Tax Notes, July 24, 2006.
What if the medium were the message?
We knew something was up when our hairdresser showed up wearing deep plum, nearly black nail polish that looked like she'd raided your correspondent's fridge.
Some readers may have looked askance at the idea of taking fashion advice from a grown woman who dresses like a suburban goth teenager. It's worse than that. For fall, fashion editors are telling all grown women to dress like suburban goth teenagers. Some readers may want to go home right now to hide the wife's fashion magazines.
But you told us the '70s revival was making way for an '80s revival! Precisely.
Goth fashion is the intersection of the '70s and the '80s, just like goth music, from whence the fashion came, mixed Led Zeppelin and Motorhead. Goth fashion is '80s black leather and lace mixed with '70s femininity. Elle has anointed as one of its fashion inspirations Pat Benatar, the '80s metal queen who was classically trained and reluctant to accept a Grammy that was foisted on her.
Yes, those influential Japanese teenage girls do embrace goth, but they don't do Morticia Adams straight up. Rather, they combine the look with schoolgirl touches that make them look like cute but weird Tim Burton characters. Japanese teenage girls are also wearing bubble skirts, which are best avoided by the non-thin.
What advice do we have for women who don't want to dress like suburban goth teenagers or Japanese teenagers or teenagers of any stripe? Some readers are wondering whether they can buy anything at all, but want to feel au courant and fulfill their responsibility to sustain consumer spending. (Beach? What beach? Fashionistas are shopping for their fall wardrobes.)
Here are some new things that readers stuck in conservative environments can enjoy:
Jewelry. Big jewelry is back, as in big yellow gold chain bracelets and gumball-sized pearls. Lose the little stuff. Lose the glassy stones. Lose anything that looks like it came off a chandelier. Glittery is over.
Leopard. Flattering to all ages and complexions, leopard print always comes back because it is a fashion editor favorite. Just avoid leopard print on raincoats or handbags.
Ruffled blouses. Universally flattering and can be worn with pantsuits. Cloth is a better way to feminize a suit than necklaces.
Shoes. Go for the wedges but avoid the platforms. Platforms are actually more dangerous to walk in than the spiky high heels. And don't try to drive a car in any of them.
Coats. There are loads of nice long coats, some with naval hardware. Capes are convenient and comfortable, but it takes presence to carry them off.
Leggings. OK, your teenage daughter — that'd be the one who isn't a goth — is wearing leggings with short skirts. Leggings are easier on the non-thin than baggy pants, the other leg covering being pushed. Your daughter nonetheless has the right idea — and do not wear spike heels with leggings.
The economic substance test is back in fashion, and some readers no doubt wish it would just go away, along with leggings and Pat Benatar. The Federal Circuit Court revived it in Coltec Industries v. United States, No. 05-5111 (Fed. Cir. July 12, 2006).
The fashion mags are yapping about layering, and there are several layers of significance to the Coltec decision. First and foremost is the court's correct application of the economic substance doctrine. By correct, we mean the court went through the statutory exercise first and thought through the remedy for a transaction lacking substance. The court made some significant statements about economic substance and reversed the nutty position taken by the trial court that the economic substance doctrine does not exist in the Federal Circuit.
Second, there is the subchapter C question, on which the court sided with Black & Decker Corp. v. United States, No. 05-1015 (4th Cir. Feb. 2, 2006). Well, gee, wasn't there a legislative fix? Yes, for this contingent liability shelter only. The basic scheme of handling transferred liabilities in subchapter C (and by extension, subchapter K) is still a mess.
Third, there remains the important tax accounting question of how the tax law should address contingent liabilities. Your correspondent was cited in the opinion, at footnote 6, for the point that contingent liabilities can be priced to everyone's satisfaction, and that when they are transferred, the tax law should accept an estimate and close the transaction.
Corporate tax discussions are on a par with fashion discussions in that both feature a lot of scurrying around and a lot of money splashed out to no great effect. Disproportionate IRS resources have been devoted to responding to every whinge from the corporate tax bar for decades, and what has been the result? Basic tax accounting questions faced by every corporate taxpayer are going unanswered, while IRS lawyers count angels dancing on the heads of pins. All of the technical problems discussed in this article are decades old and were entirely avoidable.
