Forestar Real Estate Group FOR
March 26, 2008 - 11:02pm EST by
lvampa1070
2008 2009
Price: 23.81 EPS
Shares Out. (in M): 0 P/E
Market Cap (in $M): 845 P/FCF
Net Debt (in $M): 0 EBIT 0 0
TEV ($): 0 TEV/EBIT

Sign up for free guest access to view investment idea with a 45 days delay.

Description

Forestar Real Estate Group, ticker (FOR), trades at 70-75% of fair value because it is neglected and misunderstood.  
  
FOR is overlooked by investors.  The company was spun-off f rom Temple-Inland (TIN) at the end of December.  Afterwards, S&P 500 index funds that owned TIN and received a stock dividend of FOR could be expected to sell FOR because it is not in the S&P 500.  Index funds owned 17% of TIN.  In addition, Temple was a mid-cap stock with a market cap of $5 billion, while Forestar is a small cap with a $850 million market value.  The sell at any price response by the index funds has nothing to do with the value of Forestar and presents a buying opportunity.  

Forestar is misunderstood by many investors who do bother to look.  At first, this real estate development company seems like a high-beta homebuilder.  But its development properties are over seven years old and more than 80% of these properties are in Texas, where home prices have appreciated far less and affordability is reasonable.  Also, the company has key non-homebuilding assets, such as timber inventory worth about $200 million, and oil and gas rights on 395,000 acres in Texas that are worth another $75 million.  Finally, Forestar has a variety of hidden assets that are easy to overlook, such as complex joint ventures, and a downtown Austin hotel in a coveted location worth $60 million. 

Since Forestar is neglected, misunderstood, and has no catalyst, the stock is undervalued.  FOR trades at $24, just 70-75% of my $33 estimate of the stock’s value.  A simple illustration of the disparity between price and value is that Forestar’s enterprise value per acre of land is $2,900, while the company is selling its raw undeveloped land for over $7,000.  Of course, the company has an extensive portfolio of developed properties that are selling for over $100,000 per acre.  

A detailed valuation analysis follows.

Forestar has seven categories of assets that I value separately.

1) Timberland ($115 million, or $3 per share)
2) Future development land ($525 million, or $15 per share)
3) Land in the entitlement process ($150 million, or $4 per share)
4) Development land ($360 million, or $10 per share)
5) A hotel ($60 million, or $2 per share)
6) Mineral rights ($75 million, or $2 per share)
7) Joint ventures ($115 million, $3 per share)

The company’s primary two assets are timber and land, both of which typically appreciate over time.  For example, the value of rural residential land has increased at a CAGR of 6% over the past 10 years, farmland 9%, and pastureland 10%.  Forestar owns 374,000 acres of land, of which 350,000 acres are covered with timber.  Timber appreciates over time because thicker logs can be used for higher value applications.  The business model is to maximize the net present value of each acre through one of the following paths:  

(1) to sell undeveloped land if Forestar receives a purchase offer price for it that is higher than management’s estimate of the land’s NPV,

(2) to entitle the land so that it is ready to develop and to sell it to a developer if Forestar receives a purchase offer price for it that is higher than management’s estimate of the land’s NPV, and

(3) to develop the entitled land into residential communities and commercial properties and sell lots to home builders and other businesses.

1. Timberland worth $115 million, or $3 per share

The majority of the land, or 330,000 acres, is undeveloped timberland located primarily in Georgia (80%), but also in Texas (15%) and Alabama (5%).  This undeveloped land can be divided in two buckets: land that will be forested indefinitely, and land that will be developed on the foreseeable horizon.  Forestar does not make this distinction because management believes all the land has a higher value than perpetual timberland, so the company reports these two buckets as one, and calls it “undeveloped land.” 

I estimate that the first bucket contains 135,000 acres, or 30% of the land.  During the fall of 2007, Temple-Inland sold 1.55 million acres of timberland that was encumbered by supply contracts of 12 and 20 years in length for $2.38 billion, or $1,535 per acre.  Temple retained any parcel that management judged had development potential, and hence higher value than timberland, and these acres are now owned by Forestar.  Despite management’s view, my valuation aims for a margin of safety, so I place no value on the development potential of these 135,000 acres.  Rather, I value them as simple timberland.  Transaction prices for timberland have increased over the last several years, and instead of using the recent $1,535 per acre transaction price to value Forestar’s 135,000 acres of timberland, I use a 20% discount based on the average transaction price of $1,225 over the past four years.  This average is derived from nearly 100 transactions in the US South during 2004-07, involving over 11 million acres.  (Transaction prices for timberland continue to rise, and in February in 2008, TimberStar Southwest sold 900,000 acres of Southern timberland for $1,900 per acre.  TimberStar purchased the land from International Paper in October 2006 for $1,320 per acre.)  

