PLUM CREEK FINANCIAL MODEL EXECUTIVE SUMMARY

PLUM CREEK FINANCIAL MODEL Discussion Paper No. 2 Open Space Institute Industrial Economics, Incorporated 14 September 2007 EXECUTIVE SUMMARY The Plum Creek financial model provides a framework for analyzing the financial value to Plum Creek's shareholders of the company's lake concept plan development proposal for the Moosehead Lake area in Maine, as well as possible alternatives to that proposal. The intent of the model is to inform the public discussion and give participants in the ongoing development review process a tool for evaluating the company’s proposal and potential alternatives. To illustrate the use of the model and give one measure of the value to Plum Creek of concept plan approval, we run two scenarios and present the results here. The first scenario comprises the baseline build-out scenario presented in Discussion Paper No. 1: "Baseline Development Scenario for the Plum Creek Moosehead Project Lands." The second scenario relies upon the assumptions presented in Plum Creek's April 2007 "Concept Plan for Plum Creek's Lands in the Moosehead Lake Region." By comparing the results of the two model runs, we provide an estimate of the incremental value to Plum Creek shareholders of LURC approval of the proposed plan. With the assumption that the baseline scenario for developing its lands around Moosehead Lake is the company's next-best alternative to the proposed concept plan, our best estimate of the value to Plum Creek of concept plan approval would be approximately $39 million, which is the difference between the estimated value of the April 2007 plan (approximately $106 million) and the value of the baseline (approximately $67 million). Of this difference, about $19 million can be attributed to the residential development portion of its plan, which is the difference between the estimated value for the residential development portion of its April 2007 plan (approximately $50 million) and the value of the baseline build-out (approximately $31 million). Given the uncertainty in some of the model inputs, we would bracket the $39 million estimate very roughly within a range of $30 to $45 million. Plum Creek has provided much less information on the resorts than it has on its other development components, however, and our analysis of this plan element is thus substantially more uncertain than that for the residential development portion of the plan. Based on our review of the resort descriptions outlined in the plan documents and related submissions, as well as our discussions with various experts in the resort-development field, we expect that if the plan is approved Plum Creek will develop not only the lodge or inn facilities associated with each resort area, but also a significant number of residential lots within the re-zoned parcels. As a result, we analyze each of the resorts in two parts -- a sub-parcel for the 14 September 2007 lodge or inn, and sub-parcels for the resort-housing lots. Our analysis suggests that the incremental value of Plan approval (relative to baseline) for the Big Moose Mountain Resort might range from approximately $5 to $10 million. This increment largely reflects the assumed HBU values associated with larger lot sizes under the concept plan combined with a greater proportion of standalone resort house lots relative to the baseline. Lily Bay potentially could generate between $10 million and $15 million in incremental value if the Plan is approved. Although Lily Bay is much smaller than Big Moose Mountain in both acreage and accommodations, we give it a greater amenity premium consistent with its potential positioning as a higher-end resort. Because of the limited detail provided in the plan, and greater uncertainty inherent in resort development, these results should be interpreted cautiously. Overall, the incremental increase in value associated with plan approval can be interpreted as the maximum reduction in development cash flows that Plum Creek might be willing to accept (e.g., in the form of scaled-back development or increased conservation commitments) before giving up on the concept plan and reverting to a baseline-type strategy. It provides a useful context for considering the financial value of conservation commitments that Plum Creek has proposed under its concept plan, which are explored in OSI’s third discussion paper. Along with a detailed discussion of the model, including several important caveats regarding the assumptions employed in its development, this paper contains technical appendices describing the model inputs and calculations, and full print-outs of the modeled scenarios. 2 14 September 2007 INTRODUCTION In April 2007, Plum Creek Timber Company submitted a revised application to Maine's Land Use Regulation Commission (LURC) for a comprehensive development project on 408,000 acres in the Moosehead Lake region. The proposed development includes as many as 975 house lots and the potential for two resorts allowing for an additional 1,050 accommodations. According to the company's LURC application, conservation measures proposed as balance for development will result in 90,000 acres of permanent conservation including 156 miles of lake and pond frontage, 10 miles of Moose River frontage and 153 miles of permanent trail easements within the Plan Area. Plan approval will provide the opportunity, through the Conservation Framework, to secure another 266,000-acre conservation easement, a 29,500-acre conservation sale, both within the Plan Area, and a 45,000-acre fee sale outside the Plan Area for permanent conservation. According to the company, when the Plan and Conservation Framework are fully implemented, 195 miles of permanent lake and pond shorefront conservation will be in place (as well as 32 miles of frontage on the Moose River), and 70 lakes and ponds will be protected in perpetuity. Because of its ambitious scale and potential impacts, Plum Creek's proposal has engendered significant