This shelter was made a listed transaction in Notice 2001-17, 2001-1 C.B. 730. Everyone and his brother set up a contingent liability company shelter between the issuance of Rev. Rul. 95-74 and the issuance of Notice 2001-17. The government did not start paying attention to corporate tax shelters until the late 1990s, despite ample warnings from conservative practitioners about this shelter.
A settlement offer was made in 2002. The government offered two alternatives: allowance of fixed 25 percent of the claimed capital loss on the sale of the transferee subsidiary shares, or independent binding "baseball" arbitration under which the arbitrator chooses between competing offers. (Rev. Proc. 2002-67, 2002-2 C.B. 733.) The offer does not appear to have been popular with corporate taxpayers.
Coltec Industries v. United States, No. 01-072T (Fed. Cl. Oct. 29, 2004) involved the popular contingent liability shelter. This is a do-it- yourself reserve method for taxpayers faced with large contingent liabilities such as environmental liabilities and retiree medical costs.
The taxpayer establishes a special-purpose subsidiary, with which it does not consolidate, funded with liquid assets or its own note in the estimated amount of the liabilities. The subsidiary purports to assume the liabilities and deducts them as paid. The taxpayer sells the subsidiary shares at a loss, taking the position that the contingent liabilities should not require it to reduce its basis in the shares under sections 357 and 358. (Congress enacted section 358(h) to stop this shelter and similar partnership shelters.)
Coltec had argued that the liability subsidiary would insulate the parent and its other affiliates from asbestos plaintiffs. More than 100,000 asbestos suits had been filed against some members of the taxpayer's group. The taxpayer had already centralized liability management within itself; the transaction was intended to offset a large capital gain. The taxpayer had sold the shares of the liability management subsidiary at an artificial loss three months after creating it, during the same calendar year.
Two important technical issues went unchallenged at trial. First, the taxpayer funded the subsidiary with its own note. The IRS stipulated to the value of the taxpayer's own note, about which even Court of Federal Claims Judge Susan Braden had doubts. Second, the IRS did not seriously challenge the efficacy of the liability transfer. The liability subsidiary, which had employees, continued in existence in the hands of the bank that bought the shares.
Judge Braden held that tax avoidance was not the taxpayer's principal purpose in doing the transaction, based on testimony by the taxpayer's chief executive and the reluctance of its insurers to extend coverage. Applying the business purpose test of section 357(b) narrowly to the assumption, she read the case law to say that the closer the nature of the liabilities to the taxpayer's customary business of the transferee and its continued viability, the more likely the business purpose test will be satisfied.
The Federal Circuit Court adopted instead the Black & Decker analysis of the statute. In Black & Decker, the court rejected the IRS's statutory legal arguments under sections 357(b) and 357(c). The court read section 357(c)(3) to require only that the transferor have been able to deduct payment of the liability, not that the underlying business giving rise to the liability have been transferred, as the IRS argued. The IRS had based its argument on legislative history.
"The prototypical transaction Congress had in mind in drafting section 357(c)(3) may well have been one in which a corporation exchanged liabilities as part of a transfer of an entire trade or business to a controlled subsidiary, but nothing in the section's plain language embraces such a limitation," the Black & Decker court concluded. This was wrong, but at least it was an informed choice — the court relied on commentary in this very publication.
The IRS then argued that section 357(b) boots the taxpayer out of section 357(c)(3) by virtue of section 357(c)(1), which states that section 357(c) "shall not apply to any exchange to which subsection (b)(1) of this section applies." The IRS view was that section 357(b) takes precedence over section 357(c), so that if a transferor has a tax avoidance purpose, it is not possible to escape boot treatment by keeping the assumed liabilities below the section 357(c) threshold. (Simpson v. Commissioner, 43 T.C. 900 (1965); Rev. Rul. 78-330, 1978-2 C.B. 147.)
The Black & Decker court, however, took the view that section 357(c)(3) acts by itself, as part of the basis rule of section 358(d)(2), with no requirement that liabilities exceed basis of transferred assets. In Coltec, the IRS argued that section 357(c)(3) should not apply because the transferee would be ineligible to deduct payment of the transferor's accrued asbestos liabilities.