In conclusion, this 135,000 acres is worth about $85 million, or $1,225/acre gross.  The net value is after expenses to sell the land, and taxes (book value is $300/acre).  The transaction multiple of about 15x EBITDA is line with the valuation of public timber REITs, and a modest discount to the 20-30x that timber investment management organizations like The Campbell Group (TIMOs) have paid in the past few years.  

2. Future development land that has not entered the entitlement process worth $525 million, or $15 per share

The second bucket of “undeveloped land” includes 195,000 acres that is primarily in close proximity to Atlanta.  (Forestar’s registration statement details the land by county and provides various maps.  Population growth in these counties was over 14% between 2000 and 2006.)  This land is worth more than timberland.  Although Forestar has sold undeveloped land (not in the entitlement process) consistently over the past 15 quarters for about $7,000 per acre, a margin of safety is established by estimating the sales price at just 60% of these recent transactions (discussions with community banks in Georgia suggest that lot prices are off by as much as 25% in some suburbs of Atlanta).  Based on a gross transaction value of $4,200 per acre, the net value of the land after expenses and taxes is $525 million, or $2,700 per acre.  

Forestar has sold about 3,000 acres per year during the past four years.  For comparison, Plum Creek has sold “HBU/Recreation” lands on average in the quantity of 24,000 acres per year at $3,200 per acre.  This land has different characteristics than Forestar’s land.  First, it is located in many states across the country, including places like Wisconsin and Maine.  Timberland in these areas is less productive than in Georgia, where Forestar owns land, so the hurdle rate for selling land is lower.  Also, most of this land is not in close proximity to a city, like Atlanta.  Instead, the primary use will be for outdoor recreation or vacation homes.

Also for context, about 50,000 acres of land is located in Texas, primarily waterfront property on lakes and rivers that is in close proximity to Dallas and Houston.  A large portion is located on Sam Rayburn Reservoir in Angelina National Forest.  Land five miles from the reservoir sells for around $2,500 per acre, while waterfront lots are offered for around $30,000 per acre.  

3. Undeveloped land in entitlement process worth $150m, or $4.50 per share

Forestar is seeking entitlement on 24 projects covering 26,750 acres of land.  Nineteen of the projects are in Georgia, mostly within 50 miles of Atlanta.  But there are two projects in Los Angeles County and three in Texas, primarily a large project in Lake Houston.   The entitlement process typically includes obtaining zoning, government approvals, and access to water, sewer, roads, and utilities.  (The registration statement details the location and book value of these properties.)

My estimate of fair value for the undeveloped land currently in entitlement is $150 million.  Forestar does not typically sell land that is in the entitlement process, so there are no “transaction” prices to use to value this land with.  Instead, the company sells either raw undeveloped land or entitled and developed property.    Accordingly, I derive a value for land currently in the entitlement process by approximating future cash flows for when it becomes developed and discounting the free cash flows at 20%.  A useful check on my estimate is the per acre value of $5,600, which compares reasonably with the $4,400 that Forestar has been netting by selling undeveloped land over the past several years.  The premium is warranted because these 26,750 acres can be developed far sooner than the undeveloped 195,000 acres of higher value timberland.      

My cash flow forecasts assume that no entitlements are completed for the next two years, and then 10% of the 26,750 acres is entitled each year for the following ten years.  This is an attempt at conservatism since about 7,000 acres of the 26,750 are no longer forested, so they must be fairly close to development.  During 2006, Forestar entitled five projects covering 2,151 acres, and initiated entitlement work on 4,890 acres.  After a project has been entitled, I assume the land is sold ratably over the next ten years.  These volumes are consistent with Forestar’s history.  The revenues and net margins for the development are based on recent financial reports.  For example, I assume that residential land sells for $110,000 per acre, and the selling price increases at a 3% annual rate, and that the net margin is 40%.  

By far, the largest project is Wolf Creek, a 12,180 acre development 25 miles west of Atlanta’s major airport.  Nearby residents expect this community to develop over the next 30 years and for the population to reach 55,000.  