public debate. The Open Space Institute (OSI) -- sponsor of the Northern Forest Protection Fund (NFPF), a group that has provided substantial financial support for Maine conservation purchases in recent years -is committed to playing a constructive role in helping to inform decisions about the future development in the Moosehead region. OSI's goal is to develop rigorous and objective analysis of Plum Creek's proposals and alternatives put forward by other stakeholders. OSI has collaborated with the Margaret Chase Smith Policy Center at the University of Maine in Orono and with Industrial Economics, Inc. (IEc) of Cambridge, Massachusetts to provide tools and analyses informing the public policy debate on Plum Creek's proposal. The Smith Center has assisted with technical review of the work and sponsoring outreach activities. IEc has been responsible for the design and implementation of the financial model, with land use planning support from Jerry Bley of Creative Conservation, LLC and real estate appraisal from Michael Moniz (Certified General Appraiser). This discussion paper, the second in OSI's series addressing the Plum Creek proposal, describes the Plum Creek financial model. The model provides a framework for analyzing the financial value to Plum Creek's shareholders of the company's Moosehead development proposal as well as possible alternatives. OSI's intent is to inform the public debate and give participants in the ongoing development-review process a tool for evaluating whether the company’s plan and alternative proposals can yield financially attractive returns for Plum Creek shareholders while providing conservation and other outcomes that might be important to stakeholders. The remainder of this paper is organized into four sections: • • • • Plum Creek's Concept Plan Proposal -- provides a brief overview of the proposed development. Financial Modeling Approach -- describes the overall framework, key inputs and outputs, and model caveats, limitations and uncertainties. Modeled Scenarios and Results -- illustrates the modeling approach through analysis of the financial impacts of the April 2007 proposed plan. Model Sensitivity -- demonstrates the impact on model results from varying several of the key input assumptions. 3 14 September 2007 PLUM CREEK'S CONCEPT PLAN PROPOSAL In April 2007, Plum Creek submitted to LURC a revised application (its third version) for its proposed Moosehead development plan.1 The primary element of the plan is a proposal to create up to 975 residential lots. The development would be comprised of 236 shorefront lots and 739 back lots located around six lakes and ponds. The development proposal also includes: • • • two resorts -- one near Big Moose Mountain and one at Lily Bay, with up to an additional 1,050 accommodations, an uncertain number of which could be house, townhouse or condominium lots; a donation of up to 100 acres for affordable housing in the region, plus 5 acres for the Beaver Cove town office use; and a Community Fund financed through Plum Creek donations linked to development lot sales. As previously described, Plum Creek's plan also proposes a variety of conservation measures that would result in 90,000 acres of permanent conservation, including 156 miles of lake and pond frontage, 10 miles of Moose River frontage and 153 miles of permanent trail easements within the Plan Area. Plum Creek also has linked approval from LURC of a concept plan to a binding agreement to sell to The Nature Conservancy and its partners $35 million worth of land and working forest easements on an additional 340,500 acres in and near the region.2 FINANCIAL MODELING APPROACH Overview A key premise driving our development of the Plum Creek financial model is that any realistic alternative plan must provide Plum Creek with a fair and reasonable return on its investments in Maine. Only by examining the financial implications of alternative proposals can stakeholders hope to assure outcomes that satisfy a broad range of both public and private interests. The model presented here is designed to provide a flexible framework for estimating financial returns to Plum Creek under alternative development scenarios. In our view, the starting point for any such analysis is an evaluation of a baseline development scenario. This scenario is an attempt to understand how Plum Creek could proceed in the absence of LURC concept plan approval for a Moosehead project. This baseline is the reference point that we think Plum Creek would most likely use when evaluating possible modifications to its April 2007 proposal. From a financial perspective, when changes to the concept plan proposal result in significantly lower financial returns relative to the baseline (or some other reasonably likely scenario), we would expect Plum Creek to abandon the large concept plan approach in favor of the less ambitious, more piecemeal baseline approach. Baseline issues are addressed in Discussion Paper No. 1 in detail.3 While it is obviously difficult to predict exactly what Plum Creek would be able to develop absent approval of a concept plan, the primary conclusion of the Discussion Paper No. 1 is that, barring changes in Maine's existing land 1 2 See http://www.maine.gov/doc/lurc/reference/resourceplans/moosehead.html. See Discussion Paper No. 3: "ANALYSIS OF CONSERVATION COMMITMENTS IN PLUM CREEK'S MOOSEHEAD LAKE CONCEPT PLAN" (hereafter "Discussion Paper No. 3"). See Discussion Paper No. 1: "BASELINE DEVELOPMENT SCENARIO FOR THE PLUM CREEK MOOSEHEAD PROJECT LANDS" (hereafter "Discussion Paper No. 1"). 