Quick Summary:

Despite recent appreciation, the Calpine 8.625% and 8.5% unsecured bonds (due 2010 and 2011) remain an attractive opportunity. With bonds at 105.5, the implied return based on a par plus pre and post-petition interest recovery and an emergence date between September 30^th and year-end 2007 is 17%-23%. Given the unsecured bonds should receive the bulk of the restructured equity, unless taken out at 118 to 121 (depending on the date of emergence), the potential appreciation could be even greater (ala Mirant and NRG – these restructured equities now trade well above their par plus pre-petition and post-petition claims). In essence, Calpine offers an opportunity to buy an irreplaceable set of environmentally-friendly US infrastructure assets at below replacement cost and right as industry fundamentals begin to improve. As I see it, the real downside is being taken out at 118 to 121 by the existing equity or an acquirer and the upside is owning all or a portion of the restructured equity, which could trade far higher than the bond claims.

 

At current prices, the Calpine unsecured bonds trade below plant replacement cost (<83%) and at ~10x adjusted pro forma 2007 estimated EBITDA and ~12x EBITDA minus maintenance capital expenditures. On its face, these multiples do not appear cheap, but EBITDA and free cash flow should grow substantially as the power market exits trough levels and enters mid-cycle and peak pricing over the next several years. Small improvements in spark spreads and capacity factors will be magnified via the company’s significant operating and financial leverage. Furthermore, any changes in environmental regulations should only positively impact Calpine.   

 

Since Calpine has not been written up in a while, I will start with a quick bankruptcy and timing update followed by a company and industry overview, asset description, claims summary, valuation, and investment risks section. The company and industry overview and asset description should probably be appendices, so please feel free to skip these sections as this report is very long.

 

Bankruptcy Update:

Calpine is the 6^th largest US bankruptcy by asset size in history.

Why did Calpine file for bankruptcy?

1) High leverage and limited liquidity from aggressive project development and rapid growth.

2) Difficult industry trends and natural gas volatility as spark spreads and capacity factors remained below company expectations and escalating natural gas and power prices resulted in unprofitable fixed price contracts in many instances.

3) Unsuccessful restructuring (non-bankruptcy) efforts.

 

Since filing for bankruptcy, Calpine has made significant progress divesting underperforming plants, restructuring unprofitable contracts and selling surplus or idle assets. This has occurred in conjunction with improvements in the power market and increases in plant valuations.

 

The company recently provided investors with a roadmap for its emergence from bankruptcy. With debtor exclusivity set to expire June 20, 2007, the company is on a fast-track to emerge from bankruptcy in late 2007 (CEO claims YE 2007 is the latest Calpine would emerge), before the power market recovery is fully reflected in its numbers.

 

Per court filings and the DIP refinancing presentation, Calpine intends to continue to sell non-core assets. Often these assets generate minimal or negative cash flow, so it’s an addition through subtraction play. As part of its portolio optimization strategy, Calpine announced cost cuts and rejected or renegotiated several executory contracts. The company estimates run-rate cost cuts of $170 MM by the end of 2007 via the elimination of 1,100 positions (1/3 of its workforce), non-core office closures and reductions in controllable overhead costs and $175 MM of additional savings through the rejection or restructuring of out-of-the-money contracts or uneconomic leases.

 

Recent court filings suggest a potential equity capital raise. A rights offering to existing shareholders would imply value of par plus pre and post-petition interest on the senior unsecured bonds, but would clearly cap some of the potential. My sense is that the aforementioned rights offering relates to the unsecured creditors raising funds to retire the 2^nd lien debt.  

 

Brief Company & Industry Overview:

(Largely paraphrased from Calpine’s Motion to Extend Exclusivity Filed November 22^nd)

Calpine operates the largest fleet of natural gas-fired power plants in North America, with enough capacity to meet the electricity needs of almost 27 million households, or ~3.5% of the electricity consumed in the US. This makes the company the largest consumer of natural gas in the US. The company also owns geothermal plants in Northern California which generate renewable energy with negligible air emissions.

