rough estimate of where dry gas production comes from currently: 40% shales, 15% associated gas (together 55% low cost), coal bed methane 10%, offshore 5%, and conventional onshore 30% (together 45% high cost). Every sell-side analyst break-even analysis and E&P management presentation that I have seen has the break-even cost of producing “high cost” gas well above $5 per MCF. Is it possible that the 17 BCF/d in annual production which is required to be replaced every year comes from only “low cost” sources, so that this incremental supply permanently suppresses prices? Let’s assume that associated gas is growing from 10% to 20% of production (which is very aggressive), or 6.3 BCF/d in incremental, that will get us part of the way there. Shale dry gas production has grown from 10 BCF/d in the beginning of 2010 to near 25 BCF/d currently. But around 17 BCF/d of this 25 BCF/d comes from Barnett (5), Fayettevile (2.5), Haynesville (5.9), and Marcellus (3.6). Citigroup (March 22, 2012) has done a bottoms-up estimate of field production growth and foresees both the Barnett and Haynesville declining and the Fayetteville growing only 8%. Likewise, from the Citigroup report, we can estimate that there are less than 200 rigs working in the four key shale plays; those 450 active gas rigs are working somewhere else on higher-cost acreage. When I play with the numbers, I have to suspend the normal laws of supply and demand in order to maintain a long-term “normal” price below $5 an MCF. And the average price since 2000? $5.50, which ignores the 2-3% in annual inflation. I’m willing to bet that the natural gas spot price today is materially too low and temporary.

What is SWN worth?
Since you are not going to buy the common stock unless you want to own a producer of the natural gas commodity, I am going to keep this simple. In 2014, I foresee SWN producing 760 BCF of natural gas. Assuming a minimum $5/MCF price, $1.60/MCF in profits, ratable increases in midstream profits, I generate an EPS of $2.75. Capitalized at 15x (above average financial position, above average management, low-cost producer), this generates a floor valuation of $40 a share, which I consider a low and disappointing outcome.

What about the PV-10?
Sameplot uses the recently filed PV-10 and accounting reserves in the attempt to demonstrate that SWN is worth $15 a share. I do not put much weight in the PV-10s since the 2009 SEC change due to the subjectivity in their composition and the nature of the assets of shale gas producers.

Previously, with vertical wells and geologic uncertainty, the PV-10 provided insight into the amount of resource which one could depend on the company actually possessing. With SWN and its acreage position combined with horizontal drilling, the uncertainty is less geologic and more expenses, commodity prices, and the management’s development plan. The acreage is owned, the resource is there, but the manufacturing costs and the sales prices are unknown. Reserves and PV10 numbers can swing wildly depending on the optimism in management’s assumptions.

SWN claims to do a specific EVA-like analysis for all of its capital decisions, for which it has a 11%+ IRR hurdle and is using $3.00, $3.50, and $4.00 NYMEX prices for 2012, 2013 and thereafter, which is what goes into the PV10. I believe these numbers are too low. Sameplot is taking apparently conservative management estimates and marking them to a 10-year spot market low.

Instead, I would use private market valuations of acreage sales. Everything I have seen in this regard (sell-side analysts keep exhaustive lists) makes the $15 a share valuation appear highly unlikely, in my view.