Fast pre-salt flow rates (peaking around 26kbpd on average) and low offshore decline rates (just -8.8% / year) were two drivers of high well productivity that we observed in our recent overview of 14,000 wells and 300 fields drilled in Brazil over the past decade. Overall, Brazil’s wells are more productive than other offshore basins we have analysed. In particular, the Santos basin pre-salt appears the “best of the best”, with the most productive wells of all: notably Lula’s Cidade de Angra dos Reis FPSO reached its design capacity around 90-100kbpd with just four (of six planned) wells connected & producing (Exhibit 1, Exhibit 2).

Drilling times are also declining, with the best composite wells’ drilling times likely to be 75% shorter than the first pre-salt wells. The discovery well at Lula took 141 days to drill, the average presalt well this year took 75 days, the best well took just 43 days and the best composite future wells could be drilled in just 34 days (Exhibit 3). At this rate, you could conceivably drill the c4-5 producers and 2-3 injectors required to bring on a new FPSO easily in a year with a single rig.

Will Petrobras have enough rigs after the cancellation? – We think so. Petrobras currently has 17 deep water rigs in operation in the pre-salt Santos basin area. At peak, we model that 30 wells per year must be drilled at the key five BG fields, which could imply as little as four rigs per year are required (assuming no improvement from 2012 drilling rates and a week-long turnaround to move rigs between well-sites). Even with other wells to drill in the Santos, this looks comfortable, given 17 rigs currently present and 28 further rigs under contract to be delivered over the remainder of the decade.

Pre-salt economics continue to look strong, between $5-9/boe, potentially stronger if fewer wells are required. We model an average NPV per barrel of $6.7/boe across BG’s pre-salt fields. Our numbers include c300 wells across these five super-giants, over their lifetimes. With each well costing $100M including drilling, subsea, risers and their connection to the FPSOs, the potential savings are there, albeit smaller in the context of the $160Bn of gross value locked in these five fields today.

All major states require the filing of a drilling permit before the drilling of a well. The filing for the permit is usually one of the last steps taken before drilling. Before the permit is filed for, companies incur significant expenses, including land costs, legal fees, and geological expenses, and, at that time, much of the infrastructure is in place. Thus, most well permits that are filed for are drilled. Barclays extensive analysis shows a strong relationship between drilling permit issuance and rig count. While lead time varies by cycles and states, most often it is roughly two months.

Drilling permits for the 30 states that Barclays monitors increased 0.5% in September 2012, following a 18% rise in August. Barclays believes permit activity reflects a return to normal permitting conditions following summer seasonality and indicates the early stages of a trough in the North American land market.

Offshore permits

The Bureau of Ocean Energy Management (BOEM) issues permits for offshore wells drilled in the GOL on a rolling basis. BOEM permit announcements tend to lag contract awards for the offshore drillers typically by a couple of weeks. However, offshore permit issuances are one of the final steps prior ro commencement of drilling operations and help to indicate future offshore activity levels.

Barclays believes the deepwater rig counts will surpass pre-moratorium levels by year end and reach 45-50 by 2014. In September, the BOEM issued 13 total permits for floating rigs in the US GOM, down from 25 permits announced in August and 27 permits in July. There was one new well permit issued in September (down from 3 in August) and 11 revised new well permits (down from 19 in August). Of the 13 permits issued in September, 11 permits were for exploratory work and 2 were for development jobs. Exploratory locations (shallow water adn deep water) in the GOM stand at 32, up from 26 at year end and 23 locations this time last year. Despite the dip in permits issued in September, Barclays believes permit activity in the US GOM remains healthy and suggests a continued imprevement in the permitting process, which was stifled following the Moratorium. We believe the floating rig count in the US GOM is set to increase further as more deep water rigs are delivered and migrate to the region.

The BOEM issued 28 shallow water permits (including 3 for new wells) in the US GOM in September, down from 34 in August ( 7 for new wells) and 40 in July (3 new).

The mechanics of onshore production are fundamentally different from offshore production. The largest 25 fields off Norway today were developed with 110 total wells per field on average; the largest 25 fields in the Gulf of Mexico were developed with 30 wells on average; while typical Russian fields of the same size (in kbpd) required 1,080 wells to develop: an order of magnitude higher, reflecting lower well production rates of 400-600bpd (versus 2.4kbpd in Norway and 9kbpd in the GoM deepwater).