The Federal Circuit Court in Coltec reversed the trial court's holding that contingent liabilities are not liabilities cognizable by section 358, but then, like the Black & Decker court, got hung up on the phrase "would give rise to a deduction" in section 357(c)(3). When the IRS argued that the underlying business had to be transferred, Federal Circuit Judge Timothy Dyk responded, "Nothing in the plain language of section 357(c)(3) limits the liabilities excludable to only those that were transferred along with an underlying business."
The Coltec court joined the Black & Decker court in reading section 357(b)(1) out of the law. The Coltec court believed that section 357(c)(3) operated only through the basis rule, section 358(d)(2), and was unaffected by section 357(c)(1). This the court did by torturing the word "under" appearing in section 358(d)(2) in a way that is reminiscent of President Clinton's fussing over the meaning of the word "is." (Readers, we're always in trouble when courts start referring to Webster's Dictionary.)
According to these two courts, Congress should have put everything it wanted to say about the effect of liabilities on basis in section 358. Apparently the placement of paragraph 3 in section 357(c) was pure happenstance, and nothing should be read into its inclusion in that location.
Both the Coltec and Black & Decker courts got the statutory analysis wrong. Sure, nothing in the statute requires that the underlying business be transferred along with the liabilities. But the case law does require that the transferee operate the business out of which the liabilities arose to be permitted to deduct payment of them, on which eligibility for section 357(c)(3) turns.
"Would give rise to a deduction" in section 357(c)(3) refers to the transferee. The case law, not the statute, answers the question whether the transferee gets the deduction. The IRS failed to convince the judge that it was important to take the extra step of looking at the case law to determine whether the transferee was entitled to a deduction.
Holdcroft Transportation Co. v. Commissioner, 153 F.2d 323 (8th Cir. 1946), states that a transferee corporation must capitalize payment of liabilities that did not arise from its own actions. In Holdcroft, the court held that the transferee corporation that assumed the transferor partnership's contingent and unliquidated tort liabilities could not deduct the settlement of those claims because they did not arise out of the transferee's business. Rather than being deductions, the assumed liabilities became part of the transferee's capitalized cost of the transferred assets, in the court's view, even though those assets represented the transferor's business from which the tort claims arose. It did not matter to the court that the liabilities were contingent in the hands of the transferor.
How did we get in the trouble we are in? How did we get the contingent liability company shelter? The road to hell is paved with good intentions. Here the road was paved with the good intentions of the IRS trying to ameliorate the seemingly harsh holding of Holdcroft.
In Rev. Rul. 95-74, 1995-2 C.B. 36, the IRS stated that it would not follow Holdcroft in situations described in the ruling. The IRS ruled that contingent environmental liabilities that have not been deducted or capitalized by the transferor and were assumed by a newly formed subsidiary in a section 351 exchange are not "liabilities" for purposes of sections 357(c)(1) and 358(d). The ruling assumed that section 357(b)(1) was inapplicable. The contingent liability company tax shelter was an abuse or a literal reading of Rev. Rul. 95-74, depending on one's point of view.
The ruling's contingent environmental liabilities had not yet been taken into account by the parent, an accrual-method corporation, before the transfer (and therefore had neither given rise to deductions for the parent nor resulted in the creation of, or increase in, basis in any property of the parent). The ruling depends on the notion that the liabilities are inchoate, so they cannot be explicitly transferred as real liabilities.
The ruling assumed a business purpose for the transfer, which included all the assets of an entire, polluted, dying business, so that section 357(b)(1) would not be a question. The transferee subsidiary incurred costs cleaning up the site, some of which were deductible. The IRS pointedly did not answer the question whether the transferred liabilities should be deducted or capitalized.
"These code provisions are not a model of statutory draftsmanship," Judge Dyk commented. The IRS could have won this case under the statute had it challenged the efficacy of the assumption and better explained its statutory argument. Instead, the government had to hope that the judge would bail it out with the economic substance doctrine.
Judge Braden had dismissed the government's overarching economic substance arguments, on the ground that the business purpose test of section 357(b) had been met, while noting that there was no Federal Circuit precedent on the economic substance doctrine.
Judge Braden even went so far as to argue that applying the economic substance doctrine on top of a statute would violate constitutional separation of powers. She sniffed that "it is Congress, not the court, that should determine how the federal tax laws should be used to promote economic welfare." After quoting some of Justice Antonin Scalia's "plain meaning" rants, Judge Braden stated:
The public must be able to rely on clear and understandable rules established by Congress to ascertain their federal tax obligations. If federal tax laws are applied in an unpredictable and arbitrary manner, albeit by federal judges for the "right" reasons in the "right case," public confidence in the Code and tax enforcement system surely will be further eroded.
