E&Y, PwC & KPMG Submit Joint Comments on Domestic Production Activities Deduction

Ernst & Young LLP, PricewaterhouseCoopers LLP, and KPMG LLP, on behalf of a coalition of U.S. firms engaged in the residential land development and homebuilding businesses (the "Coalition"), are pleased to offer the following comments on Notice 2005- 14¹ (the "Notice" or "Notice 2005-14") regarding the section 199² production activities deduction. The signatories of this letter have previously submitted comments on behalf of clients.³ Whereas these previously submitted comments present different perspectives on the application of section 199 to land developers and home builders, this submission represents a composite industry view.

The Coalition supporting this submission consists of leading members of industry, including:

• Beazer Homes (USA) Inc.

• Centex Corporation

• The Irvine Company

• The J.F. Shea Co., Inc.

• Lennar Corporation

• Pulte Home Corporation

• Standard Pacific Corp.

• St. Joe Company and

• The Rancho Mission Viejo Company.

The American Jobs Creation Act of 2004, enacted on October 22, 2004, added new section 199 to the Code, which provides a deduction for certain production activities. Included within the scope of this deduction is a provision applicable to "construction activities performed in the U.S."⁴ The Notice defines the term "construction" to mean:

In the explanatory section of the Notice, the Treasury describes who will be considered to be performing construction. Section 3.04(11)(e)(ii) of the Notice states that more than one taxpayer can be regarded as deriving gross receipts from construction with respect to the same activity and the same construction project. An example in that same section indicates that both a general contractor and a subcontractor can have domestic production gross receipts related to the same construction project.

Gross receipts "derived from construction" is defined in the Notice. Section 4.04(11)(e) of the Notice states that domestic production gross receipts ("DPGR") includes both "proceeds from the sale, exchange, or other disposition of real property constructed by the taxpayer in the United States" as well as "compensation for the performance of construction services by the taxpayer in the United States." The Notice, however, requires taxpayers to exclude from section 199 the gross receipts "attributable to the sale or other disposition of land" apart from improvements made by the taxpayer.

Previous Commentary on the Notice

These Provisions, read together, require taxpayers in the trade or business of real estate construction to identify the component parts of their gross receipts⁷ that are attributable to construction activities (Home Building and Land Development) while at the same time excluding gross receipts attributable to the passive holding of land over time. Members of the Coalition have previously pointed out that section 199 neither requires nor supports this exclusion. Homebuilders could be the only industry to have its most significant raw material input excluded from the section 199 tax deduction.

Moreover, as Coalition members have noted previously, implementing this distinction would require identifying gross receipts attributable to construction activities such as building the home and land improvement including, infrastructure, common area improvements, the erection of surrounding buildings, and related development activities, none of which are required to be excluded from construction DPGR because they are derived from construction activities. This undertaking would be complex for a single transaction because the data necessary to identify these elements of gross receipts is not collected for any other purpose. For a taxpayer with thousands of transactions, accurately distinguishing these elements of gross receipts would be a huge undertaking.

Members of the Coalition have also noted that very few taxpayers will be tempted to seek to incorporate gains associated with passively holding land into section 199. The section 199(b) wage limitation, the Notice section 4.04(11)(a) requirement that taxpayers seeking to claim the benefits of section 199 be actively engaged in the conduct of a construction trade or business, and probably most importantly, the strong tax incentive of the potential application of a capital gains rate that individuals, partnerships and S- corporations would have if holding land with substantial appreciation in value all work to limit the universe of taxpayers who might seek to incorporate gains associated with passively holding land into section 199. Thus, members of the Coalition continue to believe that the problem Treasury and the IRS purport to address with this provision is not widespread.

Income from Passive Holding of Land

The need for a provision that excludes from DPGR the gross receipts attributable to the passive holding of land over time apparently derives from the notion that a substantial portion of the income earned by homebuilders is derived from the passive holding of land parcels that appreciate with time. Members of the Coalition have previously expressed to Treasury and the IRS the view that this is a minor component of the income derived by taxpayers who are actively engaged in the conduct of a construction trade or business. Coalition members have also pointed out that the passive holding of land is not always a profitable undertaking, and that a provision that removes the effect of passively holding land which today may be detrimental because it reduces DPGR, may benefit taxpayers in the construction business in the future under different economic circumstances.