Russian onshore fields have a “sluggish” production profile, typically taking a decade to ramp up to peak. It takes 6-7 years to recover half of the oil from the average Norwegian / UK field. A “peaky” GoM field can recover half of its oil after 2-3 years. But in Russia, it takes as long as 20-years to recover
half of the oil at a typical field, based on the production profiles. Because of the time value of money, these “sluggish” production profiles yield 40% lower NPVs per barrel than offshore production profiles. We estimate undeveloped Russian oil resources are worth $0.7/bbl when fiscal terms are considered.

Forward-looking Russian decline rates are likely to be higher than the past, at -3.5% on average. Russian fields’ decline rates accelerate as the fields mature. Today the weighted-average barrel in Russia is produced from a field that has been in production for 30-years. Only 12% of production is from fields in their first decade of production (when production typically ramps up) vs 30% two decades ago. Incorporating this higher decline rate from very mature fields in our oil market models would subtract 0.5Mbpd of supply by 2015 & 0.7Mbpd by 2017 and increase oil price forecasts by 5%. Our 2017 Russian forecast, at 9.7Mbpd would be 0.85Mbpd below the IEA’s recently updated estimate, using this decline rate assumption.

The History and Maturity of Russian Onshore Production
Russian oil production ramped up to 11.4Mbpd at peak in 1987-8 according to the BP Statistical review, and fell precipitously after the collapse of the Soviet Union, before a reorganization of the upstream industry in the late 90’s and early 2000’s recently restored output to a new high: 10.4Mbpd this year on average (Exhibit 1). Although Western Siberia is the largest component of Russian output today, at c70% of the total country’s production, its ramp-up took-place mostly in the 1970s as Pre-Caspian and Volga- Urals production matured and declined. Today, West Siberia is in decline, while recent growth has been driven by greenfields and increased production in East Siberia, Timan-Pechora and the Far East.

Therefore the increase in Haynesville production was not so much to do with new wells being drilled, but
more the backlog of wells available for completion that had already been drilled (Exhibit 12). While a
(greatly shrunk) backlog still exists, its current level could likely sustain flattish production for only an
additional few months without new drilling.

With the backlog almost fully eroded and only ~20 rigs now drilling in the play, the chance for material
declines in the December or January timeframe remain a distinct probability, potentially to the tune of ~200 mmcfd a month (referring back to Exhibit 9 – but noting that absolute declines decrease as the overall production base falls).

Prior to October 24, 1978, the airline industry was heavily regulated by the government which caused limited price competition and new entrants. After deregulation, start-up airlines with low fares began to take market share from legacy airlines. There are two business models: Hub and spoke, which is used by legacy airlines (UAL, DAL, AMR, LCC), and point to point, which is used by start-ups (LUV, JBLU).

Since deregulation in 1978, the airline industry has collectively lost more than $37B and has seen hundreds of bankruptcies. In between, there have been cycles when the airline industry has been very profitable (1983-1989, 1993-2000, 2006-2007, 2010-2011) but years coinciding with recessions have been very bad (1990-1992, 2001-2005, 2008-2009). This has caused airlines stocks and credits to be trading vehicles rather than investments.

There are many structural problems with the industry including:

1)      High fixed costs with variable demand

2)      Airlines add too much capacity in order to take market share

3)      Low barriers to entry cause too much competition

4)      Capital intensive with high operating and financial leverage

5)      Government over-regulates operations

6)      Exogenous shocks such as terrorism and natural disasters

Supply:

The standard unit of supply is ASM (Available Seat Mile), which is defined as the number of planes times the number of seats times the number of miles flown. This is also known as capacity. Airlines can increase capacity is 4 ways:

1)      Increase utilization (each plane makes more trips)

2)      Add more seats per plane

3)      Bring back parked planes

4)      Buy new planes

ASM is a closely watched metric because if the industry increases capacity, it will most likely lead to pricing pressure, especially if capacity increases during an economic slowdown.

Demand

The measure for demand known as traffic or RPM (revenue passenger mile), calculated by the number of passengers times miles flown. Load factor is PRM/ASM (demand over supply or utilization). Yield is passenger revenue /RPM. PRASM is passenger revenue per available seat mile. RASM is total revenue per available seat mile.

Airlines individually release monthly RASM numbers for the previous month around the middle of the month. Consensus expects RASM for the industry to be up 5% in 2012. However, a slow economy could take RASM to flat in 2012.