3 4 14 September 2007 use regulations, Plum Creek would likely be able to proceed with a significant amount of development even were it to abandon its efforts to gain LURC approval for a large-scale concept plan. The foundation of the Plum Creek financial model is discounted cash flow (DCF) analysis. The DCF approach is the most widely used and accepted method for evaluating the types of real estate transactions being contemplated by Plum Creek, and we assume that Plum Creek itself uses DCF to evaluate the returns to its development proposals in Maine and elsewhere. The output from the DCF approach is an estimate of the value today of a stream of real estate development costs and property sale revenues stretching out into the future. This method allows consistent comparison of alternative deals that have different time profiles for revenues (cash inflows) and costs (cash outflows). Our model, implemented using Excel™, reflects Plum Creek's organization as a real estate investment trust (REIT). For a timber and real estate development company, REIT status confers a variety of tax benefits that can improve financial returns to the shareholders relative to other forms of corporate organization. In practice, however, Plum Creek's REIT status raises a variety of complex tax and operating issues that we need to address in the model in order to provide an accurate estimate of shareholder returns. These issues include U.S. Internal Revenue Service (IRS) requirements that lands being sold for development by Plum Creek be transferred into a taxable REIT subsidiary (TRS) where income is first taxed at the corporate level (35 percent) before being distributed to the shareholders as dividends. Our model also takes into account Maine state taxes as well as penalties associated with removing lands from Maine's Tree Growth Tax program. In calculating returns to Plum Creek shareholders, the model considers three revenue categories potentially associated with the development proposal. These include land sales and donations by Plum Creek for: • • • residential lots; resorts and associated lots; and affordable housing. Overall, the model provides a flexible framework for evaluating a range of factors that determine the value of development to Plum Creek. These factors include the location, scale and build-out schedule for development, infrastructure and development costs, sales prices for properties, as well as financial variables such as tax rates and Plum Creek's cost of capital. We also think it important to be clear about what our Plum Creek financial model will not do. We are aware that economic impacts of development, both positive and negative, are a key issue for many stakeholders. Community and conservation groups with whom we have talked stress the importance of considering the economic benefits of alternative development strategies as compared with the Plum Creek plan. We fully recognize the importance of such analyses. Nonetheless, because of resource constraints, we have not been able to address these issues. In fact, such analyses would call for a regional economic modeling approach very different from the financial-analytical approach discussed here. Financial Model: Key Inputs, Assumptions and Outputs To estimate the discounted cash flows from residential development, users of the Plum Creek financial model need to enter data and assumptions for a wide array of variables. These variables fall into three main categories: development-specific parameters, Plum Creek corporate parameters, and external financial and economic factors. Each of these is discussed below. Development Scenario Characteristics 5 14 September 2007 For financial modeling purposes, four primary types of information characterize a development scenario. First, the location of the development should be specified; for each scenario, the number and size of shorefront and back lots must be defined. Second, a development schedule is needed to specify the number of lots expected to be sold in each year, recognizing that sales further in the future will be worth less today to Plum Creek than immediate sales. Third, estimates of current sale prices are needed for the various lot types by size and location, as well as property value appreciation rates over time. Fourth, the model provides an option for varying the costs of permitting, development infrastructure (e.g., roads, utilities) and marketing. For non-residential development, such as portions of the proposed resort properties, similar information is needed. Corporate Financial and Tax Assumptions The value of the discounted cash flows from development property sales is an inverse function of Plum Creek's corporate cost of capital. We developed our estimate for this measure using a weighted-averagecost-of-capital approach that considers Plum Creek's mix of equity and debt financing. This rate provides a measure of how investors value returns to Plum Creek and is the rate used in all discounting calculations in the model. The estimated rate is 7.0 percent.4 Another factor affecting financial returns is the price used by Plum Creek for internal transfers from the REIT to the real estate TRS. Generally, the higher this internal transfer price, the lower the tax liability on the income received by Plum Creek shareholders. We currently employ a method for calculating the transfer price that applies a scalar to the final sales prices; for the analysis in this paper the scalar is set at 30 percent of current higher-and-better-use (HBU) prices.5 The model also assumes that development results in a negative financial impact to Plum Creek's cash flows in that the company can no longer harvest timber from a portion of the lands sold for development. We calculate this cash flow offset using county estimates of average timber values per acre and lot-type specific estimates of the portion of harvestable timber acreage lost. External Economic Factors The model requires several assumptions about external economic conditions. An estimate of the future level of annual price inflation in the Maine economy (assumed to be 3 percent) is used to project future changes in permitting, development and infrastructure costs. We also make the assumption that state and federal tax rates on capital gains and ordinary income will not change in the future. Finally, land that is removed from preferential