 

In total, Calpine operates around 25,000 MW of capacity (pro forma for recently announced divestitures), excluding the plants under active construction. These plants emit significantly less carbon dioxide, mercury, sulfur dioxide and other pollutants compared to coal or oil-fired plants. The majority of Calpine’s plants are combined-cycle or cogeneration facilities using natural gas-fired combustion turbines and steam turbines that operation in tandem to generate power and steam. Compared to older technologies, this process requires up to 40% less natural gas as recognized by the company lower heat rates (a measure of plant efficiency).

 

Since electricity cannot be stored and must be consumed as produced and can only move where transmission is available, the domestic electricity industry is divided into eight regions (ERCOT, FRCC, MRO, NPCC, RFC, SERC, SPP and WECC). Each region has its own regulatory regime, supply and demand dynamics and efficiencies. Certain regions are more regulated than others, or said differently, some regions have well-defined market rules and a “level playing field for sellers,” while others are run by incumbent utilities who control transmission access and own substantial competing generation. Due to transmission limitations, surplus capacity in one region often cannot be sent to areas of undersupply.

 

Demand has historically grown at 2% annually in the US. The North American Electric Reliability Council (NERC) predicts summer peak demand to grow at 1.9% from 2006-2015. While demand is fairly stable, capacity additions are often quite lumpy due to the boom and bust nature of the industry. Currently, the US has roughly 900,000 MW of capacity with approximately 231,000 MW of supply additions completed in the 1998-2005 timeframe. Recent capacity additions led to significant oversupply in many areas, but going forward new supply is expected to fall short of projected demand growth.

 

Electricity demand is expected to grow by almost 69,000 MW over the next several years compared to resource additions of fewer than 40,000 MW (TXU’s CEO expects demand growth to outstrip supply by 2.5x). Given the tightening supply and demand dynamics, I expect spark spreads to increase in the coming years as the market clearing heat rate should continue to rise in many areas. Higher cost, marginal assets, such as Calpine’s combined cycle gas turbines and peaker plants (hydroelectric, nuclear and coal assets are cheaper power generators and turn on first to serve load requirements) have the highest operating leverage to the tightening supply/demand dynamics.

 

Asset Description:

Depending on the composition of assets and fuel types in various regions, the price of electricity and profitability for merchant generation often varies significantly by location. The majority (>80%) of Calpine’s assets are in California (WECC region), Texas (ERCOT) and the Southeast (SERC). Other plants are located in Florida (FRCC), New Jersey (MAAC), New York (NYPOOL), Maine (NEPOOL), the Midwest (MAIN & MAPP) and Oklahoma (SPP).

 

ERCOT:

The ERCOT market encompasses most of Texas and is divided into five zones - West, North, Northeast, South and Houston. Total MW capacity is slightly above 70,000 MW, excluding mothballed units of roughly 8,200 MW. Calpine owns and operates around 7,500 MW of baseload and peaking generation. The company’s assets are primarily located in the Houston load zone with a couple of plants in the southern region and one in the north. Compared to the rest of the country, ERCOT’s capacity consists of a larger percentage of simple and combined cycle natural gas plants (71% vs. 39% average nationwide).

 

Electricity demand growth in Texas, especially in Houston and Dallas/Ft. Worth (ERCOT-North) is among the highest in the nation and has led to reduced reserve margins. For instance, peak demand load in summer 2006 hit levels not expected to be seen until 2008. Capacity additions from 2000-2004 (20,000 MW) have been largely worked off to the point where reserve margins in 2008 are expected to fall below the 12.5% requirement set by the ERCOT board of directors. New cheaper power sources (coal, wind and nuclear) will take time to build, which should result in strong profitability for Calpine for at least the next several years and probably longer. To bring the previously mothballed units back on line, the market clearing heat rate would have to increase, again resulting in higher margins for the more efficient Calpine portfolio.

 

WECC:

The WECC comprises all or portions of California, Oregon, Washington, Idaho, Montana, Colorado, Utah, Wyoming, Nevada, Arizona, New Mexico, British Columbia and Alberta and small portions of Texas, Nebraska, South Dakota and Mexico. The bulk of Calpine’s WECC assets are located throughout California, specifically Northern California, which is forecast to be firmly in equilibrium by 2008.