Business travels typically account for 40-45% of traffic and 55-60% of revenues (65% for legacy airlines). Business travel tends to fluctuate with economic activity and has a short bookings window (50% made 2 weeks or less in advance) while leisure bookings have a longer lead time.

Other sources of revenues:

Cargo accounts for about 3% of revenues

Ancillary fees account for about 6% of revenues

Sales of frequently flyer miles account for about 6% of revenues

Cost structure:

Fuel costs are 33% of sales : fuel is driven by flight hours but airlines with older fleets consume more gallons per seat mile. Current brent crude price of $110 per barrel equates to about $3.10 per gallon of jet fuel.

Labor costs are 28% of sales: About 50% of U airline employees and more than 90% of airline pilots are unionized. Unlike other industries, airline union contracts do not expire, but become amendable at the end of their stated term. The existing contract remains in effect until a new contract is negotiated. Unions are not allowed to strike and management cannot impose a contract without government permission.

Capital costs are 12% of sales: These costs include depreciation, aircraft rent, and net interest expense. More airlines are leasing rather than buying partly because their large NOLs (net operating losses)  do not allow them to take advantage of the tax benefits of accelerated depreciation and partly due to lower costs and increased flexibility. Although US accounting rules do not require airlines to put most lease obligation on the balance sheet, the investing community typically capitalizes these obligations at 7x annual aircraft rent.

Maintenance costs are 8% of sales

Distribution costs are 8% or sales: This includes travel agents, reservation systems, credit card fees and advertising.

The most common measure of airline costs is CASM (Cost per available seat mile).

Hard disk drives (HDD) are electro-mechanical storage devices that use rotating magnetic media to store and retrieve electronic data. They were first developed by IBM in 1956 for the IBM 360 computer as the primary means to store large amounts of data. HDDs are now used in PCs, servers, external storage arrays and consumer devices.

The form factor of HDDs has been continuously shrinking, from the original 24 inch diameter drives to 5.25 drives in the 1990s to 3.5 inch or smaller today. HDDs come in three form factors today: 1) 3.5 inch used for desktops, servers, storage arrays and consumer electronics. 2) 2.5 inch used for notebooks, servers, storage arrays, portable electronics. 3) smaller than 1.8 inch used for portable electronics.

Inside the HDD

There are two main parts in a HDD, a base and a printed circuit board (PCB)

The HDD base contains mechnical parts. The magnetized data are stored inside the disks located in a HDD, and these disks are called platters (also known as “media” or “disks”). The platters consist of substrates and the media layer, which is thinly coated magnetic material (e.g. a ferrite compound) on the substrates (typically 10-20nm) and store the actual data. Due to the mechanical nature of HDDs, the platters have to be extremely smooth and flat, as uneven surfaces could lead to head crashes and cause the HDD to fail. This is becoming increasingly important, as technology advances have been increasing the rotation speed of platters and, at the same time, decreasing the gap between the platters and the heads. Aluminum substrates are mainly used in HDDs for desktops and servers, while glass substrates, which can be made thinner and are more heat-resistant, are used in HDDs for notebooks and other non-PC applications.

Most HDDs have at least two platters and the capacity of the drive increases as the number of platters increases. Each platter is magnetized on each side, so a HDD with two platters has four sides to store data. The platters are clamped to a rotating spindle that turns the platters in unison and a motor is mounted directly below the spindle and spins the platters at a constant rate from 3,600 RPM (energy-efficient portable devices) to 15,000 RPM (high-performance servers /storage systems).

As the platters are spun by the motor, heads read and write data to the platters. There is typically one head per platter side, and they are attached to an actuator arm (head actuator), which moves the heads around the platters.

The PCB consists of several electronic components, including: (1) a Micro Controller Unit (MCU), which makes all calculations and converts signals during the read/write process; (2) a memory chip, which is used as a cache; (3) a Voice Coil Motor controller (VCM controller), which controls spindle motor rotation and movements of heads; (4) a flash chip, which stores part of the drive’s firmware (along with MCU); (5) a shock sensor, which detects excessive shock applied to the drive and send a signal to VCM controller; and (6) a Transient Voltage Suppression diode (TVS diode), which protects the PCB from power surges from external power supply.

Overview of HDD supply chain

HDD components and finished HDD products are small and lightweight, and the associated transportation cost is low. As a result, the HDD industry’s production process is divided into several stages, and each of these production processes is done in different countries, mostly in Asia. The figure below illustrates a simplified manufacturing process for HDDs.