"That holding is untenable," Judge Dyk deadpanned. He worked his way through the Supreme Court and court of claims precedent establishing the economic substance doctrine, chiding Judge Braden for wanting to ignore its existence. Judge Dyk stated:
The economic substance doctrine represents a judicial effort to enforce the statutory purpose of the tax code. From its inception, the economic substance doctrine has been used to prevent taxpayers from subverting the legislative purpose of the tax code by engaging in transactions that are fictitious or lack economic reality simply to reap a tax benefit. In this regard, the economic substance doctrine is not unlike other canons of construction that are employed in circumstances where the literal terms of a statute can undermine the ultimate purpose of the statute.
Readers know that most circuits have a disjunctive economic substance test, although often the same evidence is used to prove each element. In most circuits, the test is Rice's Toyota World v. Commissioner, 81 T.C. 184 (1983), aff'd in part and rev'd in part, 752 F.2d 89 (4th Cir. 1985).
Rice's Toyota World established a two-part alternative test: subjective business purpose, which translates as a nontax purpose, or objective profit motive, which translates as a "reasonable possibility of a profit." The court usually states those tests in the negative. In Rice's Toyota World, the Fourth Circuit stated: "To treat a transaction as a sham, the court must find that the taxpayer was motivated by no business purposes other than obtaining tax benefits in entering the transaction, and that the transaction has no economic substance because no reasonable possibility of a profit exists."
The disjunctive test can sustain deals that ought to get the boot, as the Second Circuit version did in TIFD III-E Inc. v. United States (Castle Harbour), 342 F. Supp. 2d 94 (D. Conn. 2004). In that case, the taxpayer used the self-serving testimony of its executives to sustain a business purpose for a deal that had no objective profit potential.
Both the Second Circuit, where Castle Harbour jurisdiction lies, and the Federal Circuit ascribe no significance to the existence of a tax avoidance purpose. As Castle Harbour demonstrated, this can tip the scales in favor of a taxpayer, depending on how the two main elements of the economic substance test are applied.
In a footnote, Judge Dyk rejected the Black & Decker court's interpretation that required the transaction to lack both objective profit potential and subjective business purpose. Judge Dyk would have booted the Castle Harbour deal for a lack of objective profit potential.
Judge Dyk came up with a more intelligent approach to the disjunctive test. In his formulation, if a taxpayer had no motive other than tax avoidance, then the taxpayer bore a heavy burden of showing that the transaction had economic substance. Here "economic substance" was Judge Dyk's phrase for objective profit potential and meaningful nontax economic effects, which had to be shown by objective evidence.
"While the doctrine may well also apply if the taxpayer's sole subjective motivation is tax avoidance even if the transaction has economic substance, a lack of economic substance is sufficient to disqualify the transaction without proof that the taxpayer's sole motive is tax avoidance," Judge Dyk wrote.
Judge Dyk made a nebulous statement about how a court should focus on the transaction that produced the tax benefit. "The transaction to be analyzed is the one that gave rise to the alleged tax benefit," he wrote.
No doubt some taxpayers' advisers will read that to mean that the court should focus on the narrow transaction designed to produce the tax benefits and not the surrounding circumstances, such as the derivatives that negated the taxpayer's claimed economic ownership. Narrow focus on only what the taxpayer wants the government to see is not the proper focus. Case law counsels that all of the facts and circumstances are to be taken into account. (ACM Partnership v. Commissioner, 157 F.3d 231 (3d Cir. 1998), cert. denied, 526 U.S. 1017 (1999).)
What Judge Dyk was trying to do was cut out the window dressing that the taxpayer imported into a hokey deal to produce a tax loss. The taxpayer had argued that the liability management subsidiary, the sale of shares of which produced the claimed loss, was necessary in dealing with its asbestos problem. Judge Dyk wanted to sort the wheat from the chaff. A big tax loss was not the answer to the asbestos problem.
"There is a material difference between structuring a real transaction in a particular way to provide a tax benefit (which is legitimate), and creating a transaction, without a business purpose, in order to create a tax benefit (which is illegitimate)," Judge Dyk commented. This would be the difference between UPS of America, Inc. v. Commissioner, 254 F.3d 1014 (11th Cir. 2001), a restructuring, and ACM, a hokey add-on.