 

High barriers to entry including limited site availability, environmental restrictions and high costs (unionized labor) make the California assets the crown jewel of Calpine’s portfolio with the geothermal units being the most valuable on a per kW basis. Through various programs, utilities are being urged to enter into contracts so that 20% of California’s energy supply will come from renewables by 2010. Since the fuel cost of the geysers is basically zero, the value of these assets fluctuates significantly with natural gas, which frequently sets the price of power in California. Value enhancement from increased renewable energy credits represents additional upside.

 

The Northern California market where Calpine primarily operates, while not as tight as Southern California is certainly a strategic location considering the healthy demand growth and difficulty of developing coal and nuclear facilities, or even natural gas assets for that matter. At a January meeting, the California PUC approved a measure that prohibits utilities from entering into long-term contracts to buy electricity from resources that emit substantial greenhouse gases (i.e. coal). As utility contracts expire, new contracts should be subject to this rule. In summary, the long term outlook in California is very favorable for CCGTs and geothermal plants.  

 

SERC:

The southeast is the most difficult market in the country. This region is divided into five sub-regions: Virginia & Carolinas (VACAR), TVA, Southern, Entergy and Gateway. Deregulation never took hold in this part of the country as many merchant builders had hoped when they (over)built in the early 2000s. Merchant generators have generally suffered from low margins due to the overbuild coupled with limited transmission access and a lack of organized markets in the monopolistic utility systems. Unlike other markets, the SERC is not centralized and does not have an RTO or ISO, so transactions are handled on a bilateral basis with the major utilities.  The market is controlled by the large utilities – Entergy, Southern, Tennessee Valley Authority, Duke and others, who own the transmission and frequently run their own more inefficient plants in lieu of Calpine’s more efficient units.

 

In the SERC, reserve margins are forecast to approach 31% in 2007. Natural gas plants run very sparingly as coal, nuclear and hydro represent over 65% of the market. Furthermore, roughly 1,160 MW of coal-fired capacity is under construction (to be operational by 2009), while another 7,800 MW are in different stages of development. If all of this capacity is added, reserve margins would increase ~4%.

 

Valuations have improved in the SERC as demand continues to increase and the PUCs become more independent (little by little). Additionally, the aging of certain generation stock (Entergy and Southern Company) should result in asset retirements, while investments in transmission ($6.3 BN over the next five years) should provide generators in the region with better access to other markets.

 

Other Calpine assets are located in Florida, New Jersey, Wisconsin, Illinois, Minnesota, Maine, Ontario and New York. Most of these plants are partially or fully contracted. 

 

From what I have been able to piece together from the company’s filings, there is about $7.8 BN of DIP, Calgen, CCFC and other project debt. The 2^nd liens represent an additional $3.7 BN of debt and senior unsecured bonds at face amount to another $5.7 BN. Other general unsecured claims are trickier to calculate. In the Debtors’ motion establishing sell-down procedures for trading in claims (i.e. preservation of NOL), a footnote indicates a “very rough preliminary estimate of $1.8 BN of other general unsecured claims.” I believe this number is “at market,” so assuming a 75% trading claim price at the time of this filing and netting out an estimated $100 MM for a claim reduction (retolling of Calgary Energy plant), I compute other GUCs of $2.3 BN. I think this number is conservative based on post-bankruptcy contract renegotiations in lieu of rejection.

 

Before I move on to the valuation, I want to mention my cash estimate of $2.5 BN, which includes unrestricted and restricted cash of $1.2 BN and $0.6 BN as of 11/30/06 plus asset sales (Power Systems Manufacturing, Goldendale Energy Center, Aries) and other non-core asset recoveries (claims against Solutia for contract rejection, 30% stake in Calpine Power Income Fund, etc.). Additionally, I assumed an NOL of $3.5 BN (present value closer to $1 BN) per Calpine’s estimates. The NOL should increase due to lower COD income given the bond price appreciation.

