PATRICK E. HIGGINBOTHAM, Circuit Judge:

Taxpayers appeal the judgment of the United States Tax Court which found that they fraudulently failed to declare and pay tax on approximately $1,280,000 of income. The court determined that Robert W. Lisle, along with Claude M. Ballard and Burton W. Kanter, earned the unreported income through an elaborate scheme involving the sale of influence by Lisle and Ballard at Prudential Life Insurance Co. of America, whereby Lisle and Ballard would direct business to those persons who agreed to pay a commission on the business to Kanter. Through numerous transactions involving various sham corporations and trusts, the kickbacks were distributed among Lisle, Ballard, and Kanter in a 45-45-10 percent split.

The Lisles assert that the evidence does not support the finding of fraud or the assessed deficiencies. They also allege that their due process rights were violated by the application of Tax Court Rule 183, whereby the Tax Court Judge reviewed the findings of the Special Trial Judge without making the findings of the Special Trial Judge available to them or this court. After an exhaustive review of the record, we find that the Tax Court clearly erred in determining that the government proved a deficiency due to fraud by clear and convincing evidence. However, the evidence supports the assessment of a deficiency under the less strenuous standard of a preponderance of the evidence, and we therefore affirm the deficiencies for those years not barred by the statute of limitations. Finally, we decide that the application of Rule 183 did not violate the Lisles' right to due process.

I.

It is well settled that "the courts afford IRS determinations of deficiency a presumption of correctness." To rebut this presumption, "the taxpayer bears the burden of proving by a preponderance of the evidence that the determination is arbitrary and erroneous." Once the taxpayer has established that the assessment is arbitrary and erroneous, "the burden shifts to the government to prove the correct amount of any taxes owed." In addition, when the Commissioner in his Tax Court pleadings increases the deficiency asserted against the taxpayer, he bears the burden of proof for the increase by a preponderance of the evidence. We review the Tax Court's approval of the Commissioner's determination of taxable income for clear error. To reverse the Tax Court's approval of the Commissioner's deficiency, we must find that the Tax Court clearly erred when it determined that Lisle failed to rebut the presumption of correctness of the Commissioner's deficiency by a preponderance of the evidence, or that the Commissioner failed to prove the additional deficiencies by a preponderance of the evidence.

In addition to the deficiency, the Tax Court found that Lisle was liable for a fraud penalty. Pursuant to I.R.C. § 7454(a) and Tax Court Rule 142(b), the Commissioner bears the burden of proof with respect to the deficiencies in tax and penalties for fraud by clear and convincing evidence. To sustain a fraud penalty Rule 142(b) requires proof by clear and convincing evidence both that an underpayment exists, and that some portion of the underpayment is attributable to fraud. In proving an underpayment by clear and convincing evidence, "the Commissioner may not rely on a taxpayer's failure to carry his or her burden of proof with respect to the underlying deficiency."

While we have observed that fraud must be proved by clear and convincing evidence, we have never addressed the Tax Court's rule creating two elements, each of which must be proved by clear and convincing evidence. Without challenge by the Commissioner of the Tax Court's reading of Rule 142(b), we assume that both the underpayment and the fraud must be proved by clear and convincing evidence to sustain the penalty. Here there is a significant functional overlap of the two elements, as the effort to prove underpayment and fraud is sustained by much the same evidence — establishing a kickback scheme to hide income proves both an underpayment and points toward fraud, on our facts.

We review the Tax Court's finding that there was an underpayment of tax and that a portion of that underpayment was due to fraud for clear error. We will sustain the penalty for fraud unless we find that the Tax Court clearly erred when it determined that the Commissioner, by clear and convincing evidence, established an underpayment by Lisle and that a portion of the underpayment was attributable to fraud.

A finding is clearly erroneous when, "although there is evidence to support it, the reviewing court on the entire evidence is left with the definite and firm conviction that a mistake has been committed." Whether a finding is clearly erroneous must be viewed in light of the burden of proof. If the burden of proof is by the preponderance of the evidence, the Tax Court's conclusion that a deficiency was proved would not be clearly erroneous if the Tax Court chose between competing inferences from the facts. The same evidence may fail, however, to meet the requirement that proof be clear and convincing.

II. A.

The IRS mailed notices of deficiency for tax years 1984, 1987, 1988, and 1989 to Lisle and his wife on August 15, 1991, July 24, 1991, July 2, 1992, and April 5, 1993, respectively. In connection with the notices of deficiency, the Lisles filed actions on September 9, 1991, September 23, 1991, July 16, 1992, and April 16, 1993. The Tax Court consolidated the Lisles' suit with twenty-four additional Tax Court actions involving the other participants in the alleged scheme. After a five week trial, the Tax Court filed a 600 plus page opinion and then entered final judgment in the four cases concerning Lisle and his wife on July 24, 2001. The Lisles timely appealed. Lisle and his wife died during the pendency of the actions, and their estates were substituted as parties. The following narrative is stated in the light most favorable to the government.

B.

The income deficiencies stem from an elaborate scheme of alleged influence selling, kickbacks, and money laundering through sham corporations and trusts. The case centers around five business arrangements whereby Lisle, Ballard, and Kanter assisted individuals in obtaining business opportunities or venture capital from the Prudential Life Insurance Company of America. Lisle worked for Prudential in real estate development and mortgage financing from 1950 to 1982, and Ballard worked there from 1948 to 1982. Between 1968 and 1970, they worked together in a regional office in Houston, Texas. At the request of Donald Knab, Prudential's senior vice-president in charge of real estate investments, both Lisle and Ballard moved to Prudential's corporate headquarter in Newark, New Jersey, in the early 1970s.