 Asia accounts for the entire finished HDD units manufactured, with Thailand accounting for around half of global HDD production, followed by China (30%, driven by better access to PC contract manufacturers in China), Malaysia (12%) and Philippines (5%).

Looking back, the evolution of the HDD industry in Asia started in Singapore, as Singapore had a number of location advantages for the HDD industry such as a tax incentive in the form of a full tax exemption for typically five years, and knowledgeable suppliers and labor. A number of semiconductor companies such as National Semiconductor and Texas Instruments had already shifted their production to Singapore in 1960s, developing adjacent IT industries in the region.

Starting in the 1980s, HDD companies gradually moved their lower-end manufacturing facilities from Singapore to other lower-cost countries in Asia such as Malaysia, Thailand, and more recently, China, as a labor supply shortage in Singapore began to increase labor wages in the region. Specifically, Thailand’s promotion of foreign investment through various investment incentives (e.g. a full tax exemption for 3-8 years) in the 1980s, and Seagate’s decision to shift some of its assembly facilities to Thailand, catalyzed other companies to follow shortly.

HDD component suppliers & capital equipment manufacturers

HDD component vendors design and manufacture mechanical / semiconductor components for HDD vendors. It is important to note that HDD vendors have been satisfying an increasing portion of their key component needs in-house to prevent any disruption in component supply and achieve higher profitability through vertical integration. For example, Seagate, HGST, and Western Digital procure some or most of head, substrate, and finished media in-house today. In addition, finished HDD assembly is often outsourced to the other HDD supply chain companies. For example, TDK/SAE is responsible for manufacturing finished HDD products for Samsung (90%+ of Samsung’s monthly production capacity) and Toshiba (15% of Toshiba’s monthly production capacity).

There are a number of equipment manufacturers for HDD vendors that design and manufacture capital equipment. For example, Intevac and Anelva produce sputtering systems that deposit alloy films onto substrates. Xyratex and Teradyne manufacture automated testing/inspection systems in the areas of disk assembly, head fabrication, and substrate/media manufacturing

Market shares

During the 1980s, global diversified mining companies, domestic steel companies, electric utilities, and integrated oil companies controlled coal mining assets. Corporate strategies were around maximizing volumes, and gaining market share. Electric utilities, the primary customers, had considerable pricing power over the coal producers. Many competitors exited the industry through consolidation or bankruptcy.

In 2000, a 20 year long bear market for coal found a bottom when increased electricity demand and a sharp climb in oil and natural gas prices forced utilities to look to coal producers to restock inventories and meet generation requirements. Contractual commitments at lower prices prohibited many coal producers from timely meeting those needs and thus prices began a long uptrend. Today, management teams have become more disciplined about capital investment and returns and have also moved to a market driven strategy as opposed to a production driven one.

Supply

Surface mining methods are generally used to extract coal west of the Mississippi. Large underground mines can be found in Pennslyvania and West Virgina. Eastern US coal is in tight supply due to periodic mine closures.

Demand

Coal’s total electricity generation has ranged from between 41% and 51% of the total during the past 20 years. Natural gas is becoming a substitute due to low prices. Nuclear power is losing share post the Japan earthquake.

Pricing

Pricing for Central and Northern Appalachian thermal coal is higher due to its proximity to the East Coast power plants and the supply constraints as mines close down. Powder River Basin  rices are very low because most mines are surface mines that are low cost to extract. Met Coal that is used for steel making is much higher priced due to strong demand (China) and weak supply (Australian floods).

Transportation

About 58% of mined coal output is shipped by rail, 22% by inland waterway, 11% by trucks, and 8% by conveyor belts directly to power plants.

Metallurgical Coal

Metallurgical coal (also known as coking coal) is used in the production of steel. Met coal is a low-ash, low –sulfur bituminous coal with a heating value between 12,000 and 14,000 Btu per pound, which differentiates it from thermal coal that have lower heating values. Australian  producers and Japanese buyers annually negotiate the benchmark met coal price for the April-March fiscal year. In 2008 and 2011, severe flooding impacted export volumes from Australia, which helped met coal prices rise from $100 in 2007 to $305 in 2008. Prices rose from $225 in 1Q11 to $330 in 2Q11. Since 2010, the met coal industry moved from annual fixed priced contracts to quarterly contracts. The move to shorter contracts opened the door for US exporters to increase volumes.