McKEAGUE, J., delivered the opinion of the court.
ROGERS, J. (pp. 784 - 788), delivered a separate concurring opinion.
MOORE, J. (pp. 788-791), delivered a separate dissenting opinion.
McKEAGUE, Circuit Judge.Indmar Products Co., Inc. (“Indmar”) appeals the decision of the Tax Court to disallow interest deductions the company claimed for tax years 1998-2000, and to assess accuracy-related tax penalties for those years. The interest deductions relate to a number of advances made to Indmar by its majority stockholders over several years. Indmar argued at trial that the advances were legitimate loans made to the company, and thus it could properly deduct the interest payments made on these advances under 26 U.S.C. § 163(a). The Tax Court, following the position taken by the Commissioner of Internal Revenue (the “Commissioner”), disagreed, concluding that the advances were equity contributions and therefore the company could not deduct any purported interest payments on these advances. The court imposed penalties on Indmar based on the deductions. Indmar Prods. Co., Inc. v. Comm’r., 2005 WL 419332, T.C.M.2005-32, 2005 T.C.M. LEXIS 31.
Upon review of the record, we conclude that the Tax Court clearly erred in finding *774the advances were equity. The Tax Court failed to consider several factors used by this court for determining whether advances are debt or equity, ignored relevant evidence, and drew several unsupported inferences from its factual findings. We reverse and find that the stockholder advances were bona fide debt.
I. BACKGROUND
A. Stockholder Advances to Indmar
Indmar, a Tennessee corporation, is a marine engine manufacturer. In 1973, Richard Rowe, Sr., and Marty Hoffman owned equal shares of Indmar. In 1987, after Hoffman passed away, Richard and his wife, Donna Rowe, together owned 74.44% of Indmar, with their children and children’s spouses owning the rest.
By all accounts, Indmar has been a successful company. From 1986 to 2000, Ind-mar’s sales and costs-of-goods sold increased from $5m and $3.9m to $45m and $37.7m, respectively. In addition, Ind-mar’s working capital (current assets minus current liabilities) increased from $471,386 to $3.8m. During this period, Ind-mar did not declare or pay formal dividends.
Since the 1970s, Indmar’s stockholders have advanced funds to it, receiving a 10% annual return in exchange. Hoffman started the practice in the 1970s. Beginning in 1987, the Rowes (as well as their children) began to make advancements on a periodic basis. Indmar treated all of the advances as loans from stockholders in the corporate books and records, and made monthly payments calculated at 10% of the advanced funds. Indmar reported the payments as interest expense deductions on its federal income tax returns. Consistent with Indmar’s reporting, the Rowes reported the payments as interest income on them individual income tax returns.
The parties did not initially document the advances with notes or other instruments. Beginning in 1993, the parties executed notes covering all of the advances at issue. Specifically, Indmar executed a promissory note in 1993 with Donna Rowe for $201,400 (i.e., her outstanding balance). The note was payable on demand and freely transferable, had no maturity date or monthly payment schedule, and had a fixed interest rate of 10%. In 1995, Ind-mar executed a similar promissory note with Richard Rowe for $605,681 (i.e., his outstanding balance). In 1998, when the outstanding transfers totaled $1,222,133, Indmar executed two line of credit agreements with the Rowes for $lm and $750,000. The line of credit agreements provided that the balances were payable on demand and the notes were freely transferable. In addition, the agreements provided a stated interest rate of 10% and had no maturity date or monthly payment schedule. None of the advances were secured.
Repayments of the advances were paid on demand, based on the needs of the stockholders, and not subject to set or predetermined due dates. The record indicates that between 1987 and 2000, the total advance balances ranged from $634,000 to $1.7m, and Indmar made purported interest payments between $45,000 and $174,000 each year.
The parties structured the advances as demand loans to give the Rowes flexibility as creditors. Moreover, as demand loans, the advances were treated by the Rowes as short-term debt under Tennessee law, thereby excepting interest payments from a 6% state tax on dividends and interest on long-term debts. TenmCode Ann. §§ 67-2 — 101(l)(B)(i), 67-2-102 (2005). Indmar, however, reported the advances as long-term liabilities on its financial statements to avoid violating loan agreements with *775First Tennessee Bank (“FTB”), its primary creditor, who required a minimum ratio of current assets to current liabilities.