Another important property that Forestar owns is 725 acres in Los Angeles County, California.  The land was acquired in 1997 for $3.2 million, and Forestar has invested $3.5 million on entitlement over the past ten years (so the book value is $9,240 per acre).  If this property is successfully developed, the value could easily reach $30 million ($40,000 per acre).  One of the proposed developments among these 725 acres is Hidden Creek Estates, is a 260 acre project site about 20 miles from both downtown Los Angeles and the Pacific Ocean.  Many nearby homes with three or four bedrooms sell for between $500,000 and $2 million, or about $300 per square foot for the home, and $2 to $4 million per acre for the home and lot.  Assuming the land is about 20% of the home price and Forestar sells 250 half acre lots, the residential portion of this project would be worth $50 million gross and about $25 million net of taxes.  Much of the remaining 600 acres would remain wildlife, and Forestar would capture value by selling to conservation societies or keeping the land and selling easements.  A small portion of the land would likely become commercial property.  

4. Entitled land developed or in development worth $365m, or $10 per share

Forestar owns a portfolio of 16,600 acres of real estate that was acquired beginning in 1986 and is not legacy timberland.  The book value of this land is $365 million, and the average age of the book value (time since Forestar purchased the properties) is seven years.  I think book value is a good estimate of the land’s fair value.  Properties in Texas account for 90% of the portfolio.  Between 2000 and 2006, home prices in Texas increased just 7%, compared with an increase for the entire United States of 50%, for California of 90%, Florida of 87%, and Iowa of 30%.  Accordingly, homes remain far more affordable in Texas than in other parts of the country.  Furthermore, only 75-80% of the book value is from land, while the remainder is from construction costs and other investments.  

If you find book value an overly optimistic estimate of value, then I calculated an even more conservative estimate which is $2.50/share less by taking a discount to book based on the age of the property:  75% discount for properties acquired in 2007, 50% for 2006, 25% for 2005, 10% for 1999-2004, 5% for 1998, and no discount before 1998.  

A discounted cash flow analysis suggests that $365 million is a reasonable net present value for the company’s development properties.  The first assumption is that Forestar sells 10% of its developments each year for the next ten years, consistent with the velocity of recent sales, even though an important goal of management is to increase velocity.  Second, I assume that the average prices over the past four years hold in 2008 (the average is well below 2007 levels), and then prices increase annually at 3% (note this is far below recent price appreciation).  Third, I assume a 45% net margin, and I use a 10% discount rate given the mature status of the projects.  The NPV of these cash flows is $370m.  

For context, current public market valuations of similar real estate development portfolios are sharply lower than valuations in recent years.  For example, Consolidated Tomoka (CTO) shares have fallen from $90 in 2005 to $55; today, the market values the company’s 11,800 acres at $285 million, or about $24,000 per acre.  Avatar Holdings (AVTR) shares have fallen from $82 last May to $45; today, the market values the company’s 17,000 acres at $255,000, or about $15,000 per acre.  My estimate of fair value per acre for Forestar’s 16,600 acres of development land is $22,000.  

5. Commercial properties worth $60 million, or $2 per share

Forestar owns two commercial properties: the Radisson Hotel & Suites, a waterfront hotel located at a coveted location with lakeside views in downtown Austin, and a golf course.  These properties have essentially broken even for the past three years.  My estimate of value, however, is $60 million, based on the real estate potential for the hotel and a $150,000 per room valuation; the lease with Radisson expires in 2008.  In 2007, a Hyatt just across the lake and outside of downtown sold for about $85 million, or $190,000 per room.  The property changed hands for $50 million, or $110,000 per room, in 2005.  In 2006, two hotels were sold in Austin, a Holiday Inn for $45,000 per room, and a Renaissance for $235,000 per room.  The Radisson is currently rated the following by Zagat: Rooms (17), Service (19), Dining (15), and Facilities (18), while the Renaissance, which is located six miles from downtown, is rated the following: Rooms (21), Service (21), Dining (21), and Facilities (21).  There were no ratings for the Holiday Inn and Hyatt.

6. Mineral rights worth $75 million, or $2 per share

The value of this asset would be difficult to estimate with all the necessary information, and we have only minimal information.  My estimate of $75 million is based on 5x the average net income for the past three years ($8m in 2004, $13m in 2005, $17m in 2006, and $20m in 2007).  