 

Note: My recommendation is for the 8.5% and 8.625% senior notes due 2011 and 2010. The other notes have various nuances impacting their trading value. For instance, the ULC I and II notes should have greater recoveries (up to par plus) given the potential double-dip and additional assets in the Canadian subsidiaries and the 7.75% notes due 2015 have subordination provisions requiring them to relinquish recoveries to the pre-2000 issues (up to par or par plus).

 

DIP & Project Debt                              $7.8 BN

2^nd Lien Debt                                       $3.7

Senior Unsecured Notes                       $5.7

Other GUCs                                         $2.3

Total @ Face                                       $19.5 BN

 

Cash                                                     $2.5

NOL                                                     $1.0

Net Claims                                            $16.0 BN

 

*At market, the claims increase slightly due to prepayment penalties and trading levels above par.

 

There are several ways to value Calpine. I used per kW multiples, EBITDA, unlevered FCF and FCF. It’s important to consider unlevered FCF and FCF because of Calpine’s lower pro-rata capital expenditure requirements compared to NRG, Dynegy and Mirant, or other comparables with coal-fired or older units. These companies trade around 8.5-9.0x EBITDA on average compared to ~10x for Calpine.

 

On a per kW basis, Calpine trades at ~$590 (netting out my estimated value of $3,000 per kW for the geothermal assets). Using estimates from various power consultants and construction experts, my blended replacement cost for Calpine’s plants (pro forma for geographical differences and dispatch types) is slightly above $700/kW. On this basis, Calpine trades at roughly 82% of replacement cost.

 

California:

The power market recover is region specific. Assets in California and Texas should trade at much higher valuations than assets in the SERC and SPP. Given the barriers to entry in California (i.e. rigorous environmental requirements and permitting process) and tightening supply/demand characteristics, these plants should trade above replacement cost. A plant already built will be earning substantial profits during the time a new plant is being constructed, if it even receives approval. And why would a plant operator build a plant now, especially after the previous bust cycle, if he didn’t believe the value was greater than the construction cost?  As such, I value the California baseload assets at $1,100-$1,200 per kW or $100-$200 per kW above the estimated cost for a new CCGT is Northern California, where Calpine has a cluster of plants. I value peakers at roughly 2/3s the CCGTs. My estimates for non-California plants in the WECC range from $400/kW (peaker) to $800/kW depending on location and if the plant is merchant or contracted.

 

Geothermal profitability depends on natural gas prices, which often set the price of power in California. Considering the current forward curve and the trading level of Ormat Technologies (>10x 2008 EBITDA and >$3,400 per kW), a decent comparable, I value the geysers at $3,000/kW. If anything, I think this value is conservative considering expansion opportunities and my estimated EBITDA contribution.

 

Other incremental value in California comes from development opportunities. It’s difficult to determine how many new plants can be built, but from different sources I estimate between three and five sites have received ISO approval for CCGTs. Another possibility is geothermal expansion. Calpine is evaluating an opportunity to add 50 MW of capacity for start-up in the 2010-2011 timeframe. Furthermore, Calpine controls 46,400 acres of leases at Glass Mountain, the largest undeveloped geothermal resource in the lower 48 states, which could result in 480 MW of capacity between 2011 and 2017. This additional value coupled with a potential capacity market, which is being discussed by the California PUC for possible 2009 implementation make the California portfolio even more valuable.

 

As mentioned above, the Texas market continues to tighten due to strong demand growth. Calpine’s well located plants (most clustered in the Houston load zone) coupled with improving market conditions suggest these assets should trade close to replacement cost or even higher. While the barriers to entry in Houston are noteworthy, they do not compare to California. Moreover, new build costs in Texas pale in comparison to California, so plants should trade at lower values. As such, my blended valuation for these plants approximates replacement costs of $650-$700 per kW. The Houston load zone is somewhat protected from the TXU coal plant expansion because of limited transmission from the Dallas-Ft. Worth area into Houston. Plus, TXU recently announced a scaling back of its coal expansion (11 initially, now only three planned).