In order to reconcile the treatment and execution of the advances as demand loans versus listing them as long-term debt in its financial reports, Indmar received waivers from the Rowes agreeing to forego repayment on the notes for at least 12 months. From 1989 to 2000, the notes to Indmar’s financial statements disclosed that “The stockholders have agreed not to demand payment within the next year,” and in 1992 and 1993, the Rowes signed written agreements stating that they would not demand repayment of the advances. Indmar did all of this under the direction of its accountant.
Despite the annual waivers, the Rowes demanded and received numerous partial repayments of the advances. Specifically, in 1994 and 1995, Richard Rowe demanded repayment of $15,000 and $650,000, respectively, to pay his taxes and purchase a new home. He also demanded repayment of $84,948, $80,000, $25,000, and $70,221 from 1997-2000 to pay litigation expenses, boat repairs, and tax expenses. Donna Rowe demanded repayment of $180,000 in 1998 for boat repairs. The Rowes made additional advances in 1997 and 1998 of $500,000 and $300,000, respectively. The balance of notes payable to stockholders on December 31, 2000, totaled $1,166,912.
As Indmar was a successful, profitable company, numerous banks sought to lend money to it. FTB worked hard to retain Indmar’s business, made funds immediately available upon request, and was willing to lend Indmar 100% of the stockholder advances.
In its loan agreements with Indmar, FTB required the company to subordinate all transfers, including stockholder advances, to FTB’s loans. FTB did not strictly enforce the subordination provision, however, as Indmar repaid — with FTB’s knowledge — some of the stockholder advancements at the same time FTB loans remained outstanding. As an example, when Richard demanded repayment of $650,000 to purchase a new home, Indmar borrowed the entire amount from FTB at 7.5% (the prime lending rate was 8.75%). Indmar secured the loan with inventory, accounts and general intangibles, equipment, and the personal guarantee of the Rowes. Richard Moody, the FTB lending officer who worked with Indmar on the loan, testified that he knew Indmar used the proceeds to repay Richard. Indmar had loans outstanding with FTB at the time.
As stipulated by the parties, the prime lending rate ranged from a low of 6% to a high of 10.5% between 1987-1998. In 1997, Indmar and FTB executed a promissory note for $lm that was modified in 1998. The interest rate on the note (7.85%) was below the prime lending rate. Indmar also had a collateralized line of credit with FTB. Similar to the stockholder advances, the bank line of credit was used for short-term working capital. FTB charged the following rates for the secured line of credit:
1995 9%
1996 8.75%
1997 9%
1998 8%
1999 8.75%
2000 9.5%
(rates as of December 31st of each year).
B. Claimed Deductions at Issue
On its tax returns for 1998-2000, Ind-mar claimed deductions for the purported interest payments paid on the stockholder advances. The Commissioner issued a notice of deficiency. Indmar filed a petition in the Tax Court challenging the Commis*776sioner’s decision. After trial, the Tax Court concluded that the advances did not constitute genuine indebtedness and thus the payments to the stockholders were not deductible. The Tax Court calculated a total tax deficiency of $123,735 and assessed $24,747 in penalties. Indmar timely appealed.
II. LEGAL ANALYSIS
A. Determining Whether Advance Is Debt or Equity
As a general matter, the Commissioner’s determination of a deficiency is entitled to a presumption of correctness. It is the taxpayer’s burden to prove the determination to be incorrect or arbitrary. Ekman v. Comm’r, 184 F.3d 522, 524 (6th Cir.1999).
The basic question before us is whether the advances made to the company by the stockholders were loans or equity contributions. Under 26 U.S.C. § 163(a), a taxpayer may take a tax deduction for “all interest paid or accrued ... on indebtedness.” There is no similar deduction for dividends paid on equity investments. Thus, if the advances were loans, the 10% payments made by Indmar to the Rowes were “interest” payments, and Indmar could deduct these payments. If, on the other hand, the advances were equity contributions, the 10% payments were constructive dividends, and thus were not deductible.