As the head of real estate development from about 1975 to 1982, Ballard had the ability to influence the selection of contractors and builders on projects. Ballard's staff bought, leased, and sold Prudential's real estate and supervised the property managers and leasing agents of that real estate. Ultimately, Ballard became senior vice-president in real estate operations — the highest position in Prudential's real estate operation — placing him in charge of all operations, acquisitions, sales and portfolio management of the equity investments in real estate.

At the same time, Lisle headed Prudential's mortgage operations and was responsible for lending money and buying and developing real estate. Lisle had authority to commit any loan up to $20 million and to award construction contracts. He conducted his work through a subsidiary corporation of Prudential called PIC Realty Corp. (PIC Realty), of which he was president.

Kanter began practicing law in Chicago, Illinois, in 1956. At the time of the trial and for the previous ten years, Kanter taught courses in estate and gift taxation and estate planning at the University of Chicago Law School. He has written and lectured extensively in the area of federal tax law. As a result of his expertise, Kanter had a highly successful law practice and was involved in consultation, development, and investments in a number of business fields and enterprises. Kanter represented the Pritzker family, who owned the majority of stock in the Hyatt Corporation. While at the opening of the Houston Hyatt Regency in the early 1970s, Kanter met Lisle and Ballard. Kanter and Lisle invested in partnerships together, and Kanter's law firm did estate work and established insurance trusts for Lisle.

Kanter established a number of corporations, partnerships, and trusts allegedly to receive, distribute, and disguise illegal kickbacks from five business arrangements at issue here. These five transactions and their principal participants are referred to by the parties and the Tax Court as "the Five."

Most of the payments in this case initially were made through Investment Research Associates, Inc. (IRA), which Kanter incorporated in Delaware. Solomon Weisgal, trustee of Kanter's family's Bea Ritch Trusts, owned fifty percent of the company's voting stock. Mildred Schott, a legal secretary and real estate broker, owned the remaining 50 percent of the voting stock. Bea Ritch Trusts owned 1,000 shares of common stock, and Schott owned 1,000 shares of class A preferred voting stock. No individual, business, estate, or trust owned more than fifty percent of the corporation's total voting stock. Delores Keating, a real estate sales person, served as IRA's president until 1975. Schott was president from 1975 to 1980. Lawrence Freeman served as president from 1980 to 1989. Other officers of IRA included Sharon Meyers, who was the company's secretary, and Patricia Grogan, who served as a director. It was Kanter, however, who controlled IRA at all times, directing the activities of Weisgal, Schott, Keating, Freeman, Meyers, and Grogan.

IRA employed only bookkeepers and paid no salaries in any year other than 1981 and 1982, when it paid salaries of $9,969 and $26,079, respectively. It owned controlling interests in several subsidiary corporations including Carlco, Inc., TMT, Inc., and BWK, Inc. As we will discuss, in 1983, IRA distributed all of its assets to Carlco, TMT, and BWK in a 45-45-10 percent split, which were thereafter managed respectively by Lisle, Ballard, and Kanter. The government asserts that forty-five percent of the payments from the Five to Kanter corporations were distributed to Lisle based on this arrangement.

C. -1-

The first of the Five arrangements involves J.D. Weaver and Hyatt's contract to manage the Embarcadero Hotel in San Francisco. Lisle and Ballard met Weaver, an officer of a Tenneco Corp. subsidiary, in the late 1960s. The Tenneco Corp. was working with Prudential to build the Houston Hyatt during this time. In the early 1970s, Prudential participated in a joint venture to build the San Francisco Embarcadero Hotel. Lisle supervised the building of the Embarcadero and participated in the selection of a management company to operate it.

Intercontinental Co., Del Webb Co., and the Hyatt Corporation expressed interest in the contract. A.N. Pritzker, who controlled the Hyatt Corp., hoped to manage the Embarcadero because the hotel would become the third or fourth Hyatt-operated hotel in the United States at which major conventions would be held. But Lisle was not interested in a bid from Hyatt as Hyatt was simultaneously planning a hotel to compete with the Embarcadero. As a result, Pritzker offered Weaver a share of the management fees if he assisted Hyatt. Weaver then convinced Lisle to allow Hyatt to bid on the contract.

Hyatt submitted the only bid and was awarded the Embarcadero management contract. On February 25, 1971, Hyatt agreed to compensate Weaver's corporation, K.W.J. Corp., for Weaver's help in obtaining the contract. Under the agreement, Hyatt was obligated to pay KWJ ten percent of its net cash profits from the Embarcadero management contract. Pursuant to the agreement, Hyatt paid KWJ a total of $2,589,710 for the 1976 through 1993 operating years.

In 1976, Weaver and Kanter agreed that IRA would buy all of KWJ's outstanding stock for $150,000, plus annual payments of thirty percent of KWJ's commissions from the Embarcadero management contract. This agreement was framed as a four-year option to buy KWJ's stock. In a letter dated September 27, 1979, Kanter informed Weaver of IRA's election to acquire KWJ's shares, effective retroactively to November 1, 1978. The purchase resulted in IRA obtaining seventy percent of Hyatt's commissions to KWJ, and allowed IRA to control KWJ's net worth, which was $115,084 as of January 1, 1979. As agreed, Weaver received thirty percent of the commissions.

-2-

The second of the Five involved Bruce Frey, a certified property manager, real estate broker, and insurance broker. Frey did business through his company BJF Development, Inc., which engaged in real estate development and management. Frey was the sole shareholder. In the late 1970s, Kanter introduced Frey to Ballard. From that point on, Frey began doing business with Prudential. Frey's first condominium conversion project with Prudential was the Village of Kings Creek, a 1,000 unit complex in Miami, Florida, which was owned by a Prudential pension fund that Ballard managed. Prudential and BJF participated in five other joint ventures: Calais, Chatham, Old Forge, Valleybrook, and the Greens. Prudential owned the properties and BJF converted them and marketed and sold the units.