property tax treatment under Maine's Tree Growth Tax program will be subject to penalties, which we calculate based on estimates of the assessed property values at the time of removal for the lands in question. In our discussions with Plum Creek representatives, they indicated their belief that a project-specific discount rate in the range of 15-20 percent would be more appropriate for this analysis given the risk of real estate development. We have discussed this matter at length with a variety of industry experts and investment analysts who cover timber REITs. While we understand the theoretical concepts behind the higher discount rates proposed, our research into the market-based reality indicated that Plum Creek shareholders have not imposed any greater total equity return expectations on the margin for the company's various HBU development pursuits. As a result, we believe that employing the company's overall weighted average cost of capital is a reasonable approach for our modeling purposes. That said, we have performed sensitivity analyses on the model specifically employing higher discount rates and we present the results in the last section of this paper. We have anecdotal data indicating that this transfer-price scalar value is an appropriate estimate. We have requested additional information from Plum Creek, but as of this writing have not received any more detail. 5 4 6 14 September 2007 Assumptions for Resorts Plum Creek's April 2007 plan proposes the development of two resort areas: a large resort near Big Moose Mountain (covering approximately 4,400 acres, with accommodations for 800); and a smaller one at Lily Bay (covering an 800-acre parcel, with accommodations for 250). Based on our review of the resort descriptions outlined in the plan documents and related submissions, as well as our discussions with various experts in the resort-development field, we expect that if the plan is approved, Plum Creek will develop not only the lodge or inn facilities associated with each resort area, but also residential accommodations including single family homes, townhouses and condominiums within the re-zoned parcels. As a result, we analyze each of the resorts in two parts -- a sub-parcel for the lodge, with the land priced at a moderate premium to large blocks sold for other non-residential uses; and sub-parcels for the resort housing lots, priced at a premium over those analyzed elsewhere in the model to reflect the added amenity value of the associated resorts. Plum Creek has provided much less information on the resorts than it has on its other development components. Thus, our analysis of this plan element is more speculative than that for the rest of the plan. Nonetheless, the assumptions we employ are not inconsistent with Plum Creek's proposals and our analysis should provide insight into the possible financial returns associated with resort area development. Model Results The model results are presented in a table that summarizes major input assumptions and the overall discounted cash flows associated with the development scenario. These cash flows are reported separately for each major component of the development revenue streams, and are summed to show the DCF value of the modeled scenario as a whole. MODELED SCENARIOS AND RESULTS To estimate the financial implications for Plum Creek of LURC concept plan approval, we compared the financial returns to Plum Creek shareholders under a projected baseline development scenario with those expected from Plum Creek's April 2007 proposed plan. This section describes the model input assumptions for the two scenarios, reports the results of the analyses, and discusses key uncertainties inherent to the model structure and the input assumptions. Baseline Development Scenario The baseline scenario represents our estimate of what Plum Creek would be able to develop absent the kind of rezoning process envisioned in the region-wide concept plan.6 For the residential lots, the baseline assumes that Plum Creek could create lots using existing zoning rules without triggering legislative or regulatory changes restricting the use of such an approach. We also assume in the baseline that Plum Creek would likely be able to develop an 800-unit resort at Big Moose Mountain, with 200 accommodations in a lodge, approximately 25 percent of the accommodations (200 units) being individual home sites and the balance (400 units) being multi-unit dwellings (i.e., townhouses and condominiums). Finally, the baseline assumes that Plum Creek would not donate either the affordable housing site or the Community Fund. 6 See Discussion Paper No. 1 for more detail. 7 14 September 2007 The baseline development scenario requires numerous assumptions, many of which are quite uncertain and have important implications for financial returns to Plum Creek's shareholders. A primary value of the model is that it allows testing of the implications of alternative assumptions. Under any reasonable set of assumptions, however, our analysis suggests that the most important drivers of financial returns in any scenario are the number and prices of residential lots, their location, and the timing of sales. Based on our analysis, we estimate that under current zoning regulations (i.e., two-infive; adjacency; and management class 3 lakes and level 2 subdivisions), Plum Creek could develop on the order of 618 residential lots, and the sales of those lots would be spread fairly evenly over the first 20 years of a 30-year development period, with a slight slowdown thereafter. As covered in Discussion Paper No. 1, we estimate that 252 of these lots would be shorefront lots (216 2-acre lots and 36 100-acre kingdom lots) and the remainder would be back lots (3-acre lots with scenic views or other amenities). We have not assumed that Plum Creek would develop every legally feasible back lot because of high infrastructure costs, low lot values, and/or potential interference with ongoing