Over the years, courts have grappled with this seemingly simple question in a wide array of legal and factual contexts. The distinction between debt and equity arises in other areas of federal tax law, see, e.g., Roth Steel Tube Co. v. Comm’r, 800 F.2d 625, 629-30 (6th Cir.1986) (addressing the issue in the context of the deductibility of advances as bad debt under 26 U.S.C. § 166(a)(1)), as well as bankruptcy law, see, e.g., In re AutoStyle Plastics, Inc., 269 F.3d 726, 750 (6th Cir.2001). The Second Circuit set out the “classic” definition of debt in Gilbert v. Commissioner: “an unqualified obligation to pay a sum certain at a reasonably close fixed maturity date along with a fixed percentage in interest payable regardless of the debtor’s income or lack thereof.” 248 F.2d 399, 402 (2d Cir.1957). “While some variation from this formula is not fatal to the taxpayer’s effort to have the advance treated as a debt for tax purposes, ... too great a variation will of course preclude such treatment.” Id. at 402-03. The question becomes, then, what is “too great a variation”?
To determine whether an advance to a company is debt or equity, courts consider “whether the objective facts establish an intention to create an unconditional obligation to repay the advances.” Roth Steel, 800 F.2d at 630 (citing Raymond v. United States, 511 F.2d 185, 190 (6th Cir.1975)). In doing so, courts look not only to the form of the transaction, but, more importantly, to its economic substance. See, e.g., Fin Hay Realty Co. v. United States, 398 F.2d 694, 697 (3d Cir.1968) (“The various factors ... are only aids in answering the ultimate question whether the investment, analyzed in terms of its economic reality, constitutes risk capital entirely subject to the fortunes of the corporate venture or represents a strict debtor-creditor relationship.”); Byerlite Corp. v. Williams, 286 F.2d 285, 291 (6th Cir.1960) (“In all cases, the prevailing consideration is that artifice must not be exalted over reality, whether to the advantage of the taxpayer, or to the government.”).
The circuit courts have not settled on a single approach to the debt/equity question. We elucidated our approach *777in Roth Steel, setting out eleven non-exclusive factors for courts to consider:
(1) the names given to the instruments, if any, evidencing the indebtedness; (2) the presence or absence of a fixed maturity date and schedule of payments; (3) the presence or absence of a fixed rate of interest and interest payments; (4) the source of repayments; (5) the adequacy or inadequacy of capitalization; (6) the identity of interest between the creditor and the stockholder; (7) the security, if any, for the advances; (8) the corporation’s ability to obtain financing from outside lending institutions; (9) the extent to which the advances were subordinated to the claims of outside creditors; (10) the extent to which the advances were used to acquire capital assets; and (11) the presence or absence of a sinking fund to provide repayments.
800 F.2d at 630. No single factor is controlling; the weight to be given a factor (if any) necessarily depends on the particular circumstances of each case. Id.; see also Universal Castings Corp., 37 T.C. 107, 114, 1961 WL 1057 (1961) (“It is not enough when examining such a precedential checklist to test each item for its presence or absence, but it is necessary also to weigh each item.”), aff'd, 303 F.2d 620 (7th Cir.1962). In essence, the more a stockholder advance resembles an arm’s-length transaction, the more likely it is to be treated as debt. AutoStyle Plastics, 269 F.3d at 750.
B. Standard of Review
We review the Tax Court’s factual findings for “clear error” and its application of law de novo. Vision Info. Servs., LLC v. Comm’r, 419 F.3d 554, 558 (6th Cir.2005); Holmes v. Comm’r, 184 F.3d 536, 543 (6th Cir.1999). The circuits are split on whether the debt/equity question is one of fact or law, or a mixed question of fact and law. Jaques v. Comm’r, 935 F.2d 104, 106-07 (6th Cir.1991) (collecting cases); Roth Steel, 800 F.2d at 629 (collecting cases). Earlier panels of this court have held that the question is one of fact. Jaques, 935 F.2d at 107; Roth Steel, 800 F.2d at 629; Smith v. Comm’r, 370 F.2d 178, 180 (6th Cir.1966).1 Accordingly, we review the Tax Court’s findings for clear error.
Under the clear error standard, we give deference to the Tax Court’s factual findings and inferences drawn from those findings. Holmes, 184 F.3d at 543 (citations omitted). “Where there are two *778permissible views of the evidence, the fact-finder’s choice between them cannot be clearly erroneous.” Anderson v. City of Bessemer City, 470 U.S. 564, 574,105 S.Ct. 1504, 84 L.Ed.2d 518 (1985). Moreover, we afford even greater discretion to any credibility determinations made by the Tax Court. Id. at 575, 105 S.Ct. 1504.
