Implications of a Shift Toward Natural Gas Power Generation

A nice piece of anecdotal color out today from Steve Mitnick (of Olivery Wyman via MW) on the transition of operators from Coal to Gas.

Highlights:

• There are more utilities with natural gas divisions that are providing tangible earnings streams when regulated utilities are not. (i.e., Exelon and Progress Energy recently set plans to shut down coal plants and open up NG plants)
• One challenge facing natural gas verses coal is price volatility, which NG producers and power companies are increasingly focused on
• Mitnick forecasts NG supplying 30% of U.S. power generation (Coal supplies about 45% of, while gas currently supplies about 23%)
• He foresees a tidal wave of pipeline development down the road and definitely storage will become increasingly important.

Implications:

• More hedging needs for natural gas from less price-sensitive commercial operators may likely create more opportunities for price-sensitive speculators
• Understanding how this will impact the volatility of the commodity may be a source of alpha generation going forward...

Higher Natural Gas Volatility to Come?

As I continue to work on my thesis, I have been spending an increasing amount of time thinking about the dynamics influencing the volatility of the natural gas market. With front-month contracts for natural gas down 29% since Dec-31 (see chart), historical volatility has also declined to multi-month lows, from a high of 131% (30D annualized HVOL) in Sep-09 to a current low of 30% (a 1.2 percentile event over the last year).

Accordingly, I think there is opportunity for notably higher NG volatility in the medium-term for the following fundamental reasons:

From a supply-side, conventional production (i.e., Barnett, etc.) is not going away anytime soon, and with a marginal cost for production higher relative to Shale gas ($5-8/Mcf vs. economical Shale gas at sub-$5/Mcf), it will help to establish a higher NG floor once demand comes back online. While shale gas is unarguably the future of our domestic natural gas supply story, supplying an expected ~29bcf/d, we are still likely going to be ~30+ Bcf/d short of demand in the longer-term, based on current U.S. NG consumption of ~60-80 Bcf/d (depending on season).

I think there is reason for more economical / lower marginal cost shale gas to not be in a place yet to replace conventional gas the way some sell-side analysts prosteletize (For instance, on May 27, 2009, one energy analyst noted, “Shale — along with other low-cost unconventional gas — could provide 75% to 90% of new gas supply over the next several years and set the marginal cost of new supply”). My line of thinking is as follows:

1. In the Marcellus shale, arguably one of the most economical shale plays (slightly ahead of the Haynesville shale), the pipeline infrastructure is not yet established in the region to provide adequate take-out capacity at a rate that is commensurate with current levels of production.
   
2. Current Appalachia production is ~2.5bcf/d, expected to grow to ~6 Bcf/d by 2013.
   
3. This is a lumpy process as new pipelines come online to match growing production.
4. However, pipeline capacity will get there in time: the major pipelines are still a few years away, with 6.9 Bcf/d of take-out capacity announced by Q413 (a bulk of which is back-ended).
5. Since the Appalachia area has heated up, companies have announced 6+ Bcf/d of new pipeline capacity - Key players in the Pipelines: EP, EQT, NFG, WMB, SE, D, NI).
   
6. In the meantime, I believe that this may contribute to higher volatility over time, particularly as greater drill-bit emphasis from operators is placed on unconventional / shale gas, while traditional (horizontal) wells dry out, reaching their EUR.
   

From a demand-side, I think there is a reasonable argument for NG demand coming back online late this year, perhaps sooner than may be anticipated, largely attributable to the substitution effects with Coal:

1. the pick-up in coal exports, which was 15.2 mm short tons in Q309, up from 13.0 mm short tons in Q209, implies that we are exporting ourselves into a tighter coal market. I believe the Q4 data released by the EIA later this month will give evidence to a continuation of that trend.
   
2. Increased economic competitiveness for power producers will compel a switch from Coal to NG.
3. Coal is still the lowest cost for energy production, but it is losing its economic competitiveness (see chart): For the U.S. as a whole, Coal energy production costs 19.8 $/MWh, down 11.6% YoY, whereas NG is 40.56 $/MWh, down 23.2% YoY, all as of Nov-09.
4. Coal prices are already up 9% from Q309, and I think this trend will continue if we get evidence of sustained larger coal exports to Canada and EM.