Zeus Ventures, Inc., one of IRA's wholly-owned subsidiaries, was a limited partner in the Prudential and BJF partnership. Zeus was to receive five percent of BJF's developer's fees and twenty percent of its profits on all prior and subsequent condominium conversions of Prudential properties. Under this agreement, BJF paid Zeus over $1,000,000 between 1980 and 1985.

-3-

The third arrangement began when Kanter invited William Schaffel, a real estate broker, to have dinner with Lisle and Ballard in New York City in 1979. Schaffel had previously met Lisle briefly but had not done business with him. He had not met Kanter or Ballard before the dinner. At the meeting, Kanter asked Schaffel if he wanted to arrange the financing for a casino hotel to be built in Atlantic City, a deal which did not involve Prudential. Kanter said he would introduce Schaffel to the appropriate people in return for fifty percent of Schaffel's fees from the deal. Schaffel agreed, but the transaction never materialized.

Later, Schaffel agreed to give Kanter fifty percent of any fees earned by doing business with Prudential. This agreement also applied to finders fees that Schaffel obtained for procuring business from Prudential for real estate developer Bill Walters and for Torcon, Inc. Schaffel's first deal with Prudential was negotiating the sale of the IBM headquarters in Lexington, Kentucky, to Prudential. Schaffel dealt with Ballard in its initial stages, but as usual, the deal went through the local Prudential office. As agreed, Schaffel paid fifty percent of his broker's fee to IRA.

In late 1979 or early 1980, Schaffel introduced Benedict Torcivia, Torcon's sole shareholder and chairman of the board, to Ballard. Before this introduction, Torcon had done no business with Prudential. However, after the introduction, Torcon did "quite a bit" of business with Prudential, also first approved by the local Prudential office. Schaffel also introduced Walters to Prudential executives including Ballard. Schaffel assisted Walters' companies in obtaining financing from Prudential for two buildings located in Aurora, Colorado. Prudential agreed to contribute about $30 million in financing to the Ramada Renaissance Hotel project, and about $15.6 million to the Cherry Creek Place II project. In written agreements, Walters' companies acknowledged that Prudential participated in the ventures primarily as a result of Schaffel's efforts and, therefore, he was entitled to compensation.

Between 1979 and 1983, Schaffel paid IRA a total of $1,184,876 as a result of his business dealings with Prudential, which ended when Ballard and Lisle left that company. He then began doing business with Travelers Insurance Co., Lisle's new employer. Although he initially paid IRA a portion of his fees for these deals, he stopped making the payments. Schaffel told Kanter that the deals were not with Prudential, and were therefore not covered by their agreement. Kanter persuaded him that, because he continued to get business as a result of Kanter's introduction of Schaffel to Lisle, their agreement applied. Schaffel resumed making the payments, only now they were made to a different Kanter corporation, Holding Co.

-4-

The fourth involves Kenneth Schnitzer. During the 1960s and 1970s, Kenneth Schnitzer conducted business through Century Development Corp. and was engaged in real estate development in Houston, Texas. In 1974, CDC acquired Fletcher Emerson Co., a property management and cleaning company, for $1.3 million. After its purchase, Fletcher Emerson was renamed Property Management Systems, Inc., and Schnitzer became the company's chairman of the board and chief executive officer.

In 1974, Schnitzer arranged a meeting with Ballard, whom he first met when Ballard was working in Prudential's regional office in Houston, and offered Prudential a fifty percent interest in some PMS management contracts in return for more business opportunities with Prudential. Prudential declined because it managed pension plans that owned some of the properties, and owning both the property and the property management company would create a conflict of interest. Prudential did thereafter give PMS additional business.

Schnitzer then offered PMS stock to Kanter. On November 7, 1977, three years after initially meeting with Ballard, and after talking to Ballard about Kanter's ability to send business to PMS, particularly through Kanter's contacts with Pritzker and Hyatt, CDC agreed to sell forty-seven and a half percent of the PMS common stock to IRA's predecessor for $150,000. The sale took place on February 14, 1978.

Eventually, Prudential became PMS's largest customer. In March 1979, Schnitzer informed Kanter that he wanted to repurchase the PMS stock because its business with Pritzker had not increased as Schnitzer had anticipated. On November 30, 1979, CDC reacquired IRA's PMS stock for $3.1 million payable in installments over ten years. IRA received a total of $4,590,388 in connection with the stock repurchase. The increase in value was due in large part to the increase in PMS's business with Prudential, which began after Schnitzer's initial meeting with Ballard in 1974.

-5-

The final transaction in the Five involved John Eulich, a real estate developer. Eulich met Lisle and Ballard in the mid-1960s through his transactions with Prudential, and Kanter in the late 1960s or early 1970s through Pritzker. In 1975, Eulich participated in the formation of Motor Hotel Management, Inc., a hotel management business. Eulich became MHM's majority shareholder and chairman of its Board of Directors. His role was to find management contracts and financing for the company. During that year, MHM obtained seventeen management contracts, each part of a joint venture with Prudential as lender and a Eulich corporation as developer.

John Connolly owned the Gateway Hotel Management Co., which managed the Gateway Hilton Hotel in Newark, New Jersey, a management contract obtained through Prudential. Connolly also obtained a contract to manage the Midland Hilton Hotel in Texas, which Prudential owned.

Eulich, Kanter, and Connolly formed a partnership called the Essex Hotel Management Co. Essex's partners were MHM (47.5% interest), Connolly (5% interest), and Kanter's entities, IRA (26.125% interest) and Holding Co. (21.375% interest). Essex entered into representation and marketing agreements with GHM and MHM, effective January 1, 1982, requiring Essex to perform liaison functions. In return, Essex was to receive a large part of GHM's and MHM's management fees. Of Essex's partners, only MHM performed any consulting or liaison services. IRA and Holding Co. did not contribute money or services to the partnership but received a total of forty-seven and a half percent of its distributions.