Yet, while our review is deferential, it is not nugatory. As we explained in Holmes, “an appellate court may reverse a lower forum’s factual finding for clear error when, even though the record contains some evidence in support of the finding, consideration of the overall evidence leaves the reviewing court ‘with the definite and firm conviction that a mistake has been committed.’” 184 F.3d at 543 (quoting Concrete Pipe & Prods, of Cal., Inc. v. Constr. Laborers Pension Trust, 508 U.S. 602, 622, 113 S.Ct. 2264, 124 L.Ed.2d 539 (1993); Anderson, 470 U.S. at 573, 105 S.Ct. 1504). If the Tax Court’s factual findings “were induced by an erroneous view not supported by substantial evidence or made without properly taking into account substantial evidence to the contrary,” clear error exists. People Helpers Found., Inc. v. City of Richmond, 12 F.3d 1321, 1329 (4th Cir.1993). Moreover, as explained by one of our sister courts which also applies the “clear error” standard to the debt/equity question,
[The lower court] must apply relevant legal principles, and a factual determination made in disregard of the applicable principles of law, or made through gross overemphasis on one relevant principle to the exclusion of others, will be reversed because of the application of an improper legal standard.... The characterization of the ultimate issue as one of fact, and the resulting diminution of the scope of review, does not, of course, affect the requirement that the legally relevant factors be applied in making the determination.
Piedmont Minerals Co. v. United States, 429 F.2d 560, 562 n. 4 (4th Cir.1970); see also Flying J, Inc. v. Comdata Network, Inc., 405 F.3d 821, 829 (10th Cir.2005) (“Whether the [lower] court failed to consider or accord proper weight or significance to relevant evidence are questions of law we review de novo.”) (quoting Harvey ex rel. Blankenbaker v. United Transp. Union, 878 F.2d 1235, 1244 (10th Cir.1989) (citing Pullman-Stan’d v. Swint, 456 U.S. 273, 291-92, 102 S.Ct. 1781, 72 L.Ed.2d 66 (1982))).
With these principles in mind, we now turn to the Tax Court’s decision in this case.
C. Roth Steel Factors
After discussing some, but not all, of the Roth Steel factors, the Tax Court concluded that the Rowes’ advances were equity contributions. Specifically, it found the following factors weighed in favor of equity: (i) Indmar did not pay any formal dividends (although this is not one of the Roth Steel factors); (ii) there was no fixed maturity date or obligation to repay; (iii) repayment came from corporate profits and would not be paid if there were not sufficient profits; (iv) advances were unsecured; (v) there was no sinking fund; and (vi) at the time advances were made, there was no unconditional and legal obligation to repay. The court found that several factors weighed in favor of debt: (i) Ind-mar reported the advances on its federal income tax returns as interest expenses; (ii) external financing was available; (iii) Indmar was adequately capitalized; (iv) the advances were not subordinated to all creditors; and (v) the Rowes did not make the advances in proportion to their respective equity holdings. The court concluded that the factors favoring equity “certainly *779outweigh” those favoring debt. Indmar, 2005 T.C.M. LEXIS 31, at *15.
As explained below, we find that the Tax Court clearly erred in concluding that the advances were equity contributions rather than bona fide debt. The Tax Court failed to consider several Roth Steel factors. It also did not address in its analysis certain uncontroverted testimony and evidence upon which the parties stipulated. Consideration of all of the record evidence in this case leaves us “with the definite and firm conviction that a mistake has been committed.” Holmes, 184 F.3d at 543.
1. Fixed Rate of Interest and Interest Payments
The first factor to which we look is whether or not a fixed rate of interest and fixed interest payments accompanied the advances. Roth Steel, 800 F.2d at 631. The absence of a fixed interest rate and regular payments indicates equity; conversely, the presence of both evidences debt. Id.; 7 Mertens Law of Fed. Income Tax’n § 26:28 (“A fixed interest rate is indicative of a deductible interest payment.”) (collecting cases). In its findings of fact, the Tax Court determined that the advances were made with a 10% annual return rate. The court also found that Indmar made regular monthly interest payments on all of the advances.