 

One thing to note is that both Calpine (400 MW addition to Deer Park) and NRG are planning to add natural gas capacity (peakers and intermediate plants) in Houston. Presumably, NRG expects to earn a decent return on capital (12-16%) or the plant would not be built. NRG estimates that the cash cost (excluding preparation and finance costs) for these plants will be less than $500/kW. Assuming $450-500/kW costs, this implies roughly $520-575/kW for the CCGT and obviously less for the peaker. These plants are not scheduled to be completed until early 2009, so two good years will be missed.  Since these plants are being repowered, the all-in cost should be 30-40% cheaper than a greenfield project. This implies greenfield costs of $675-$800/kW and is one way to triangulate replacement costs in Houston.

 

SERC:

The SERC assets are even tougher to value because they generate limited EBITDA (many estimate negative). Location and transmission access are key factors in valuing these plants. To evaluate these plants, I looked at project debt comparables, the KGen IPO and the recent and proposed Kelson Energy and Entegra refinancing.

 

Prior to the announced refinancing by Kelson Energy, Magnolia (Mississippi), Redbud (Oklahoma – SPP) and Cottonwood (Texas SERC) traded above $400 per kW. Pro forma for the refinancing, the first and second lien leverage is above $350/kW. Recent appraisals (from a Kelson Energy consultant) value these assets at an average of $625/kW. The KGen deal priced at over 20x EBITDA and around $300/kW (adjusting for the peaker). The Entegra trading values imply a high $300 per kW value for the Arkansas plant in the Entergy region of SERC with an EBITDA multiple over 35x.

 

Some of the Calpine plants have partial contracts and the units located in the VACAR and TVA sub-regions of SERC are likely better than those located in the Southern or Entergy sub-regions. My blended estimate is slightly below $300/kW, which is probably light considering the aforementioned comparables. A $50/kW increase amounts to $280 MM of additional value.

 

I valued the other plants in Calpine’s portfolio at $300 to $800/kW depending on the location and whether it was merchant or contracted. For the plants in Oklahoma, I used SERC-type multiples and for some of the recently constructed and contracted plants I used multiples implying replacement cost. Some of Calpine’s newer plants (i.e. Mankato in Minnesota) were built only after signing contracts, so these should trade at or above replacement cost.

 

The sum of this is a valuation of $17.5 BN (including little value for plants under development and for the California permits). Including the cash, NOLs and other non-core assets, I compute approximately $21 BN of value or par plus post petition interest on the bonds (>120).

 

California Geothermal                                                          $2.2 BN

Remaining California & Other WECC                                  $6.0

Texas                                                                                  $5.2  

SERC                                                                                 $1.6

Other Assets                                                                       $3.0

Cash                                                                                   $2.5

NOLs                                                                                 $1.0

Total                                                                                   $21.5 BN

 

*Geysers valued at $3,000/kW compared to Ormat’s trading value of >$3,400/kW. My EBITDA est. suggests a higher valuation.  

**For reference, in a bankruptcy court filing to get approval to build the Russell City plant in California, Calpine estimated that upon completion with a power purchase agreement in place, the facility would be worth $1,081-$1,402 per kW.

***Mach Gen with plants in Athens, NY, South Haven, MI, Tonopah, AZ and Charlton, MA trades at ~700/kW.

****~$2.3 BN of the GUC are other liabilities, which would have lower pre and post petition claims than the senior notes.

*****According to Citibank, there is a 0.77 r-squared relationship between capacity margins and enterprise value per MW.

 

Cash Flow Valuation:

Using pro forma 2007 EBITDA of ~$1.6 BN, Calpine trades at ~10x EBITDA. I adjusted the company’s 2007 EBITDAR estimate to exclude rent payments and added the additional cost cuts expected to be realized in 2008. Since the projections were based on the forward curves from June 2006, I factored in improved market heat rates and capacity factors, net of the decline in the forward gas curve. These adjustments result in EBITDA of over $1.6 BN compared to the company’s $1.454 EBITDAR projection.

 

This multiple is net of cash, non-core assets and NOLs. On its face, this is expensive. However, significant growth is expected and maintenance capital expenditures should be minimal, so EBITDA is a good cash proxy. As an example of the growth potential, Reliant expects wholesale margins to increase 44%, 50% and 20% in 2007-2009. While Reliant is a bit of a different animal (coal plants and PJM exposure), I think the company’s projected growth presents a good case of how quickly cash flow can grow.