From 1982 to 1988, Essex reported $1,334,601 in commission fee payments from GHM. During the same period, Essex reported $1,563,412 in commission fee payments from MHM. In addition, Essex made the following distributions from 1982 to 1989: $788,452 to IRA; $645,028 to Holding Co.; $150,899 to Connolly, and $1,433,551 to MHM. The distributions to all of Essex's partners from 1982 to 1989 totaled $3,017,930.

D.

In 1982, three corporations were formed: Carlco, TMT, and BWK. At the end of 1983, IRA acquired the common stock of these three companies and listed them as its subsidiaries on its 1983 federal income tax return. Also in 1983, IRA liquidated KWJ, which it had purchased from Weaver. KWJ's assets and Zeus's accumulated funds were distributed to IRA. By the end of 1983, IRA had accumulated $4,771,445 from payments flowing from the Five arrangements.

In 1984, Kanter directed IRA's president to distribute funds it had received in those transactions in the following ratios: forty-five percent to Carlco, forty-five percent to TMT, and ten percent to BWK. It was then agreed that Lisle would manage Carlco's assets, Ballard would manage TMT's assets, and Kanter would manage BWK. Notably, Carlco, TMT, and BWK were not listed as subsidiaries on IRA's federal income tax return for 1984.

On December 31, 1984, IRA transferred its partnership interest in Essex to Carlco, TMT, and BWK in the 45-45-10 ratio. IRA never informed Essex of this transfer, and therefore continued to receive payments from Essex, which IRA then transferred to Carlco, TMT, and BWK. After 1984, IRA also distributed the payments it received in connection with the PMS stock repurchase to Carlco, TMT, and BWK in the 45-45-10 ratio. Likewise, in 1984, after the corporation called KWJ had been liquidated, the KWJ partnership was formed. Carlco and TMT each had a forty-five percent interest and BWK had a ten percent interest in the KWJ partnership. Weaver forwarded the Hyatt commission checks to the KWJ partnership which paid Weaver his thirty percent share, and the balance was distributed to Carlco, TMT, and BWK. Weaver never informed Hyatt of the liquidation of KWJ corporation or the formation of the KWJ partnership.

From 1982 to 1989, first the KWJ corporation and then the KWJ partnership paid "consulting fees" of $1000 per month to each of two of Lisle's and two of Ballard's adult children. The payments were allegedly for their submission of proposed real estate deals. Lisle's children testified that they spent very little time reviewing properties and that none of the deals they proposed were consummated by the KWJ partnership. In February 1990, after the IRS began checking Lisle's, Ballard's, and Kanter's tax returns, Kanter terminated the fee arrangement on behalf of IRA. In a letter, Kanter indicated that it appeared as though no services had been performed by the children for a number of years but that IRA did not intend to seek reimbursement for payments made.

At trial, Lisle admitted that he had complete control over Carlco. He claimed his function in Carlco was to invest its assets in municipal bonds. However, he never obtained a fee for his services. Lisle and his family members were Carlco's officers and directors, having signatory authority over the Company's accounts. In 1989, when Lisle and his wife Donna moved from Connecticut to Texas, Carlco's money was deposited into an account at the North Dallas Bank. Lisle maintained signatory authority over that account. The Christie Trust, which Kanter created in 1983 for the benefit of Lisle's wife and children, owned all 300 shares of Carlco's issued preferred stock. IRA owned all 1000 shares of Carlco's common stock.

In addition, between 1973 and 1980, Lisle established three grantor trusts: RWL Cinema Trust, RWL Cinema Trust II, and the Basking Ridge Trust. Lisle's wife and children were the beneficiaries of these trusts. Kanter's companies made loans totaling $220,000 to Lisle and the three trusts from 1974 to 1990. Neither Lisle nor the trusts paid any interest on these loans. IRA wrote off some of these loans as worthless in 1987. Lisle never reported the discharge of this indebtedness as income on his 1987 return or any other subsequent return.

III A.

As there is no evidence of payments from the Five directly to Lisle, the Tax Court's opinion as well as the Commissioner's argument hinge on Lisle receiving money paid to the various Kanter corporations. The linchpin of the government's theory that there was an arrangement among Lisle, Ballard, and Kanter to sell influence in return for kickbacks is Lisle's actual receipt of monies equal to forty-five percent of the payments from the Five — the government's theory simply cannot stand if Lisle did not receive the money. We ask then whether any of the money paid to the Kanter corporations can be attributed to Lisle, first in light of the clear and convincing evidence standard required for a finding of fraud, then under the more lenient preponderance of the evidence standard.

The Tax Court relied on three rationales for attributing payments by the Five to the Kanter corporations to Lisle: (1) Carlco was the alter ego of Lisle; (2) by the assignment of income doctrine, under which Lisle, the person who earned the income, is responsible for the tax on the income regardless of an assignment of that income to the Kanter corporations; and (3) the Commissioner could reallocate the income paid to the Kanter corporations to Lisle pursuant to § 482.

B.

The first theory is that the Kanter corporations were incorporated pocketbooks, and that Carlco was the alter ego of Lisle. Where a corporation is the alter ego of an individual, the government ignores the corporate identity and assigns the income to the controlling individual. Here, the source of the payments to Carlco is largely irrelevant in determining whether Carlco was the alter ego of Lisle, since the question of alter ego turns on control and use.