Smith Barney homebuilding industry analyst Stephen Kim recently released an analysis that addresses directly the concern, expressed by many investors, that the profits reported recently by homebuilders consist largely of profits derived from the passive holding of land.⁸ Mr. Kim makes several points, under the heading "Gauging the Importance of Land Profits," that are directly relevant to the assumption Treasury and the IRS rely upon to support the need for an exclusion from gross receipts attributable to the passive holding of land over time:

This study strongly supports the views expressed previously by members of the Coalition that the passive holding of land is a very minor component of their business profits.

Land Proceeds Where Construction is Undertaken

Given the modest amounts involved, and the massive complexity associated with stripping out these amounts from DPGR, the Coalition suggests that Treasury and the IRS reconsider the need for its land proceeds exclusion. In the event that a land proceeds exclusion is still deemed necessary, the Coalition believes it would be appropriate to create a safe harbor that would permit taxpayers to claim, without limitation, the gross receipts attributable to construction where they are actively engaged in the conduct of a construction trade or business, otherwise satisfy the requirements of section 199, and undertake more than a de minimis amount of construction activities.

Treas. Reg. § 1.460-1(b)(2)(ii) provides a mechanism to ensure that taxpayers undertake more than a de minimis amount of construction activities and could be used to implement this suggested safe harbor. Treas. Reg. § 1.460-1(b)(2)(ii) provides that:

Example 1. Corporation A purchases a finished lot for $40,000 on June, 1, 2005, with all zoning, permitting, platting, engineering, and infrastructure work completed, from a developer and begins construction of a single family home. Costs associated with constructing the single family home total $120,000.

The home is sold on November 30, 2005 for $200,000.⁹ Total basis in the home, also referred to as cost of goods sold ("OGS"), at the time of sale is $160,000 ($120,000 in construction costs and $40,000 land purchase price¹⁰).

Corporation A's determination of whether or not it has met the proposed safe harbor advocated by this comment letter is as follows:

      A Construction Costs Excluding Land                $120,000

 Domestic Production Gross Receipts                     $200,000

The home is sold on November 30, 2009 for $1,100,000.¹³ Total basis in the home just priorto sale is $550,000 ($100,000 in construction costs and $450,000 land purchase price).

Corporation B's determination of whether or not it has met the proposed safe harbor advocated by this comment letter is as follows:

      A Construction Costs Excluding Land                     $100,000

 Unadjusted Domestic Production Gross Receipts          $1,100,000

Simplified Land Exclusion

In the event that representatives of Treasury and the IRS conclude that some adjustment to the section 199 computation is necessary to take into account the fluctuation in land values, if any, attributable to the passive holding of land, the Coalition requests that Treasury and the IRS provide taxpayers in the real estate construction business some form of relief from the massively complex proposal currently incorporated into the notice. Previous submissions point out that the proposed exclusion would be extremely difficult for taxpayers to comply with and for the IRS to administer, in large part because every tract of land is unique and no two tracts of land have identical value. The Coalition is concerned that, without further guidance on this matter, its members will be left to devise procedures based on statistical sampling involving thousands of individual transactions and involving valuation methodologies. Upon examination, the IRS frequently devotes substantial time and energy to the examination of both statistical sampling approaches and valuation methodologies. The Coalition believes that implementing the Notice without further guidance would make the computation of the deduction overly complex and would lead to needless controversy.

The Coalition proposes that Treasury and the IRS consider a flat 5% reduction, to be applied to the total amount of a taxpayer's qualified production activities income ("QPAI") attributable to construction, that a taxpayer could elect to use in lieu of determining the actual amount of inflation or deflation on land. The Coalition understands that a 5 percent reduction will not represent the actual gain on land attributable to the mere passage of time on each and every property sale. The recently released independent study discussed above would indicate that a 2 percent reduction is the best available proxy for an actual reduction, if one is necessary. In order to be conservative, and to take into account the fact that a larger amount might be appropriate in different economic circumstances, and to encourage Treasury and the IRS to abandon the massively complex requirement contained in the Notice, the Coalition is proposing a larger 5% reduction. The Coalition is proposing that this reduction be elective in order to enable taxpayers, if they so chose, to use their actual experience.