The judge also revived an older concept that seems to have been forgotten recently — that dealing with related parties should be subject to strict judicial scrutiny. We live in a foolish era when grown men and women who should know better go to great effort to try to tease arm's-length prices out of intragroup transactions. Back when the consolidated return rules ignored intragroup transactions, the economic substance doctrine looked at related-party transactions with a jaundiced eye. (Higgins v. Smith, 308 U.S. 473 (1940).)
"Arrangements with subsidiaries that do not affect the economic interest of independent third parties deserve particularly close scrutiny," Judge Dyk wrote, citing Gregory v. Helvering, 293 U.S. 465 (1935), and Basic Inc. v. United States, 549 F.2d 740 (Ct. Cl. 1977).
Judge Dyk correctly concluded that business purpose is a legal, not a factual, matter, so that he owed no deference to Judge Braden's ruling. With the fact of tort law against it, Coltec was careful to argue that the creation of the contingent liability subsidiary would make it more attractive as a takeover target and would put up a barrier to claims against the parent.
Judge Dyk responded that the creation of the subsidiary to manage asbestos liabilities was not the transaction at issue. The question, in his mind, was not the transfer of management of liabilities but the new subsidiary's assumption of those liabilities. Foolishly, the IRS had not seriously challenged the efficacy of the liability transfer. Had it done so, both judges would have seen clearly that no business purpose was accomplished by an ineffectual transfer of asbestos liability.
Judge Dyk was struck by the fact that the subsidiary had managed another affiliate's liabilities without assuming them. He correctly concluded that any management purpose could be accomplished without building a high basis into the shares of the subsidiary. He concluded that the tax gimmick was not necessary to this purpose.
The judge was not amused by the self-serving testimony of Coltec's executives, from which he quoted extensively. He commented that "economic substance is measured from an objective, reasonable viewpoint, not by the subjective views of the taxpayer's corporate officers."
On the question of the subsidiary providing any protection against claims on the parent's assets, Coltec conceded that the new subsidiary's assumption of liabilities would not shield it or the operating subsidiary that incurred the liabilities from tort claims. Judge Dyk noted that the accommodating banks that bought the shares had to be indemnified against the asbestos poison by Coltec. He recognized the liability assumption to be ineffectual as to third parties — that was the meaningful economic effect that was missing in this case. Coltec was left with a transaction that had no effect outside its little group.
"The problem is that there is no objective basis for suggesting that the assumption of these liabilities by another subsidiary . . . would in any way ameliorate this veil-piercing problem," he wrote. "Looking at the transaction objectively, there is no basis in reality for the idea that a corporation can avoid exposure for past acts by transferring liabilities to a subsidiary."
Courts in economic substance cases frequently fall down on the remedy. If the arrangement is not a partnership, what is the unwind? Here the proper remedy easily suggested itself. Judge Dyk ignored the liability transfer and treated the share basis as though it had reflected only the subsidiary's remaining assets.
The tax law is dead set against premature recognition of expense and loss, as indicated by section 461(h) and the all-events requirement. But financial accounting, with its stated goal of conservatism in income recognition, requires reporting companies to face up to losses early. Financial accounting requires recognition of losses when they are probable. Probable means the event giving rise to the loss is likely to occur. Contingent losses must be accrued when they are probable and can be reasonably estimated.
Prudent business practice requires the same recognition of contingent liabilities, which is why the tax shelters being discussed here depend on a dichotomy between fair market value and the statutory tax definition of liability. That level of probability and pricing has not been good enough for the tax law, which delays recognition of a contingent liability until economic performance occurs on the liability under section 461(h).
Ultimately, a broad, inclusive definition of liability would be better for the system, in the long run, rather than a definition that permitted exclusions. Exclusions have created a road map to abuse. All of the variants of the BOSS shelter stemmed from taxpayers creating their own exclusions from the definition of partnership liability in the absence of guidance from the tax administrator and the courts. On the corporate side, the contingent liability shelter depended on a badly drafted statutory exclusion.