We have established a non-exhaustive list of factors to consider in determining whether a subsidiary is the alter ego of a parent corporation, but the question is ultimately determined by examining the totality of the circumstances. Although this list was developed in the context of a parent corporation and its subsidiary, it has been adapted and applied to the relationship between an individual and a corporation. Additional factors we have considered in the context of a corporation as alter ego of an individual include: "the total dealings of the corporation and the individual, the amount of financial interest the individual has in the corporation, the ownership and the control that the individual maintains over the corporation, and whether the corporation has been used for personal purposes."

The Tax Court relied on several pieces of evidence to conclude that Carlco was a sham corporation and the alter ego of Lisle. First, the court concluded that contrary to the position of the petitioners, Lisle, Ballard and Kanter were far more than mere managers of Carlco, TMT and BWK, respectively. The court noted that "[p]etitioners used the funds for their personal benefit." This conclusion, as it applies to Ballard and Kanter, is well supported by the evidence — specifically, the findings that both diverted hundreds of thousands of dollars from their respective corporations for personal use.

With Lisle, the government's case quickly thins; it can only cite two incidents where Lisle potentially diverted Carlco's funds for personal use. First, Lisle used $3,000 of Carlco's funds to pay a receivable on the books of Administration Co., ostensibly for his grantor trust, RWL Cinema Trust. The government contends that this payment was for personal purposes. The only evidence to support this conclusion is that the notation on the check stated "Payment for Loan," and while Lisle and the Lisle family trusts owed money to Kanter and his corporations, the government offered no evidence that Carlco owed Kanter any money. The court concluded, without additional evidence, that the payment was for the loan to RWL Cinema Trust.

The other alleged personal use of Carlco funds was the KWJ partnership's payment of consulting fees to Lisle's and Ballard's adult children. KWJ paid two of Lisle's children $1,000 a month each between 1982 and 1988. The total paid to Lisle's children was $158,000, all of which was reported by the children as income. Prior to 1984, IRA owned and controlled KWJ, which it purchased from Weaver. After IRA distributed its assets to Carlco, TMT, and BWK, Carlco received a forty-five percent share of the KWJ partnership which continued to pay Lisle's and Ballard's children the consulting fees. Although both children had backgrounds in real estate and were paid to submit proposed real estate deals to KWJ, which they did, none of these deals were ever pursued and the children received the payments regardless of how many potential deals they submitted.

It is not clear how this is evidence of an abuse of Carlco for personal gain by Lisle. IRA, undeniably controlled by Kanter, was the sole owner of KWJ before the formation of the partnership when KWJ first started paying the consulting fees. Kanter or one of his employees at IRA was responsible for initiating the consulting fees to Lisle's children. Neither the court nor the government cites any evidence that Lisle used Carlco to cause the KWJ partnership to continue to make the payments once the partnership was formed. There is no evidence that Lisle exercised any power over Carlco to that end, since IRA retained majority ownership of Carlco, TMT, and BWK after the partnership was formed, and it was Kanter, on behalf of IRA, who eventually terminated the payments. There is no evidence that Lisle used his position at Carlco to ensure that payments to his children continued. This speculation, combined with the possible payment of a $3,000 personal loan, will not support the finding of alter ego.

The court pointed to Lisle's admissions that as manager of Carlco's assets, he had "complete authority" over Carlco and "full discretion" over the use of its funds; that Lisle maintained possession of Carlco's records at his home; that the address for Carlco was the same as Lisle's personal residence and when Lisle moved to Texas, Carlco's address was changed to Lisle's new residence. In addition, Lisle, his wife, and brother had signatory authority over Carlco's corporate accounts. From 1984 through 1988, Lisle's brother was Carlco's president, and Lisle's wife was its vice-president. In 1989, Lisle was Carlco's president, his wife was its secretary, and their son was its vice-president. Although the government has not cited to any abuse of these positions or the signatory authority held by Lisle and his family, aside from the $3,000 payment already discussed, it argues that the mere fact that they could have accessed Carlco's accounts means that it was their alter ego — the fact that they chose not to does not diminish the control they had over Carlco, and thus negate a finding of alter ego.

Again, this speculation cannot sustain a finding of alter ego. It was undisputed that Lisle was managing the assets of Carlco for the purpose of investing in municipal bonds. That he used his personal residence as the address for Carlco and that he had signatory authority over its assets then means little. That said, this is evidence that Lisle exercised a great deal of control over Carlco.

The government also relies on the fact that in a letter from Administration Co. distributing money to Carlco, Lisle and Carlco were referred to interchangeably. The letter stated that a check payable to Carlco, Inc. was enclosed and that "your 45% comes to $63,000." The letter also states that another check represents interest earned "through the time that we sent the various monies to you as a transfer of funds." The government contends this indicates that Lisle was the true owner of Carlco's funds. While this is suggestive of a failure to distinguish between the corporation and the individual, it provides little support for finding an alter ego.

The court's final piece of evidence is that a trust created by Kanter for the benefit of Lisle's wife and children purchased 300 preferred shares of Carlco. According to Kanter, this was done to disperse IRA's ownership of Carlco so that Carlco was no longer included in IRA's consolidated group for tax purposes. Kanter testified that this was necessary so that Carlco could maximize the tax benefits of its investments in municipal bonds. He stated that the family trusts were sold the shares so that Lisle's investment decisions would not be second-guessed by a minority owner. The Tax Court rejected this argument, noting that IRA still controlled 100% of the 1000 shares of common stock and thus its ability to second-guess Lisle remained unimpeded. The court reasoned that the preferred stock was issued to give Lisle an effective ownership interest in Carlco through his family trust, thereby facilitating Lisle's use of Carlco as an incorporated pocketbook.