forestry activities. Instead, we have attempted to identify the most profitable lot configuration. For example, we included a smaller number of high-end “kingdom lots” (valued at about five times the value of a premium shorefront lot) rather than a larger number of shorefront lots. Table 1 summarizes our assumptions in the baseline development scenario about the location, size and number of residential lots, and the estimated timing of lot sales. This table does not include lots associated with the Big Moose Mountain resort. A full printout of the baseline development model input assumptions and output, including the resort analysis, is included as Appendix A-1. Table 1 BASELINE DEVELOPMENT LOT SUMMARY Financial Model Input Assumptions Total # of Lot Sale Timing (# of lots by year) Lots 2008 2013 2018 2023 2028 2033 24 192 36 252 177 189 366 618 5,032 Lot Location/ Lake or Pond Shorefront (2 acres): - Moosehead - Other Lakes/Ponds Kingdom Lots (100 acres) Subtotal Shorefront Lots Back Lots (3 acres): - Premium7 - Quality8 Subtotal Back Lots TOTAL LOTS Total Acres 2038 0 0 5 5 0 0 0 5 500 10 77 5 92 11 38 49 141 821 12 40 5 57 47 48 95 152 889 4 29 5 38 41 45 86 124 824 0 32 5 37 30 20 50 87 714 0 13 5 18 35 32 67 85 727 0 0 5 5 13 6 19 24 557 To estimate current prices for residential lots, we collected data on recent comparable second home lot sales in the Moosehead area and generated valuations for the lot types used in our analysis. Generally, this analysis indicates that shorefront lots on Moosehead Lake will have the highest sale prices and back lots at other locations will have the lowest prices. Table 2 shows a summary of the lot prices we use. "Premium" back lots are those assumed to be near the Moosehead-Greenville area, with deeded water access and nice views. 8 7 "Quality" back lots are those with two out of three "Premium" criteria. 8 14 September 2007 Table 2 RESIDENTIAL LOT PRICES (2007 Dollars) Financial Model Input Assumptions DEVELOPMENT LOTS Lot Description "Lot Type" Large Lot HBU Appraisal Lot price Lot size (acres) $/Acre Baseline Build-Out 10 Lot Price Lot Size $/Acre Concept Plan 11 Lot price Lot size (acres) $/Acre $200,000 2.5 $80,000 $152,500 2.5 $61,000 RESORT LOTS12 Baseline Build-Out: Big Moose Mountain (with premium) Lot Price $210,000 $156,000 20% Lot Size 2 2 $/Acre $105,000 $78,000 Concept Plan: Big Moose Mountain (with premium) Lot price $240,000 $183,000 20% Lot size (acres) 2.5 2.5 $/Acre $96,000 $73,200 Lily Bay (with premium) Lot price $300,000 $228,750 50% Lot size (acres) 2.5 2.5 $/Acre $120,000 $91,500 $114,000 3 $38,000 n/a $107,500 4 $26,875 $72,500 4 $18,125 $42,500 4 $10,625 $175,000 2 $87,500 $130,000 2 $65,000 $575,000 100 $5,750 $95,000 3 $31,667 $65,000 3 $21,667 $35,000 3 $11,667 9 Moosehead 1 $225,000 3 $75,000 Shorefront Other Lakes/Ponds 1 $175,000 3 $58,333 Back Lot Kingdom Lots Premium 2 $575,000 100 $5,750 $120,000 5 $24,000 Quality 3 $80,000 5 $16,000 Other 4 $50,000 5 $10,000 n/a n/a n/a n/a $129,000 4 $32,250 $161,250 4 $40,313 n/a n/a n/a n/a n/a To estimate sale prices for future years, we assume that all lots appreciate in value at 3 percent per year for the first 5 years, and then 4 percent per year thereafter.13 9 Information obtained from Michael Moniz (Certified General Appraiser). Baseline lot prices assume, on average, a 25-percent discount off of the Large Lot prices shown. 10 11 Concept Plan lot prices assume that half of the proposed lots will be Large Lots and half will be standard (i.e., sized like those in the Baseline). 12 13 Resort lots are priced at the premia shown in the leftmost column. As with other key model input parameters, we perform sensitivity analysis on this assumption at the end of this paper. 9 14 September 2007 To place an economic value on the baseline scenario, we also need to estimate a variety of other economic parameters. One important element is the price at which Plum Creek transfers development land into its real estate TRS. We estimate the transfer price by assuming that the company treats the land as a wholesale block transfer at un-entitled land values. To effect this assumption, we apply an adjustment scalar to the year-zero appraised HBU price. As a hypothetical example, for a shorefront lot on Moosehead Lake with a 2013 projected sales price of $260,837, the transfer price in 2008 would be $67,500 (see Table 3). This transfer price reflects the removal of the 3-percent per year appreciation in property value that we assume occurs while the shorefront land is held in the TRS prior to sale, with the result (i.e., the $225,000 HBU value in 2008) multiplied by the 30-percent adjustment scalar. For permitting, development and marketing (PD&M) costs, we apply a percentage factor to the HBU appraisal at the time of the land sale. Based on our discussions with several reviewers of earlier model versions, we estimate for the baseline scenario that PD&M costs will be 25 percent of the HBU price for each lot sold. We also need to correctly account for a variety of taxes in order to estimate Plum Creek shareholder returns. Almost all land transfers are taxed by Maine. This Land Transfer Tax is calculated based on the transfer (or sale) price, and is split between the seller (i.e., the REIT, in this example) and the buyer (i.e., the TRS, in this example).14 For purposes of estimating cash flows, both Maine State and IRS rules permit Plum Creek to defer the related taxes until the year when the lot is sold (2013 in this example). In addition, the Tree Growth Tax penalty is estimated as 20 percent of the difference between the valuation of the land for timber and the assessed value of the land when it is sold (with assessed values being much lower than HBU sales values in almost all cases; in this example our assumption is that the assessed value would be 40 percent of the HBU value). Finally, we similarly compute the taxable income at the TRS based on the HBU sale to a developer after subtracting all of the Maine taxes (including the Tree Growth Tax penalty) and the development and infrastructure expenses, with the balance being taxed at the marginal Federal rate (35 percent). Table 3 shows a schematic of the various tax and cash flow calculations. The Land Transfer Tax is calculated by rounding the transfer/sales price up to the nearest $500.00, then dividing by $500.00, and then multiplying the result by $2.20. The resulting tax is split 50/50 between the seller and the buyer. 