The fixed rate of interest and regular interest payments indicate that the advances were bona fide debt. In its analysis, however, the Tax Court took a different view. Rather than analyzing these facts within the Roth Steel framework (i.e., as objective indicia of debt or equity), the Tax Court focused instead on why the Rowes made the advancements: it concluded that the Rowes “characterized the cash transfers as debt because they wanted to receive a 10-percent return on their investment and minimize estate taxes.” Indmar, 2005 T.C.M. LEXIS 31, at *11. Yet, neither of these intentions is inconsistent with characterizing the advances as loans.
For tax purposes, it is generally more important to focus on “what was done,” than “why it was done.” United States v. Hertwig, 398 F.2d 452, 455 (5th Cir.1968). “In applying the law to the facts of this case, ... it is ‘clear that the objective factors ... are decisive in cases of this type.’ ” Raymond, 511 F.2d at 191 (quoting Austin Village, 432 F.2d at 745). It is largely unremarkable that the Rowes wanted to receive a return from their advances. Most, if not all, creditors (as well as equity investors) intend to profit from their investments. Bordo Prods. Co. v. United States, 201 Ct.Cl. 482, 476 F.2d 1312, 1322 (1973). As long as the interest rate is in line with the risks involved, a healthy return on investment can evidence debt.
Of course, “[ejxcessively high rates would ... raise the possibility that a distribution of corporate profits was being disguised as debt. Were such the purpose of an exorbitant interest rate, the instrument involved would probably not qualify as debt in form.” Scriptomatic, Inc. v. United States, 555 F.2d 364, 370 n. 7 (3d Cir.1977) (citing William T. Plumb, Jr., The Fed. Income Tax Significance of Corporate Debt: A Critical Analysis & A Proposal, 26 Tax L.Rev. 369, 439-40 (1971)). The Tax Court found that the 10% rate exceeded the federal prime interest rate during most of the period at issue, as well as the rate charged by FTB on several of its loans to Indmar.
The record indicates that the 10% rate was not an “exorbitant interest rate” under the circumstances. Indmar had a collateralized line of credit with FTB, which, *780similar to the stockholder advances, was available for short-term working capital. The rate charged by FTB for the line of credit ranged between 8%-9.5%, a rate not much below the fixed rate of 10% charged by the Rowes. The rate differential makes financial sense when considering the differences in security — the FTB line of credit was secured while the Rowes’ advances were not.2
As for the Rowes’ desire to minimize their estate taxes, this also offers little to the analysis. “Tax avoidance is entirely legal and legitimate. Any taxpayer ‘may so arrange his affairs that his taxes shall be as low as possible; he is not bound to choose that pattern which will best pay the Treasury; there is not even a patriotic duty to increase one’s taxes.’ ” Estate of Kluener v. Comm’r, 154 F.3d 630, 634 (6th Cir.1998) (quoting Helvering v. Gregory, 69 F.2d 809, 810 (2d Cir.1934) (L.Hand, J.)). The desire to avoid or minimize taxes is not itself directly relevant to the question whether a purported loan is instead an equity investment. Rather, such a desire is only tangentially relevant, by acting as a flag to the Commissioner and courts to look closely at the transaction for any objective indicia of debt.
Far from proving the Commissioner’s position, the existence and consistent payment of a fixed, reasonable interest rate strongly supports the inference that the advances were bona fide loans.