Example 3. Assume the same facts as in Example 2 above. Assume further that Corporation B elects the proposed 5 percent reduction. Corporation B's section 199 deduction would be computed as follows:

 DPGR                                         $1,100,000

Members of the Coalition believe that the land proceeds exclusion proposed in the Notice is not warranted. In the event that Treasury and the IRS feel the need for additional protection to address the possibility that taxpayers may inappropriately seek to claim section 199 benefits for gains associated with the passive holding of land, the Coalition suggests the existing section 460 rules could provide an adequate mechanism. If Treasury and the IRS reject this suggestion, the Coalition requests that relief be provided from the massively complex land proceeds exclusion proposed in the Notice, and proposes that an elective 5 percent exclusion from the cumulative amount of a taxpayer's QPAI derived from construction be permitted.

We would be pleased to discuss these comments. If you have any questions, please contact Diane Herndon at (202) 327-8819, Steven Friedman at (703) 747-1940, Brian Meighan at (202) 414-1790, Harry L. Gutman at (202) 533-3044, or Carol Conjura at (202) 533-3040.

Many housing bears voice concern that the builders' profits are mostly derived from the passive holding of land. As we discuss in this report, the evidence suggests otherwise. Rather, by using options, the builders often willingly pay more in order to minimize land risk. We believe the passive holding of land parcels that appreciate with time has only driven the builders' margins 200bps higher than normalized. Overall, we conclude that the advent of supply constraints in the homebuilding industry has been a far greater force behind the builders' improved margins.

In the event of a downturn, we estimate that the reversing of cyclical land profits would drive margins down only about 400bps from current levels. Given the group's current valuation relative to history, we believe that investors expect profit margins to decline almost twice this amount. Finally, we believe home prices drive land prices, not the other way around.

OPINION

We frequently hear from skeptics that the homebuilders probably generate all or most of their profit from the passive holding of land positions. Since builders construct houses on land that was purchased several years earlier, they reason, there is a natural embedded profit for the builders from simply holding land. Some even go so far as to suggest that when home prices rise, it is simply because the land on which the home sits has risen in value. For their part, the builders don't help matters much by consistently refusing to estimate the contribution of land appreciation in their numbers. So, the debate rages on, with bulls pointing insistently at 30%+ annual earnings growth spanning more than a decade, and the bears feeling very much like they are being offered the last two seats on Titanic's maiden voyage.

In our opinion, there can be little doubt that the builders have benefited from this kind of land arbitrage, and that rising land costs will in fact pressure builder margins in the future. However, to our eyes, the builders' profit model incorporates more than mere land speculation. As we discuss below, there are three primary profit drivers at the builders: 1) The profit from the actual construction of the home (500bps); 2) the passive holding of land parcels that appreciate with time (+/-200bps), and 3) the competitive advantage that the large builders wield over their smaller brethren, which allows them to acquire the land in the first place and produce the homes in a more efficient manner (700bps). In our experience, investors tend to overlook this last driver to profits, which includes the public builders' lower cost structure relative to their smaller peers and other supply constraints that have emerged in the past decade to drive margins up for the large, public companies. This has been the linchpin to our bullish thesis on the public homebuilders over the past several years. By failing to factor in this emerging profit driver, investors underestimate the level of operating margin that the builders can sustain. Overall, we conclude that, with the advent of supply constraints in the homebuilding industry that favor large publics, the "normalized" pretax margin for our names is roughly 12% (200bps lower than current), and in a downturn, the negative effect of cyclical land profits could drive margins down about 400bps. Given that valuations in the group are nearly half their own historical peak multiple, we conclude that investors expect profit margins to decline almost twice this amount.

Gauging the Importance of Land Profits

At the heart of the discussion of land margins is this basic understanding: that land is purchased in advance of the home sale, and in the meantime can fluctuate in value. Many investors seem to view this as a damning acknowledgement, tantamount to admitting that the builders are nothing but land speculators dressed up as "real companies." However, the closer one looks at the details, the less this prejudice seems to fit the facts. The available data shows that the builders do not behave as if their profits are primarily generated from land; on the contrary, they seem to go out of their way to pay more for land in order to minimize land risk. Moreover, land costs as a percentage of revenues have remained stable while operating margins have skyrocketed, and some builders that own no land whatsoever still manage to generate very strong margins.