14 10 14 September 2007 Table 3 EXAMPLE REIT LAND TRANSFER AND SALE 3-Acre Shorefront Lot on Moosehead Lake Annual Inflation $600 1998 REIT Acquisition Cost Trees Land HBU Price: $225,000 2008 3% $260,837 2013 =====> Transfer to TRS (installment sale) =====> HBU Sale to Developer $375 $260,462 $260,837 ($65,209) Total PD&M 30% $375 $225 $600 $375 $224,625 $225,000 ($56,250) $67,500 3% Transfer price (HBU Price * scalar) CASH FLOWS: HBU sales revenues Assessed value of property (% of HBU value) $104,335 40% 20% Maine tree growth tax penalty (based on assessed value change) 23% Lost stumpage value REIT portion of Maine transfer tax on REIT->TRS transfer (deferred) 25.00% P,D&M expenses (at TRS) TRS portion of Maine transfer tax on REIT->TRS transfer (deferred) TRS portion of Maine transfer tax on TRS HBU sale Taxable TRS income for Maine income taxes $106,658 8.93% Maine income tax on TRS HBU sale TRS Ordinary income taxable at Federal rate $97,133 Federal income taxes 35.00% TRS note repayment to REIT TRS Dividend Cash Flows to S/Hs REIT Dividend Cash Flows to S/Hs TOTAL DIVIDEND CASH FLOWS 7.00% $260,837 ($20,747) ($88) ($149) ($65,209) ($149) ($574) ($9,525) ($33,997) ($67,500) $62,900 $67,500 $130,400 $81,207 Discount Rate DCF Value Total Cash Flows (2007) Applying this approach to the baseline build-out and the resort-potential property near Big Moose Mountain, we estimate annual after-tax cash flows to Plum Creek in the baseline scenario going 30 years into the future. We assume that Plum Creek uses a discount rate of 7 percent, the company's weighted average cost of capital, to estimate the present value of these cash flows. Baseline Scenario Results Using the assumptions described above, we estimate that the baseline scenario would yield discounted cash flows of $67.3 million (in 2007 dollars) for Plum Creek's shareholders. In laymen's terms, this DCF value can be thought of as the after-tax amount someone would need to pay Plum Creek's shareholders 11 14 September 2007 today to convince them to forego development altogether if the baseline scenario is the most profitable alternative development option reasonably available to the company. The DCF values for the components of the baseline development plan are summarized in Table 4, which replicates the related summary page from the model. Table 4 PLUM CREEK DEVELOPMENT in MOOSEHEAD LAKE REGION Contribution to Shareholder Value in 2007 Dollars BASELINE DEVELOPMENT SCENARIO Property Sales Residential Development Lots Resort Property Descriptions BASELINE DEVELOPMENT SCENARIO DCF Value $31.0 $36.3 $67.3 BIG MOOSE MOUNTAIN LODGE and RESORT LOTS TOTAL ($ millions) Plum Creek April 2007 Proposed Concept Plan The primary differences between the baseline development scenario and Plum Creek's April 2007 proposal relate to the number, size and location of residential lots, and the development of a second resort in Lily Bay. The April 2007 plan proposes construction of 975 residential lots located around the six lakes and ponds targeted for development by Plum Creek. Plum Creek did not indicate how rapidly development would occur at each location so, absent other information, we assume that the residential lot development begins in year 0 and occurs at a uniform rate over 15 years at each of the proposed locations.15 The complete schedule is shown in Appendix A-2. Sale prices differ from those in the baseline scenario to reflect the larger lot sizes proposed by Plum Creek, and were summarized earlier in Table 2. Discounting assumptions are the same as in the baseline scenario. Based on input we received from several model reviewers, however, we use different assumptions than in the baseline for the PD&M costs and the assessed value estimates for purposes of Tree Growth tax penalties.16 As stated earlier, our view is that with any resort development, Plum Creek will include a variety of residential lots (some shorefront and some back lots) within its resort areas. In its resort proposal, Plum Creek has not limited the number of individual house lots, or the size of these lots, resulting in more land being allocated to such development then in the baseline (approximately 570 more acres). Table 5 below summarizes our model input assumptions for the two resorts. This development schedule assumption results in sales of 65 lots per year over a 15-year period, not including resort lots. We have not made any effort at this point to verify whether the local real estate markets are positioned to absorb development at this level, and we test the sensitivity of the schedule assumption later in this paper. For the proposed plan scenario, we estimate that PD&M costs will be 20 percent of the HBU price of each lot sold. We use a lower percentage here than in the baseline with the assumption that concept plan approval will remove many of the hurdles associated with piecemeal, lot-by-lot permit applications. Thus, the permitting-cost component of the PD&M cost estimate for the proposed plan is lower than that for the baseline. In the proposed plan scenario, we assume that property will be transferred out of Tree Growth classification in large blocks, and will be assessed at 25 percent of HBU values. In contrast, we assume that the transferred parcels will be much smaller in the baseline, leading to higher per-acre assessments, and we use an assumption of 40 percent of HBU values for our assessed values in the baseline. 