But tax policymakers are still reluctant to go for a broad definition of liability, so section 358(h) and its partnership analogue, a regulation, define liability broadly only for purposes of getting at the known tax shelters built around the prevailing narrow statutory definition. Section 358(h) is an add-on, when the policymakers should have gone back and rewritten section 357(b), saying that it controls section 357(c).
Guidance under section 358(h) is in a holding pattern. Section 358(h) is self-enforcing, but the IRS has yet to figure out how it interacts with section 357. The IRS has the power to resolve this question in regulations, but will not interpret section 358(h) until the contingent liability litigation is finished.
Another important question is what happens to "orphan" liabilities when the trade or business that incurred them has been sold, and the seller either kept the liabilities or indemnified the buyer. This is a common question in business transactions that are not tax shelters. Literally, under section 358(h), the buyer would have to reduce its basis in the acquired assets for an assumed liability even if the seller indemnified or retained the liability.
A big part of the larger problem is that the treatment of contingent liabilities is a tax accounting question that was wrongly classified under subchapters C and K in IRS organization because the question was framed as how to treat contingent liabilities in mergers and acquisitions. The subchapter C and K specialties are noted for the reluctance of specialists to think out of the box or attempt to change long-standing, admittedly stupid rules.
The reticence of the corporate types is demonstrated by the baby step that has been taken in the direction of the treatment of some liabilities in asset acquisitions. The IRS is continuing its practice of answering one little question at a time, rather than facing the larger question head-on. Not only would the baby steps be unnecessary if the larger question were faced, but they are counterproductive insofar as progress toward answering the larger question is concerned. (See, e.g., Rev. Rul. 95-45, 1995-1 C.B. 53; Rev. Rul. 95- 8, 1995-1 C.B. 107.)
Suppose a target company has a reserve set aside for nuclear decommissioning, but it is not a qualified fund under the special rule allowing reserves, section 468A. This is a common situation; many sellers have both section 468A funds and unqualified nuclear decommissioning funds. The normal section 338 rule would ignore the existence of the unqualified fund, even if it were a formal trust, treating it as cash in the allocation of the purchase price.
Temporary reg. section 1.338-6T(a)(5) allows a taxpayer that acquires a target that has an unqualified fund to allocate some of the purchase price to that fund as though it were a freestanding entity (technically shares of a corporation). It requires the fund's assets to be netted against the decommissioning liability in valuation of the fund. It does not involve the use of an estimate or change the purchase price.
Even this baby step is inadequate. The netting rule means the unqualified fund will have little basis allocated to it, making investment decisions potentially taxable to the buyer because built- in gain would be created. Utilities argue that section 461(h) economic performance should be deemed to have occurred when the unqualified fund is transferred to the buyer because federal law makes the buyer liable for decommissioning. This would give the buyer a fair market value basis in the unqualified fund.
Whether the buyer should get basis for an assumed liability, and the seller should have income for relief of the liability, are fundamental tax accounting questions that the IRS has the authority to address in a regulations project. There is no current regulations project to deal with contingent liabilities. An attempt to maintain such a large project died a slow death about a decade ago.
A new, larger section 461 project should be created and assigned to IRS tax accounting people, with the corporate and partnership types limited to an advisory role. This is properly a section 461 project — when and how to recognize contingent liabilities. The subchapter C and K tails are being allowed to wag the tax accounting dog, and the resulting answers are predictably piecemeal and inadequate. We end with Judge Dyk's quotation of our previous article in his footnote 6:
[A] sound alternative to deferred purchase price adjustments is for the parties to estimate the contingent liabilities at the time of sale, with the seller taking the estimate into income and the buyer adding it to its basis in the purchased assets. This would allow the seller to walk away. Valuing contingent liabilities, although difficult, turns out not to be as difficult as incorporating those estimates into tax rules that are premised on precision and symmetrical results.
About Tax Analysts
Tax Analysts is an influential provider of tax news and analysis for the global community. Over 150,000 tax professionals in law and accounting firms, corporations, and government agencies rely on Tax Analysts' federal, state, and international content daily. Key products include Tax Notes, Tax Notes Today, State Tax Notes, State Tax Today, Tax Notes International, and Worldwide Tax Daily. Founded in 1970 as a nonprofit organization, Tax Analysts has the industry's largest tax-dedicated correspondent staff, with more than 250 domestic and international correspondents. For more information, visit our home page.
For reprint permission or other information, contact firstname.lastname@example.org