The Builders Pay to Own Less Land . . .

The builders today own less land than they once did. This is a curious detail that is worth exploring. For, if builders derive much of their value through their land holdings, one might expect them to be striving to own as much as possible. However, the reverse appears to be true: As shown in the chart below, the average public builder in 1989 owned roughly 5-7 years worth of land, in addition to the lots they controlled through options. This heavy land supply, however, brought with it heavy debt loads and inflexibility, and subsequently, many builders either went bankrupt or took massive writedowns on their land assets. In the 1990s, the builders made a conscious effort to reduce their land exposure, and by 2001, the average builder owned a mere 2-3 years' supply of land. While we have seen land holdings increase somewhat over the past two years, the fact remains that the builders today still own considerably less land than they once did. The builders buttress this owned land supply with roughly 4 years' worth of land options, slightly more than the 3.1 yrs they had in 1989.

It should be understood that this reduction in owned land and resulting increased percentage of lots controlled through options has a cost associated with it: since the land seller continues to shoulder market risk, there is an option premium that the builder must pay over the purchase price. This effectively increases the builder's land cost, in exchange for reduced risk.

So, by relying increasingly on options to control land, the builders have intentionally been diluting their land arbitrage margins. If, as some suspect, the builders derive their profits primarily from land investment, this increased option activity would be counterproductive. Therefore, the builders' own actions suggest that their profits are tied more to the construction of homes than passive land investment.

. . . by Using Land Options to Reduce Volatility and Risk.

The increased use of options minimizes the risk of owning particular parcels of land and should result in less overall margin volatility through the cycle.

• First, since options mute the upside to a builder's land margin, this results in lower cyclical peak homebuilding margins. So, when the downcycle comes, homebuilding operating margins have less far to fall.

• Second, like any call option, the purchaser benefits from reduced downside risk. With owned land, if the market value of the land declines below its original purchase price, the builder sustains a loss from the investment that worsens as land prices continue to drop. With an option, however, if market prices for land are below the strike price, the builder will simply choose not to exercise the option, or purchase land in the open market, instead. In either case, the builder's downside is less than if the land had been purchased outright. Of course, if the builder has a significant percentage of his capital in options on lots that all move down in value at the same time, the builder's exposure could be magnified by the increased leverage, but we note that the public builders generally have very little capital at risk with their options, and that the underlying lots provide improved diversification.

• Third, while in theory, a builder who holds an "out of the money" owns worthless paper, in practice, land sellers are often willing to renegotiate the original option down to a lower strike price. In other words, the builders get their cake and eat it too: if land prices rise from the time the option is set, they own the right to purchase the land at below-market prices; if, on the other hand, land prices fall from the time the option is set, the seller lowers his price. This claim, often made by the builders themselves, was corroborated for us during a conference we held in NYC last year, where three national land banking firms (who write land options to builders) acknowledged that in a downturn, they would likely renegotiate the terms downward. As we have discussed in numerous reports, the pool of qualified buyers for large tracts of land has become smaller and more homogeneous over the past decade, with the major public builders having similar capital costs and requiring similar rates of return. As a result, in a situation where land prices have dropped to a point where one builder would be willing to walk away from an option, it is highly unlikely that the seller would be able to find another willing buyer at that price. Thus, it is in the option writer's interest to renegotiate with the option holder, as doing so entails significantly lower deal costs. Thus, with options, trough margins are protected, overall margin volatility is reduced, and their balance sheets are preserved. Altogether, we believe this favorable confluence of results argues for higher price/earnings multiples than in the past, the builders currently trade at much lower P/E multiples.

Ultimately, if land profits were the driver behind the builders' rising margins, land costs as a percentage of revenues should be dropping. However, Ryland shows that its land costs as a percentage of revenues have changed very little over the past five years. Apparently, the builders' material, labor and financing costs have risen more slowly than land costs. Again, this seems to run counter to the notion that rising land profits have been driving builder results.