16 15 12 14 September 2007 Table 5 PLUM CREEK CONCEPT PLAN PROPOSED RESORTS Financial Model Input Assumptions Average HBU/Acre17 Resort (Development Years) Accommodations/Units Acres (2007 Dollars) Big Moose Mountain 800 4.446 200 2,136 $5,000 - Lodge/Inn (2009) 600 2,310 $35,378 - Residential Lots (2009-2014) Lily Bay - Lodge/Inn (2011) - Residential Lots (2009-2014) 250 100 150 777 77 700 $5,000 $47,303 Finally, we analyze the costs to Plum Creek that result from donating the affordable housing parcel in Greenville, as well as building the Community Fund outlined in the plan.18 April 2007 Concept Plan Scenario Results Under the assumptions outlined above, we estimate that the April 2007 plan scenario would yield discounted cash flows of $106.5 million for Plum Creek's shareholders, which represents the total value to Plum Creek of development if the concept plan permit is approved. The results indicate that concept plan approval yields a DCF value for the residential lot development of $50.5 million, compared with $31.0 million in the baseline development scenario, reflecting more and higher-valued lot sales taking place sooner relative to the baseline. Figure 1 shows a comparison of cumulative DCF values over time between the baseline and concept plan for residential lot (the data underlying this chart are detailed in Appendices A-1 and A-2). 19 Moreover, the second resort at Lily Bay and the more expansive Big Moose Mountain resort configuration in the concept plan would add $13.0 million and $6.7 million, respectively, to Plum Creek's incremental returns. The DCF values for the components of the April 2007 development proposal are summarized in Table 6. Interestingly, while the concept plan would allow Plum Creek to develop more residential lots and yield higher returns for the company's shareholders relative to the next best alternative, it would do so with the development of about the same residential-lot acreage as in the baseline during the first 15 years, and fewer residential-lot acres overall during the 30-year plan period. This is largely due to the inclusion of 100-acre kingdom lots in the baseline development scenario. In reality, only a few acres on each of these lots will actually be developed in the foreseeable future. 17 18 The Average HBU/Acre figures are weighted averages. See Table 2 for underlying price detail. We calculate the cost of the Community Fund directly as a function of the residential lot sales, as specified in the plan documents. See Appendix A-2 for full details. It is important to note here that the residential lot development comparisons we have performed and charted in Figure 1 do not include the residential back lots that might be developed within Plum Creek's proposed resort areas. We have excluded those lots from our comparisons because the assumptions around them are significantly more uncertain than the detailed lot development defined elsewhere in the concept plan proposal. Including the resort-area residential lots as we have roughly modeled them in our analysis would change the comparisons shown in Figure 1. 19 13 14 September 2007 Figure 1 COMPARATIVE RESIDENTIAL-LOT DEVELOPMENT CASH FLOWS Contribution to Shareholder Value in 2007 Dollars RESIDENTIAL LOT CASH FLOWS $60 $50 Cumulative DCF ($mm) $40 $30 $20 $10 $0 2008 2010 2012 2014 2016 2018 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 Baseline PC Concept Plan Table 6 PLUM CREEK DEVELOPMENT in MOOSEHEAD LAKE REGION Contribution to Shareholder Value in 2007 Dollars APRIL 2007 CONCEPT PLAN Property Sales Residential Development Lots Resort #1 Resort #2 Affordable Housing Property Descriptions APRIL 2007 CONCEPT PLAN BIG MOOSE MOUNTAIN LODGE and RESORT LOTS DCF Value $50.5 $43.0 $13.0 ($0.0) $106.5 LILY BAY RESORT INN and RESORT LOTS 100-ACRE SITE DONATION TOTAL ($ millions) Incremental Value of Concept Plan Approval Under the assumption that the baseline development scenario for Plum Creek's lands around Moosehead Lake represents the company's next-best alternative to the April 2007 proposed concept plan, the value to Plum Creek of concept plan approval would be $39.2 million, which is the difference between the DCF value for the April 2007 proposal and that of the baseline build-out scenario (i.e., $106.5 - $67.3 million). 14 14 September 2007 One interpretation of this incremental value is that it represents the maximum reduction in development cash flows that Plum Creek might be willing to accept (e.g., in the form of scaled-back development or increased conservation commitments) before giving up on the concept plan and reverting to a baselinetype strategy. This analysis does not include the financial value of Plum Creek’s proposed conservation commitments under the concept plan (“conservation for balance”). A third discussion paper assesses the financial value the company is likely to forgo from the donation of easements on 90,000 acres of land.20 MODEL SENSITIVITY The modeled scenarios presented above make clear that using the Plum Creek financial model requires numerous assumptions about the nature and scale of development, real estate prices, taxes, and more general economic conditions for many years in the future. We have researched values for the core inputs to the model, such as land prices, development costs, taxes, inflation, and Plum Creek's cost of capital. Nonetheless, one of the most important factors, the schedule for selling properties over the next 30 years, remains highly speculative. There is no crystal ball (or existing study) that offers much certainty about this dimension. Our purpose in developing the model is not to provide a definitive value for LURC assessment of the Plum Creek concept plan proposal; rather it is to give stakeholders a tool for understanding the financial implications to Plum Creek of different development schemes. As such, we have encouraged interested stakeholders to become knowledgeable about the model and its inputs, and to develop their own scenarios for analysis. To illustrate the impact that some of