Forestar reports earnings data for Natural Resources.  Revenue comes from selling timber, leasing oil and gas mineral interests to exploration and production companies, and collecting royalties for the sale of oil and gas.  Since the value from the timber cash flows is captured in my valuation of the timberlands, the value of the mineral rights is based only on the earnings from the oil and gas lease and royalty income.  

Forestar’s average net income from mineral rights over the past three years is $15 million.  Growth appears promising, given that Forestar has leased just 105,000 acres out of 622,000 total.  The mineral rights should be separated into two buckets.  First, is the more promising area of Texas and Louisiana where the company owns mineral rights in 365,000 acres.  Second is the less promising area of Alabama and Georgia where Forestar owns mineral rights in 257,000 acres.  So far, E&P companies have leased acres on 96,000 of the 365,000 acres in the western region, while they have leased only 9,000 of the 257,000 acres in the eastern region.  

During 2006, Forestar booked $10.6 million in lease revenue for 105,000 acres of leases ($100 for each acre leased), and $17.4 million in royalty revenue for 32,000 acres of productive land ($540 in royalties for each productive acre).  If the company leased the remaining 260,000 acres in the more promising western region where it has 395,000 acres, and the royalty success rate was the same 30% as to date, revenues would approximate $70 million for a single year.  For context, there were 322 producing wells on Forestar’s land as of October 30, 2007, up from 290 on June 30.

7. Joint ventures worth $115 million, or $3 per share

Forestar has three joint ventures.  The most important is one that was formed on January 3 following the spin and the SEC registration statements because the Market is not likely well aware of it.  I think it is worth $50 million, at least.  The value is based on Forestar’s 58% stake in a joint venture with The Jones Company.  

The Jones family owned 24,000 acres of land that is about 30-40 miles northwest of Atlanta, but the land was encumbered because Forestar owned a lease entitling it to timber harvested from the land.  The lease expired in 2025, when Forestar had the right to purchase the land at market value.  Members of the Jones family, who had inherited the property, sought to capitalize on the land’s proximity to downtown Atlanta by developing the land for residential and commercial use instead of growing timber on it.  Forestar was in a very strong bargaining position vis-à-vis the Jones family given the lease, so Forestar picked up about $30 million in value by exchanging the lease rights to 24,000 acres for 58% ownership of 17,000 acres of the land.  The JV includes just 17,000 acres rather than the entire 24,000 acres because 7,000 acres of the original property, largely wetlands, were sold to the state for wildlife preservation at a value of $6,500 per acre.  My estimate for the remaining land, which has greater development potential, is $8,100 per acre gross, and $5,100 per acre net.  

The other two joint ventures are with Cousins Properties, and Forestar owns 50% in both.  The first, called CL Realty, is focused on Texas.  CL Realty is developing sixteen projects that have a total investment of $152 million, of which 76% has been equity.  The average age of these projects is four years, since on average they were started in 2004.  CL Realty has approximately 8,400 lots remaining for sale, and 550 acres of land.  If CL Realty sold the remaining lots for $30,000 each at a 30% net margin (in line with the past four year average), then the lots would be worth $74 million, and the additional acreage might be worth another $3 million at a very conservative $8,000 per acre.  Forestar’s share would be nearly $40 million.

My estimate of Forestar’s interest in CL Realty is $40 million, or about 0.75x book.    

The second joint venture with Cousins, called Temco Associates, is focused on Atlanta.  Temco is developing four residential communities near Atlanta that have a total investment of $36 million, and Temco also owns nearly 7,000 acres of land.  Only 5% of the investment has been financed with debt.  Two of the projects have been essentially completed, and the two that remain are five years old.  Temco has approximately 1,600 lots remaining for sale, and has been selling nearly 500 per year.  If Temco sold the remaining lots for $50,000 each at a 30% net margin, then the lots would be worth $17 million.  But the undeveloped land is worth much more:  if the acreage is worth $8,000 per acre gross, it nets to $37 million.  Forestar’s 50% share of the two would be $27 million.  

My value of Forestar’s interest in Temco is $25 million, or about 0.75x book. 

Catalyst

The spin-off itself was a catalyst. Transparency has increased and is improving, and the company is a takeover candidate.

Management had not been granted options yet in FOR (most hold TIN options) until the the recent board meeting in February.
    show   sort by    
      Back to top