Bottom line:

• I agree that Shale gas is the new hope for the long-term, but in the near-term there are structural obstacles for it to fully replace traditional wells in the way that I think the sell-side is envisioning.
• Further, increased substitution from Coal into NG will help to bring demand back online later this year, likely in the energy consuming summer months.
• Once demand picks up, even slightly, I think we will see some violent volatility moves in the NG.
  
• Don't get me wrong, I am still bearish on the near-term outlook for NG prices - we are still materially over-supplied, but I think there are reasonable catalysts that can correct this dynamic in the medium term, which may not be given enough consideration.
  

The Secrets of Skewness

A question came up the other day regarding how to think about skewness in the volatility surface of equity options.

Thinking of volatility skew as approximated by the relative slope of the volatility surface for a given time to expiry, the skew can be a valuable indicator that shows option traders' biases towards a particular stock or index. Expensive options are a strong indicator of pending changes in a stock's price. The measure of expensiveness is the options implied volatility.

The ATM or vega-weighted implied volatility measure tells only whether the overall stock volatility is high, but doesn't yield any information about the markets sentiment of the stock's impending directional movement.

Under a Black-Scholes framework, options of the same maturity should in theory have the same implied volatility across strikes. However, in reality this is not the case. Supply-demand dynamics for individual option contracts can distort the price (and thus implied vol) at different points across the volatility surface's term structure and strike range.

The disparity of implied volatility between calls / puts indicates the position of traders in aggregate (volume alone does not imply net buying or selling). For instance, if OTM call ivol is higher than same strike/expiry put ivol, it may imply that traders favor an increase in the underlying.

To really understand the implications of volatility skew for impending price movements, one needs to analyze the volume of individual option contracts. Here is where volume can add color:

• High volume and low volatility may indicate that option contracts are being sold.
• High volume and high volatility may indicate that option contracts are being purchased.
  
  Additionally, it is important to consider Time Skew vs. Strike Skew:

• Time Skew: measures the spread between option ivol for contracts with same strike but different time to expiry.
• Strike skew: measure of spread in ivol for contracts with different strikes, but same time to expiry

Together, combining Implied Volatility and Volume, we can infer the following relationships:

• Higher ivols in NTM / ITM calls vs. puts may indicate a majority of call buyers, which can be viewed as bullish.
• An abundance of volume in the OTM call, coupled with low ivol, can indicate OTM call sellers. While this is a relatively neutral indicator of direction, it may suggest that traders aren't anticipating the stock rising beyond the OTM call's strike price.

So, what does this all mean in practice, what is the skew of the S&P 500 (SPX) telling us now, and are options "fairly" priced today?

While, except for a brief respite in mid-November, equity index options have not been this fairly valued in nearly a year (measuring IVOL/30D HVOL), meaning that realized volatility has more closely matched the volatility implied by options prices (see chart at right).

From a distribution standpoint, the ratio is relatively middle of the road, at a 44th-percentile relative to the past year (see histogram at right).


I constructed an equal-weighted index that looks at 80%, 90%, 95%, 105%, 110% and 120% strike options' implied volatility relative to the ATM implied volatility. The index captures the degree of slope, or skew, in the ivol term structure on any given day.

In looking at the past year's distribution of this index, we see that today the degree of skew in the SPX is incredibly high, or a 5th-percentile event (see charts at right). This high-degree of skew has persisted well over the past month.

However, SPX skew is the highest among global indices, despite extensive focus on the deteriorating macros of Europe and concerns over China. Furthermore, an important implication is that SPX correlation has nearly doubled, which can be seen to indicate the degree to which the market has focused more exclusively on the macro. This coupled with a flattening of the VIX term structure (at right) due to upward pressure on spot/near-term contracts may be indicative of market expectations for higher volatility due to an impending pull-back.

The sovereign macro risks seem to be inherently driving the market...

Superior Well Services (SWSI) - GARP?