the key inputs have on the incremental value of concept-plan approval, we performed a variety of sensitivity runs on the modeled scenarios. To assess the impact of a change in baseline development scenario lot numbers, we ran that scenario assuming both 20-percent greater and 20-percent fewer lots.21 The results of these lot-number sensitivity runs are shown in Table 7. Table 7 VALUE OF APRIL 2007 CONCEPT PLAN PROPOSAL Contribution to Shareholder Value in 2007 Dollars ($mm) Baseline Assumption High (+ 20%) Mid-Point Low (- 20%) # of Lots 742 618 494 Baseline Development Scenario DCF Value $73.5 $67.3 $61.1 April 2007 Concept Plan DCF Value Value of April 2007 Concept Plan $33.0 $106.5 $39.2 $45.4 20 21 See Discussion Paper No. 3. These two sensitivity runs were purely mathematical in nature in that they did not make any specific assumptions about the locations of the lots added or subtracted. 15 14 September 2007 Similarly, we performed several sensitivity runs on both the baseline and the April 2007 plan scenarios to determine the effect that changes in the most important input assumptions would have on the incremental value of the plan. Table 8 summarizes the results. Independently varying shorefront HBU price inflation, HBU price appraisals and PD&M costs upward and downward over ranges that are quite possible, results in variation of the plan’s value of between approximately 9 to 12 percent in both directions. As can also be seen, a higher discount rate decreases the overall DCF value of any scenario but increases the incremental benefits of more rapid development.22 Table 8 INCREMENTAL VALUE OF PLAN SENSITIVITY TO OTHER KEY INPUTS Contribution to Shareholder Value in 2007 Dollars Input Value DCF ($mm) ∆ from Modeled Estimate ($mm) ($4.1) $4.5 ($3.9) $3.9 ($3.6) $3.6 ($9.8) $6.3 ($13.4) $10.6 % -10.5% 11.5% -10.0% 10.0% -9.1% 9.1% -25.0% 16.0% -41.9% 33.2% HBU Inflation (yrs 1-5) Low 0% $35.1 As Modeled 3% $39.2 High 6% $43.7 HBU Prices (vs. current appraisal) Low -10% $35.3 As Modeled Current $39.2 High +10% $43.1 PD&M Costs (% HBU) Low +10% $35.6 As Modeled (Baseline/Plan) 25% / 20% $39.2 High -10% $42.7 Concept Plan Development Schedule (years) assuming 7% Discount Rate Low 30 $29.4 As Modeled 15 $39.2 High 8 $45.4 Concept Plan Development Schedule (years) assuming 12% Discount Rate Low 30 $18.5 As Modeled High 15 8 $31.9 $42.4 Accelerating residential-lot development (i.e., not including resort-related lots) from 15 to 8 years using the 12-percent rate yields an increase in the value of plan approval of $10.6 million. If the analysis is run using Plum Creek's recommended discount rates ranging from 15 to 20 percent, the benefits of accelerated development become even more significant. Using a 15-percent discount rate, accelerating development from 15 to 8 years results in an $11.9 million increase in the value of plan approval; using a 20-percent discount rate increases this benefit to $12.8 million. 22 16 14 September 2007 Finally, as mentioned earlier, Plum Creek has provided much less information on the resorts than it has on its other proposed development pieces. We have relied on a variety of sources to support our preliminary resort scenarios described above, some as filed with LURC and others gained from external research and conversations with resort-industry personnel. Those sources notwithstanding, we want to emphasize that our analysis of these development components contains far more uncertainty than that for the rest of the plan. As a result, we conducted sensitivity analysis on a number of key variables that affect valuation of the resorts, as presented in Table 9. As with the residential development lots, the resort values are sensitive to changes in HBU prices and the development schedule, and highly sensitive to changes in the discount rate. Table 9 RESORT-VALUE SENSITIVITY TO KEY INPUTS Contribution to Shareholder Value in 2007 Dollars Baseline Input Value Concept Plan Lily Bay Total ∆ from Modeled Estimate Incremental Resort Value in ($mm) % Concept Plan $18.0 $19.7 $21.4 $17.7 $19.7 $21.7 $17.6 $19.7 $21.8 $23.9 $19.2 $19.7 $22.3 $19.7 $16.5 $14.0 $19.7 $17.5 $16.2 $14.6 $12.4 ($1.7) $1.7 ($2.0) $2.0 ($2.1) $2.1 $4.2 ($0.5) $2.6 ($3.2) ($5.7) ($2.2) ($3.5) ($5.1) ($7.3) -8.7% 8.7% -10.0% 10.0% -10.7% 10.7% 21.4% -2.5% 13.1% -16.3% -28.9% -11.2% -17.6% -26.0% -37.2% Big Moose Big Moose Mtn Mtn # Resort Accommodations as Lots (BMM/LB) -10% 540/135 $33.6 $39.7 As Modeled 600/150 $36.3 $43.0 10% 660/165 $39.0 $46.3 HBU Prices (vs. current appraisal) Low -10% $32.6 $38.6 As Modeled Current $36.3 $43.0 High +10% $40.0 $47.3 Lily Bay Lot Premium (% over Concept Plan Lots) Low 25% $36.3 $43.0 As Modeled 50% $36.3 $43.0 High 75% $36.3 $43.0 Very High 100% $36.3 $43.0 Resort Lot Development Schedule (# years) Low 16 $32.6 $40.3 As Modeled 8 $36.3 $43.0 High 4 $37.8 $46.0 Resort Lot Development Start (year) As Modeled 2009 $36.3 $43.0 5 years out 2014 $30.2 $35.8 10 years out 2019 $25.7 $30.5 Discount Rate As Modeled 7% $36.3 $43.0 10% $32.5 $38.4 12% $30.3 $35.7 Higher Range 15% $27.4 $32.3 20% $23.4 $27.6 $11.9 $13.0 $14.2 $11.7 $13.0 $14.3 $10.9 $13.0 $15.1 $17.2 $11.5 $13.0 $14.0 $13.0 $10.9 $9.3 $13.0 $11.6 $10.7 $9.7 $8.2 17 14 September 2007 OSI continues to be available to work closely with stakeholders who would like to analyze alternative development scenarios to help inform the public's views on Plum Creek's proposal. For each requested scenario, we will provide stakeholders with a template for the required model inputs, and will use those inputs to estimate DCF results like those shown in this paper, including detailed model output tables similar to those in the appendices. We intend to make these results publicly available along with our own discussion and analysis of the input assumptions and results. For further information, please contact: Ann Czerwonka, Industrial Economics, Incorporated -- 617 354 0074 Peter Howell, Open Space Institute -- 212 290 8200 18