The court found that the restrictions on the preferred stock were ambiguous and speculated that it could have been worth much more than the $1650 the trust paid for it. The certificates authorizing the preferred shares granted the board of directors authority to fix the preferences of the preferred shares, but the record does not contain any evidence of resolutions passed by the board. The stock certificates contain several restrictions, two of which were important to the court: "Redemption by company at any time upon 10 days notice at 105 percent," and "Priority on liquidation equal to original purchase price per share." The court speculated that "105 percent" could refer to retained earnings rather than "par value" as Lisle asserted. Likewise, "original purchase price" could include not only the purchase price, but a share of the remaining assets or a value for the uncompensated services of Lisle. This inference is too tenuous and taxes the plain meaning of the restrictions.

Considering the evidence in light of the factors we have articulated in the past and the totality of the circumstances, the court erred when it concluded that clear and convincing evidence established that Carlco was Lisle's alter ego. Lisle personally had no interest in Carlco. Although his wife and children were beneficiaries of a trust which owned some preferred stock, the government failed to prove that the value of that stock was not minimal in comparison to the total assets of Carlco. While Lisle did exercise complete control over Carlco's assets in his role as manager, there is only the slimmest evidence that he used Carlco's assets for personal purposes. Control alone will not support a finding of an alter ego relationship.

That is not to say that there is no evidence to support a finding of alter ego. That Kanter and Ballard both used BWK's and TMT's assets for their personal benefit suggests Lisle owned Carlco's assets and could have done the same. But it is also plausible that Lisle took his position as manager of Carlco's assets seriously, and honestly managed the money entrusted to him. The potential for a manager who has control over a company's assets to abuse his position cannot alone be the basis for abandoning the corporate form. As we have explained, the other evidence relied on by the court is also insufficient. In sum, there is no clear and convincing evidence to warrant the conclusion that Carlco was the alter ego of Lisle.

C.

In addition to the alter ego theory, the court alternatively relied on the assignment of income doctrine to attribute forty-five percent of the income from the Five to Lisle. The basic principle of the assigned income doctrine is that the person who earns the income is responsible for the tax on that income, regardless of whether the individual assigns that income to another person or corporation. "[I]t is a well established rule that income is taxed to the person who earns it, regardless of any `anticipatory arrangements and contracts however skillfully devised to prevent the salary when paid from vesting even for a second in the man who earned it.'" The government argues, and the court found, that Lisle was the actual earner of forty-five percent of the income from the Five that was paid to the Kanter corporations, and thus Lisle is liable for the taxes on his share of that income.

The Tax Court's opinion, as complete as it is in most respects, does little to distinguish among the actions of Lisle, Ballard, and Kanter. Rather, the court refers to them collectively as the "petitioners." This common brush of the individual actions of the taxpayers bleeds the court's analysis, as it presupposes the question to be answered in charging that the actions of Kanter and Ballard are attributable to Lisle. On this record, it is not sufficiently clear that Lisle, rather than Kanter, earned the income from the Five.

The Tax Court begins its discussion of the assignment of income doctrine by stating that:

The record shows that Kanter was in control of negotiations concerning the amount of commissions and that he earned those commissions by performing the work for them. He directed members of the Five where to make payments. The various entities were entirely subject to Kanter's control: he set up the entities, and managed the entities in that Meyers, Schott, Weisgal, and Freeman were subject to his control. There is no evidence that IRA, Holding Co., or any of the other entities earned these funds.

These findings point to Kanter, not Lisle, as the earner of the payments from the Five.

The court continues by stating that "[p]etitioners handled the accounts as if they were their own, moving funds around from location to location and using the funds for their personal benefit." As we discussed, there is little evidence to support the claim that Lisle used Carlco funds for his personal benefit. As it did in its discussion of the alter ego doctrine, the court attributed behavior to Lisle for which Ballard and Kanter are responsible — it is they, and not Lisle, who used the IRA subsidiaries they managed for personal gain. This clustering of the "petitioners," as opposed to Kanter as the mover of the funds from location to location is contrary to the evidence and the court's own findings:

Kanter did virtually all of the planning and implementing of the transactions. The officers, directors, and trustees signed documents and entered transactions as Kanter directed including issuing and redeeming stock, liquidating corporations, purchasing and selling stock, distributing funds, and executing contracts and agreements. There is very little evidence that IRA or the other entities had anything to do with these transactions other than to be the named recipients of the checks.

There is little evidence that Lisle had anything to do with the shuffling of money from the Five between the various Kanter entities.

Other findings by the court also point to Kanter as the true earner of the income, not Lisle. The court noted that "[e]ven though the payments were made to various corporations, it is clear that the other parties to the transactions viewed IRA, Holding Co., their subsidiaries, and Kanter as one and the same." Schaffel paid IRA half of all commissions he earned on deals involving Prudential. The court found that all payments to IRA, including those from Schaffel, were split 45-45-10 among Ballard, Lisle, and Kanter. Yet when Lisle moved to Travelers, Schaffel initially refused to continue the payments on business he was now receiving from Travelers. Kanter protested, arguing that he was still getting business as a result of Kanter's introduction of Schaffel to Lisle. Schaffel acquiesced, and resumed making the payments; however, they were now directed to Holding Co. rather than IRA. The court found that none of the payments to Holding Co. were ever distributed to Lisle. In arguing that IRA and Holding Co. were sham corporations, the court noted that "[i]f IRA, rather than Kanter, had been the true party in interest . . . the payments for the Travelers deals would have been paid to IRA." The court failed to explain why, if Lisle was selling his influence, he would receive forty-five percent of the commissions while at Prudential, but none when he moved to Travelers even though Schaffel was now receiving business from Travelers. This suggests that Kanter, and not Lisle, was the true earner of the commissions from Schaffel. Similarly, in discussing the purchase of the Schnitzer-PMS stock, the court stated that Schnitzer "sold it at a bargain price for Kanter's services."