Thesis: The bullish thesis for SWSI is predicated upon increased demand for their services related to the ongoing recovery from a capacity bubble within the pressure pumping market. Increased demand for SWSI's hydro fracing is due to greater per-well intensity in the Marcellus Shale, where growth is expected to be sustained even despite a lousy gas price environment. The market does not fully appreciate the magnitude of this dynamic's impact on SWSI's top-line, and still has some concern over the firm's financial leverage (The bulk of principal payments are not due until 2013 ($207.1mm))

Company Profile: $570mm Market Cap, $785mm Enterprise Value. SWSI provides well-site solutions to oil and natural gas companies. The company offers technical pumping services and down-hole surveying services. Its technical pumping services include stimulation services, such as fracturing and acidizing, which are designed to improve the flow of oil and natural gas from producing zones; It serves regional, independent oil and natural gas companies in the Appalachian, Mid-Continent, Rocky Mountain, and southeast and southwest regions of the United States. The company operates service centers in Pennsylvania, West Virginia, Oklahoma, Mississippi, Alabama, Michigan, Arkansas, Utah, Texas, Virginia, New Mexico, Louisiana, Ohio, Kansas, North Dakota, Colorado, and Wyoming. As of December 31, 2008, Superior Well Services, Inc. owned a fleet of 1,628 commercial vehicles and 116 logging and perforating trucks and cranes. The company was founded in 1997 and is based in Indiana, Pennsylvania.

The Story: Once near the brink of bankruptcy, SWSI stock has surged ~170% over last the 6 months to become one of the top performing oilfield service stocks in the sector. The company has a strong presence in the Marcellus Shale, providing over 30% of the horsepower to operators in the region (SWSI also has good exposure to plays in the Haynesville, Barnett and Bakken as well).

The Marcellus in particular possesses good economics and is expected to grow, even despite the current tepid natural gas price environment. SWSI’s large exposure to the Marcellus relative to its size makes the company an excellent pure-play to capitalize on the growth of that region.

Valuation: A 6.3-6.5x EV/EBITDA multiple on 2011 is a reasonable target, given SWSI's size and the valuation history of its comps (RES, BJS and KEG), implying a $20-20.6 per share price target.

Disclosure: I have been Long SWSI since Jan 4th, 2010 at an average fill of $15.10.

What is the Optimal Inflation Hedge?

Yesterday's FT noted the rising demand amongst investors for inflation-safe debt. The UK on Monday saw the second best tender, or mini-auction, for inflation-protected bonds since October 2008. Notably, inflation expectations have risen sharply in the U.S., particularly because of the credit-easing policies and record low nominal rates. Last week, there was strong investor demand for a $10bn 10-year US inflation-linked bond. Its bid-to-cover ratio was 2.65 (versus the average of 2.14 for the past five new sales).

Using the break-even rate, which is the difference between conventional U.S. treasury bond yields and inflation-linked bond yields, we can see that current inflation expectations are roughly in line with pre-Lehman levels. Worth note, since last Friday's disappointing employment report, 5Y and 20Y expectations tightened by nearly 10bps through today.

It is important to put this in context. Looking at a histogram of implied inflation expectation levels since Jan-08, we appear to be back near median forecast levels (histogram at right). It is also important to notice the modest correlation that these expectations have on the price level of both gold and oil (see regressions at right).


This begs a few very important questions for mitigating risk to our portfolios in the future; namely, what are the levels of inflation that we can anticipate in the future, when, and how best can we hedge ourselves for the long run?

I am a believer in Mauldin's view that we can expect to see a "statistical recovery", the statistics (i.e., GDP, etc.) are positive but it certainly doesn't "feel" like a recovery. First, year-over-year comparisons are looking better, since 2008 was horrific. Second, inventory levels are about as low as they will go. In the way GDP is figured, a reduction in inventory reduces GDP (perhaps GDI is an even better barometer...). That was a negative figure for most of this recession. Simply because inventories not falling any more, it is easier to get a positive GDP. Lastly, the onetime benefits for GDP from the fiscal stimulus are coming due (toughly 90% of the 2.2% growth in GDP in the third quarter was attributable to the stimulus), and we can already see the effects of federally-sponsored growth (auto sales fell in Dec-09 0.8% after a 1.2% Nov-09 gain, slumping in Sep-09 after the "cash-for-clunkers" plan expired).

I am always skeptical of outliers and when the momentum of the herd moves to quickly. Looking at how quickly inflation expectations have resumed to pre-crisis levels, despite a tremendously negative macro, gives me great pause in the bubble taking place in commodities like gold and oil. This has been a shift away from my more tactical-oriented momentum-based trading of options on the GLD in recent months. I am cautious with how quickly risky assets and inflation-sensitive commodities like gold have risen in recent months. Granted the past is not indicative of future performance, but if you look at past recessions, it is not all that unusual (8 out of 11 times) for there to be positive GDP quarters in the midst of an ongoing recession.