Sizing Up Passive Land Profits

The notion that the builders are recording outsized land profits today stems from the fact that when a home is delivered, the land on which the home was built was often purchased several years ago. In any given year, therefore, there is a portion of profit derived simply from the passive holding of land -- which many investors view as lower-quality earnings. After all, they reason, merely sitting on land that is going up in value seems more like speculation than vocation. So as builder margins have expanded over the years, many skeptics have suggested that these moves have largely come on the back of rising land values -- and have been reluctant to ascribe a multiple to builder earnings.

As we will discuss later, we believe that this line of reasoning distorts the cause-effect relationship between land and home prices. Moreover, it gives the builders no credit for navigating an increasingly difficult land entitlement process effectively. However, sticking to the basic question, is there a way we can estimate how much of the builders' profit comes from this cyclical land "arbitrage"? We believe there is, and that the amount is less than most people think.

Passive Land Profits Rise Early in the Cycle . . .

We start with the basic assumptions that land prices move coincident with home prices, although with greater amplitude, and that the builders move through their land inventory in a first-in, first-out manner. From this, we deduce that the land arbitrage opportunity is really only a factor during the first several years of a housing upeycle or downcycle -- in other words, in and around the trough of the home price cycle. For instance, at the trough of a cycle, the builders are delivering homes on land that was purchased close to the peak, and so margins are quite depressed. Then, as home prices improve entering the recovery, the builders build on land that was purchased at progressively lower prices until at some point, they are delivering homes on land that was purchased at the trough of the cycle and recognizing the greatest amount of passive land profit. Beyond this point, home prices may still increase, but so will the land cost underlying the homes. Common sense suggests that this crossover point occurred somewhere in the mid- to possibly late-90s. In any event, it seems far-fetched to believe that land price arbitrage could be contributing to margin gains now, nearly 15 years since the last noticeable home price trough!

We next analyze the historical operating margins for CTX and PHM to help us quantify how much of a swing factor this land arbitrage represented to homebuilder margins during the late 80s and early 90s. We use these two builders since they were the only companies that were reasonably comparable to today's public builders with respect to geographic diversification, debt ratings, and conservative land policies. The rest of the builders were either not publicly traded or were far too small and geographically concentrated and therefore much more susceptible to swings in local housing factors. While CTX and PHM have numerous factors that prevent their historical margin figures from being exactly comparable to current data (historical numbers include land profit, more recent ones do not), their results provide the best window into how today's large builders might perform in a similar downturn.

Interestingly, during the housing crash of 1990 and the resulting recovery, Pulte's and Centex's homebuilding operating margins gyrated only 100-300bps, and seemed to oscillate around the 5% level. Some of this may be tied to aggressive SG&A reduction; Pulte's gross margin declined closer to 500bps peak-to-trough (we do not have gross margin data for Centex). However, even this more pronounced move suggests that normalized operating margins for the builders cyclical land profits influenced normalized margins by no more than +/-250bps back in those years.

. . . Which is Not When the Builders Saw Their Big Margin Gains.

The data seems to demonstrate that the lion's share of the builder's margin improvement has occurred AFTER 1995 -- which is precisely when we have argued that the industry witnessed the emergence of supply constraints. From 1995 to 2003, margins for CTX and PHM rose roughly 600-700bps in a very steady fashion; likewise, our composite of 13 public builders shows gross margin improvement of 700bps since 1995.

This margin improvement occurred AFTER the period where cyclical land profits could have reasonably been adding to margins, and during a period where new home price growth was fairly contained by historical standards. Only in mid-2004 did we begin to see the rate of new home price growth surge to prior peak levels -- too late to explain the run-up in delivered margins over the past ten years.

Summing up, we believe that the emergence of supply constraints in the mid-90s -- not passive land investment -- drove normalized operating margins up 700bps from the 5% range to the 12% level. Amidst these constraints, the public homebuilders, who were able to gain access to the capital and the land necessary to construct homes benefited from tighter inventories, market share gains, volume discounts . . . and higher margins.