The conclusion that Kanter was the true earner of the payments is consistent with the testimony of those members of the Five who testified. As the court noted, at trial, all of the witnesses associated with the Five explicitly denied that the payments were "kickbacks" or "payoffs" for Ballard's and/or Lisle's help in steering business to them. These same witnesses did confirm, however, "that they entered into these arrangements in exchange for Kanter's influence in obtaining business." It is clear that the Five sought the assistance of Kanter in obtaining business, not only through his contacts with Lisle and Ballard, but also through his numerous other business contacts, such as the Pritzker family. Thus, the evidence is far from clear and convincing that payments were for the services of Lisle, rather than Kanter.

The government argues that Kanter could not have assisted the Five without the help of Ballard and Lisle, and that Lisle and Ballard needed Kanter's assistance in laundering the money. Therefore, they agreed to the 45-45-10 split, regardless of whether Ballard or Lisle was most responsible for securing the business for Kanter's acquaintances. Thus, while Kanter negotiated the deals and directed the payments from the Five to the various Kanter corporations, Lisle and Ballard earned a portion of those payments and are responsible for taxes on their portion.

While it is true that Kanter could not have procured additional business for the Five without the assistance of influential friends such as Ballard, Lisle, and Pritzker, it does not follow that Ballard, Lisle or Pritzker therefore earned a portion of the commissions paid by the Five to Kanter. At the same time, if Kanter, Ballard, and Lisle did in fact agree to work together and split the income, they would each be liable for taxes on their share of that income. The court concluded that "it is clear from the record that Kanter, Ballard, and Lisle agreed to share and did share the money from the Prudential transactions in a 45-45-10 split." We cannot agree.

The assertion that Lisle was the true earner of forty-five percent of the payments because he received forty-five percent of the proceeds, and therefore we will assign forty-five percent of the proceeds to Lisle because he earned them is circular. Yet both the Tax Court and the government walk close to this line of reasoning. For example, while trying to prove that Lisle was selling his influence, the government relies in part on the "fact" that he eventually received forty-five percent of the payments. Later, when trying to prove that Lisle received forty-five percent of the payments as income, the government argues that he earned it by selling his influence. The question is what is the evidence of Lisle actually receiving any of the payments which the Five made to Kanter, without simply assuming that he earned them and therefore the funds were his. The government relies on several pieces of evidence, the three most substantial being Lisle's control of the assets of Carlco, the payment of consulting fees to Lisle's children, and the loans to Lisle and the trusts which benefitted Lisle's family.

We have discussed Lisle's control over Carlco and concluded that the evidence does not support the application of the alter ego doctrine. For the same reasons, we cannot agree with the court's conclusion that Lisle received forty-five percent of the payments by managing Carlco's assets. While the counter proposition is not without purchase, as with the alter ego doctrine, we simply cannot agree that there is clear and convincing evidence that the funds in Carlco belonged to Lisle. At most, Lisle used $3000 of Carlco's millions of dollars in assets for personal gain.

The government argues that in light of the other payments to Lisle from IRA which we will discuss, we can assume that Lisle owned the funds in Carlco, but unlike Ballard and Kanter with respect to TMT and BWK, simply chose not to use them. The government analogizes the situation to a bank account which an individual owns but from which he chooses not to withdraw any money. While it is true that failure to use the funds in the account is not proof that the individual is not the owner of those funds, the analogy fails when the evidence of ownership of the account is lacking in the first place. Here, we do not find clear and convincing evidence that Lisle was the true owner of Carlco's assets.

The government and the court also rely on the payments to Lisle's children and the trusts which benefitted Lisle's family as evidence that a portion of the money paid to IRA was in fact earned by Lisle. We have discussed the payments to Lisle's children, and as we noted, there is no evidence of Lisle's involvement in authorizing them. It is true that Kanter, Ballard, and Lisle could have had an arrangement to split the funds from the Weaver deal, and accomplished this by funneling some of the money through KWJ to Lisle and Ballard's children. Alternatively, it is also possible that Kanter hired the children of a close family friend who both had some real estate experience to bring potential real estate investments to his attention. In fact, one of Lisle's children was otherwise employed full-time by Kanter. The question is whether the evidence supports the conclusion that Lisle was funneling money from the Weaver deal to his children.

When the Weaver transaction is examined closely, the evidence is far from clear that Lisle and Ballard were receiving kickbacks laundered through Kanter to their children. First, one must assume that Ballard and Lisle agreed with Weaver to split the commission, and that after meeting Kanter, decided to use him to help launder the money years later. Weaver received a commission from Hyatt for his help in getting Hyatt the Embarcadero contract in 1970 or 1971. Ballard and Lisle did not meet Kanter until 1972. The agreement between Kanter and Weaver for the purchase of the Hyatt commission payments through the purchase of KWJ corporation was not reached until 1976, and the purchase did not take place until 1979. The funds from commission payments were not disbursed to Carlco, TMT, and BWK until 1983, although small payments to Lisle's and Ballard's children began in 1982. If Lisle had agreed to sell his influence to Weaver for a cut of the commission Weaver received from Hyatt, it is not credible that he would rely on a deal formulated years later by Kanter, a man he did not even know at the time he allegedly sold his influence to Weaver. And if the commission paid to KWJ belonged to Lisle, Ballard, and Kanter in a 45-45-10 split, there is no explanation for why KWJ made payments to the children which were not in proportion to this split. There may be some overarching rationalization for these inconsistencies, but we have not discerned it.

Nor does the government explain how Lisle was able to get the other partners in the Embarcadero deal to give the contract to Hyatt. Lisle represented Prudential, only one of the prominent partners involved in the deal. Speculation that he threatened to block the deal unless Hyatt paid the commission is not supported by the record. The government offered no evidence to refute the testimony that Hyatt got the deal because it submitted the only bid.