Getting Back to Business!

After a much needed respite from exams, it's time to get back to business.

Let's start the New Year by discussing updates to the portfolio over the past four weeks.

The Gold Trade:
It looks like the violent interruption in gold's strong trend during early December was likely just a reversion to the mean. It was during that correction that GLD options appeared to be reasonably priced relative to the past several months (see Chart #1, at right). However, it looks like the short-term will be characterized by turbulent trading, with a skew toward the upside (see Chart #2, at right).

The fundamental rationale for being long-gold (ideally not in USD-terms) still seems intact. As highlighted in today's Gartman Letter, the logic behind the problematic nature of the EUR as a reserve currency strengthens the notion of owning gold, not in US dollar terms but in terms of the EUR, Sterling and the Yen (particularly in light of recent comments made by Naoto Kan, the newly appointed Finance Minister in Japan). Gold is now being seen by more and more governments as reserve currency. More so, with a weakened EuroZone, which is increasingly threatened by the weighing divergence from member nations' (i.e., the PIIGs), it would seem reasonable that capital, which may have been earmarked for euro-denominated assets, on the margin, may more likely be allocated to gold denominated in euro terms.

Energy - Fundamental Book: XTO Bought up by XOM
On Dec-14, ExxonMobil (XOM) announced that it had reached an agreement to buy XTO Energy in an all-stock transaction valued at $41 billion, including approximately $10 billion of XTO debt. The deal implies a ~25% premium for XTO shareholders based on Dec-11 prices.

I first recommended getting long XTO's equity and neutral on CDS in early Mar-09. The thesis was that XTO was more highly levered than peers, and the market was uncertain that they could successfully digest their mammoth acquisitions of 2008 and develop the acquired resources successfully.

However, from a technicals standpoint, XTO initially underperformed relative to key peer Anadarko (APC) by nearly 22% over the subsequent 5 weeks due to crude's nearly 25% rally. After some humbling analysis, I discovered the reason was related to XTO's tighter correlation to natural gas prices and greater hedging toward crude. However, this exercise validated that the underlying fundamental thesis was still intact: because of XTO's best-in-class hedging program and strong management expertise, they would likely maintain the flexibility to successfully development acquired resources on schedule in 2009 while reducing debt through stronger-than-expected operating cash flows. The market had not appreciated their strong CF position via hedging and operational expertise.

XTO's strategic move to acquire promising, high-profile shale assets in 2008 (45 Tcfe resource base and 14 Tcfe of proved reserves at YE08) has paid off. The transaction price comes out to about $0.90 / Mcfe of 3P (proved, probable and possible) reserves and ~$2.88/Mcfe of proved reserves.

In the end, gross of hedging costs (which detracted approximately 950 bps from performance; Jun-09 38 strike ATM puts bought in advance of May 6, 2009 earnings call), the trade returned approximately 48% cumulative, or 78% annualized.

Looking at the bigger picture, the deal meaningfully expands Exxon’s exposure to North America natural gas prices and unconventional natural gas resources. North America unconventional gas has been dominated by the E&P companies, most notably XTO and CHK amongst many others. Now, Exxon will need to demonstrate that it can develop XTO’s resources at an even lower all-in cost than XTO. I would not be surprised if this is the catalyst which sets off massive M&A activity in the shale space, particularly as the Majors look to secure a longer-term energy outlook amidst record low NG prices.

Technical Options Trade: Short Brazil via EWZ Puts
The long Brazilian Jan-10 80 strike ATM EWZ put position, recommended on Dec-2 was closed on Dec-8 when the +/- 40% PnL threshold was reached (this is my internal trading guideline for 1M ATM options where time to expiry is greater than 20 days). Again, I failed to participate in even further upside due, perhaps, to too conservative risk limits and return thresholds; had I held on to the position until Dec-21 (theoretically assuming perfect market timing), the position would have generated twice the return, with 79.6% cumulatively over a three week period. As I get more experience under my belt, I will come back to re-evaluate these guidelines, but in the meantime, the over-arching credo is that the market can remain irrational far longer than I can remain solvent... so, I'll stick to the side of prudence in the meantime.