Thus, by subtraction, passive land investment likely accounts for only the remaining 200bps in current margins. We would agree that if new home price growth persists at current levels, it would cause reported margins to spike even farther above normalized in FY06 and beyond. However, at this point, with operating margins averaging roughly 14%, we believe a reasonable downside scenario would incorporate not much more than a 400bps contraction in margins over a two-year period.

Home prices set landprices, not the other way around.

This point, although perhaps a bit basic, needs to be stated directly, for there are some investors who suspect that the profits the builders make from selling homes is derived almost entirely from land inflation. We think this confuses the relationship of home prices to land prices. After all, land prices increase primarily when the home to be built on it rises in value. In almost all cases relevant to this discussion, there are no competing uses for the land other than homebuilding -- certainly, lots in coastal California are not prized primarily for their agricultural value! Rather, as home values inflate, those remaining land parcels that are suitable for residential development become increasingly valuable. In other words, market prices for homes set market prices for land. To think it works the other way around is to suggest that when one acquires a pet dog, he is bringing home a tail that happens to be attached to a canine.

Of course, we are not saying that land has no intrinsic value -- rather, we contend that land in residential areas derives most of its value from the home(s) around it. For instance, if two neighboring homes are similar in all respects, except that one home sits on 3 acres while the other sits on 1 acre, one would naturally pay more for the larger plot, even if it was determined that no additional construction could take place on the property. However, the value of these extra 2 acres is greatly enhanced by the fact that it provides privacy and views to the home that it surrounds. The same 2 acres would, for example, be worth far less to anyone who lived on the other side of town. Similarly, even in highly land-constrained markets on the coasts, small parcels that cannot be developed for residential use are virtually worthless when compared to land that is approved for construction. On the whole, a homesite generally accounts for about 25-35% of the total selling price of the home; if most of the profit resided in the land, we suspect market prices for land would represent a much higher percentage of the total.

This relationship between house and land is worth bearing in mind when weighing whether to trade builders on earnings or book value. We believe that when one buys a homebuilder, he is buying into an operation that converts land into a consumer good. Land apart from a house has little value, and a house with no land to put it on is not much use to anybody. The ability to combine the two is where the value is created, and as we have discussed in numerous earlier reports, we believe this process is becoming increasingly difficult for small, under-capitalized companies. The large, public builders, by contrast, have consistently shown an ability to acquire and develop land, manage the building process, and produce housing communities in desirable locations all over the country, both urban and suburban. Thus, we conclude that the large public builders, with their strong balance sheets, growing economics of scale, sophisticated land acquisition teams, and long track records of success should trade primarily on Price/Earnings multiples, rather than Price/Book.

² All section references are to the Internal Revenue Code of 1986, as amended ("I.R.C." or the "Code"), unless noted otherwise.

³ See Ernst & Young LLP submission dated February 15, 2005, PricewaterhouseCoopers LLP submission dated February 18, 2005, and KPMG LLP submission dated May 9, 2005.

⁴ § 199(c)(4)(A)(ii).

⁵Id.

⁶ § 4.04(11)(d) of Notice 2005-14.

⁷ Gross receipts for homebuilders include proceeds attributable to the home, land (unimproved) and land improvements.

⁸ Stephen Kim, A Solid Foundation: The Good Dirt on Land Profits June 2, 2005, at www.smithbamey.co (registration required). This document, reformatted for readability, is incorporated herein as Attachment A.

⁹ Gross receipts associated with any tangible personal property included in the home are assumed to be less than five percent of the gross receipts from the sale of the home.

¹⁰ Carrying costs required to be capitalized under section 263A are omitted to simplify the example.

¹¹ The use of a calculation for a single home is for illustration purposes only and is not intended to infer that homebuilders should be expected to do this calculation at this level of detail due to the administrative cost of doing the calculation.

¹² Other Costs are assumed to be zero in order to simplify the example.

¹³ Gross receipts associated with any tangible personal property included in the home are assumed to be less than five percent of the gross receipts from the sale of the home.

¹⁴ Other Costs are assumed to be zero in order to simplify the example.

¹⁵ For purposes of this example, it is assumed that an independent appraisal was done and that the appreciation in land attributable to the passage of time was $25,000. The increase in land value attributable to B's construction activities are not excluded from section 199.