Finally, the government attempts to rely on the court's finding that a letter from Weaver to Kanter forwarding a commission check from Hyatt which instructs Kanter to "deposit and issue appropriate checks to the participants" refers to Kanter, Ballard, and Lisle. Pursuant to the sale of KWJ to Kanter, Weaver was to receive thirty percent of the commission, and KWJ, now owned by a Kanter corporation, was to receive the remainder. There is no support in the record for the conclusion that "participants" refers to Lisle, Ballard, Kanter, and Weaver, as opposed to just Kanter and Weaver.

We conclude that while the government's theory of Lisle's role in this elaborate scheme is plausible, it is not proved by clear and convincing evidence. We cannot agree that because Lisle's children received $1,000 a month in questionable commissions from KWJ, Lisle was the true earner of forty-five percent of all the payments from the Five.

Finally, the government argues that Lisle and trusts established for the benefit of Lisle's wife and children borrowed money from Kanter corporations and did not repay it. Unpaid loans to Lisle and these trusts total more than $220,000 between 1974 and 1990. As no interest was ever paid to the Kanter entities for these loans, and a large portion of the amount was written off as worthless in 1987, the court determined that these were never legitimate loans, but were part of the scheme to distribute kickbacks to Lisle and Ballard.

Once again, we cannot agree that clear and convincing evidence supports this conclusion. The government does not explain how these loans can be characterized as kickbacks when they began before the Five ever made any payments to Kanter's corporations. In fact, many of the largest loans to the Ballard and Lisle family trusts occurred before the first payment by a member of the Five in 1977. Of the $220,000 loaned to Lisle and his family's trusts over the sixteen years cited by the court, $68,000 was loaned before 1977. While this does not mean that later loans were not distributions of income from the Five, it casts serious doubt on that conclusion.

We are left with the definite and firm conviction that a mistake has been committed. We cannot agree that the government has established by clear and convincing evidence that Lisle earned forty-five percent of the payments from the Five, and therefore the assignment of income doctrine cannot be applied. As we will explain, these findings are not without evidentiary support. It is rather that to support a penalty for fraud, the government bore the burden of proving a deficiency in income by clear and convincing evidence and we cannot agree with the Tax Court that this burden has been met.

D.

We now turn to the court's final theory for attributing forty-five percent of the payments from the Five to Lisle, the application of 26 U.S.C. § 482. The court applied § 482, reasoning that the petitioners, Lisle, Ballard, and Kanter, distributed income among the various Kanter corporations in order to avoid paying taxes on the income. Thus, the court reasoned that under § 482, that income could be reallocated from the Kanter corporations to the petitioners. Even if § 482 can be applied to individuals, the absence of clear and convincing proof that Lisle was the earner of forty-five percent of the payments from the Five makes § 482 inapplicable.

IV.

We next turn to the court's findings regarding the deficiency, independent of fraud. Here, the government enjoys a presumption of correctness for the deficiencies stated in the original notices, and must prove any additional deficiencies asserted before the Tax Court only by a preponderance of the evidence.

We have already discussed the government's case at length. From our review of the record, we cannot say that the court clearly erred in finding that the government established the deficiencies by a preponderance of the evidence. Thus, even if Lisle shifted the burden of proof to the government on all deficiencies as appellants argue, the government has met its burden. While much of the evidence is equivocal, when the burden of proof is by a preponderance of the evidence, we will not find clear error if the evidence supports either of two theories.

We affirm the Tax Court's finding that Lisle owes additional taxes for the years 1987, 1988, and 1989. However, without a finding of fraud, the assessment of taxes for 1984 is barred by the statute of limitations. The court relied on 26 U.S.C. § 6501(c)(1), which provides that the tax may be assessed at any time in the case of fraud. With no finding of fraud the limitations period is three years.

The government argues that appellants waived this issue because they failed to renew the argument in their opening brief. While we may in our discretion decline to consider issues not raised in an initial brief, we choose to address the issue here. The appellants raised the issue of statute of limitations in the Tax Court and renewed their argument that the court's finding of fraud was erroneous before this court. The government suffers no prejudice by the absence of the issue in the opening brief.

V.

Appellants also raise a due process challenge to the Tax Court's application of Tax Court Rule 183. Briefly stated, appellants argue that it is a violation of their due process rights for the Tax Court to not make the Special Trial Judge's initial report available to the parties. This issue was recently addressed by the Seventh Circuit in Kanter v. Commissioner of Internal Revenue, and by the Eleventh Circuit in Ballard v. Commissioner of Internal Revenue, the companion cases to this one. We find the reasoning of the Seventh and Eleventh Circuits direct and persuasive, and adopt it here. We hold that the application of Rule 183 in this case did not violate Appellants' due process rights.

VI.

In conclusion, we find that the Tax Court clearly erred when it concluded that the government had proved by clear and convincing evidence that Lisle had earned forty-five percent of the payments from the Five and therefore owed taxes on that income. Without clear and convincing evidence of a deficiency, the element of fraudulent intent is not reached, although in this case the two are nearly indistinguishable. We therefore reverse the Tax Court's finding of fraud. As a result of our reversal of the Tax Court's finding of fraud, we also reverse its conclusion that § 6501(c)(1) applies and that the three-year statute of limitations does not bar the assessment of taxes for 1984. We affirm the Tax Court's ruling sustaining the assessment of a deficiency for years 1987, 1988, and 1989. We remand the case to the Tax Court for the limited purpose of recalculating the deficiencies and additions to tax, consistent with this opinion. REVERSED in part, AFFIRMED in part, and REMANDED.