 

Calpine’s last presentation prior to filing for bankruptcy (3Q05) shows a contractual portfolio that becomes significantly more merchant in 2007 and 2008 and beyond. For example, the ERCOT portfolio declines from 49% hedged in 2006 to 26% and 24% hedged in 2007 and 2008. WECC decreases from 58% in 2006 to 29% in 2010. SERC falls from 69% and 54% in 2006 and 2007 to 37% in 2008 and 28% in 2010.  As power markets tighten and Calpine’s “open” megawatt portfolio increases, EBITDA should continue to grind higher.

 

From a comparables perspective, Mirant, NRG, Dynegy and Reliant trade at lower EBITDA multiples than Calpine, but I would argue these names have higher capital expenditure requirements and don’t have as much EBITDA upside to market heat rate expansion. Project bank debt issues with depressed EBITDA are valued at over 20x. Mirant recently sold natural gas assets to LS Power at 14-15x EBITDA and Dynegy recently purchased natural gas assets from LS Power at 10x 2007 EBITDA.

 

Furthermore, the aforementioned EBITDA multiple of ~10x includes little contribution from the SERC assets. Another way to view the EBITDA multiple is to subtract the estimated SERC value (on a per kW basis) from the enterprise value, while removing the cash flow contribution. Doing this reduces the EBITDA multiple to the low 9s.

 

Capital Expenditures and Unlevered FCF:

Since these plants are relatively new, the ongoing capex requirements should range from $100 to $200 MM annually. Plant operators estimate $1.5 per MWh of capital is required to maintain the plant. If Calpine generates 100,000 MWhs of capacity (2006 estimate of 85,000), $150 MM will be required to maintain the plants. On a recent call, Calpine indicated 2005 maintenance capex approximated $250 MM. Splitting the difference and using $200 MM results in a 2007 unlevered FCF multiple of ~12x.

 

Free Cash Flow:

Real quick, on a 2007 pro forma free cash flow basis, we should have a restructured equity trading at ~14x FCF or under 12x, excluding the SERC. This contemplates a global restructuring of Calpine’s project debt. To make the numbers easy, I assume all project debt is refinanced at 8%. I add in extra debt to pay make wholes and other fees. Capex is $200 MM and I estimate taxes using a 38% tax rate (and assuming a depreciation shield). In the example, the 2^nd liens notes and unsecureds share the equity, even though the 2^nd liens will almost certainly be taken out via a rights’ offering.

 

Equity Case:

While this report is intended for the bonds (which I have always liked more), the Calpine stock at $1.73 is also intriguing. At current prices, the equity market capitalization is nearing $1 BN. The equity committee proposed hiring Perella Weinberg at a $150,000 per month fee plus a success fee based on recoveries by stockholders. This arrangement is fairly unorthodox and may represent Perella Weinberg’s view on its ability to recover substantial proceeds for the equity holders via an outright sale to a strategic or private equity buyer or valuation fight.

 

Calpine bulls point to a valuation of $25 BN or 10x “mid-cycle” estimated EBITDA of $2.5 BN. Power plant builders generally require 12-16% cash on cash ROEs. Assuming 70% debt (@ 8%) and 30% equity and $200 MM of capital expenditures results in roughly $2.5 BN of EBITDA for Calpine’s plants and geysers. Another data point is Pete Cartwright’s (former CEO) estimate of market equilibrium. Obviously he was a power market bull, but his mid-cycle estimate of 80,000-100,000 of EBITDA per MW equates to $2.0-$2.5 BN.

 

Figure every billion of value above the amount of debt equates to $1.85 per share. This equity upside is tempered by the near-term emergence from bankruptcy and a potential cramdown. At the very least, equity holders should get warrants, so downside is fairly limited. We own both bonds and equity.

 

 

Conclusion: At current levels, the bonds have attractive risk-return characteristics. In my worst case scenario, bondholders end up with a 17-23% return and in my best case they receive the bulk of the restructured equity, which could trade far higher than the unsecured pre and post-petition claims of 118 to 121. While more risky, the Calpine equity also exhibits an asymmetric risk-return profile at $1.73 per share.