2. Written Instruments of the Indebtedness
“The absence of notes or other instruments of indebtedness is a strong indication that the advances were capital contributions and not loans.” Roth Steel, 800 F.2d at 631. In its analysis, the Tax Court found that Indmar “failed to establish that, at the time the transfers were made, it had the requisite unconditional and legal obligation to repay the Rowes (e.g., the transfers were not documented).” Indmar, 2005 T.C.M. LEXIS 31, at *15 (emphasis added).3
The Tax Court focused on only half the story. For years 1987-1992, the Rowes did make advancements without executing any notes or other instruments. Beginning in 1993, and for all the tax years at issue in this case, the parties executed notes of loans and lines of credit covering all of the advances at issue, as the Tax Court noted in its findings of fact. Yet, in its analysis of the Roth Steel factors, the Tax Court was silent as to the subsequent execution of notes. After-the-fact consolidation of prior advances into a single note can indicate that the advances were debt rather than equity contributions. See, e.g., Dev. Corp. of Am. v. Comm’r, T.C.M.1988-127, 1988 T.C.M. LEXIS 155, at *11. The Tax Court erred by focusing on the initial lack of documentation without addressing the subsequent history of executed notes.4
*7813. Fixed Maturity Date and Schedule of Payments
“The absence of a fixed maturity date and a fixed obligation to repay indicates that the advances were capital contributions and not loans.” Roth Steel, 800 F.2d at 631. Based on the Rowes’ waivers, the Tax Court concluded that there was no fixed maturity date or fixed obligation to repay. While correct, we find that this factor carries little weight in the final analysis. The parties structured the advances as demand loans, which had ascertainable (although not fixed) maturity dates, controlled by the Rowes. Piedmont Minerals, 429 F.2d at 563 n. 5 (“The absence of a fixed maturity date is a relevant consideration, but it is far from controlling. The maturity of a demand note is always determinable by its holder.”). Furthermore, the temporary waiver of payment does not convert debt into equity “since [the stockholders] still expected to be repaid.” AutoStyle Plastics, 269 F.3d at 751.
Where advances are documented by demand notes with a fixed rate of interest and regular interest payments, the lack of a maturity date and schedule of payments does not strongly favor equity. To give any significant weight to this factor would create a virtual per se rule against the use of demand notes by stockholders, even though “[m]uch commercial debt is evidenced by demand notes.” Piedmont Minerals, 429 F.2d at 563 n. 5; see also AutoStyle Plastics, 269 F.3d at 750 (cautioning against a “rigid” rule that the factor always indicates equity).
4. The Source of Repayments
“An expectation of repayment solely from corporate earnings is not indicative of bona fide debt regardless of its reasonableness.” Roth Steel, 800 F.2d at 631 (emphasis added). Repayment can generally come from “only four possible sources ...: (1) liquidation of assets, (2) profits from the business, (3) cash flow, and (4) refinancing with another lender.” Bordo Prods., 476 F.2d at 1326 (quoting Plumb, supra, at 526).
The Tax Court found that the “source of repayments” factor favored equity. It relied upon Richard Rowe’s testimony that Indmar was expected to make a profit and that repayment “has to come from corporate profits or else the company couldn’t pay for it.” Indmar, 2005 T.C.M. LEXIS 31, at *14. The full colloquy from the testimony, however, is more equivocal:
Q. ... [A]t the time that you made these advances, were you anticipating that the repayment was going to come from corporate profits?
A. Yes, sir. It has to come from corporate profits or else the company couldn’t pay for it. Unless it made profit — and I have always believed from the first day we started, that we were going to be profitable.
Q. Was it your understanding and intent, at the time you made these advances, that if the company was not, in fact, profitable, you would not be repaid?
A. I had no intentions of not being repaid, sir.
Q. Why is that?
A. I believe it’s me. It’s my personality.
Q. Is that because you intended to make a profit?
A. Yes, sir.
There are at least two plausible ways to read this testimony. One can read it the *782way the Tax Court apparently did— Rowe’s testimony was, at best, contradictory: repayment must come from profits, but he had no intention of not being repaid, regardless of the company’s fortunes. Given the apparent contradiction, one should focus on the statement against Ind-mar’s interest: Rowe admitted that repayment of the advances “has to come from corporate profits or else the company couldn’t pay for it.” If repayment “has” to come from profits, then this would imply that repayment was tied to the company’s fortunes, suggesting the advances were equity contributions.
Another way to read the testimony, however, is that Rowe, as a small businessman and unsecured creditor, believed that full repayment of all of Indmar’s debt required a thriving, successful business, which, ultimately, required profits. In other words, struggling companies near or at bankruptcy do not repay their debts, at least not dollar for dollar. Under this reading, his testimony is consistent with debt. In fact, we sounded a similar note in an earlier decision addressing the debt/equity issue: “One who makes a loan to a corporation also takes a risk, and while he may receive evidence of an obligation, payable in any event, often such obligation is never paid. ... ‘All unsecured loans involve more or less risk.’ ” Byerlite, 286 F.2d at 292 (quoting Earle v. W.J. Jones & Son, Inc., 200 F.2d 846, 851 (9th Cir.1952)).
If there was no other evidence to support one view or the other, we could not say that the Tax Court’s reading was clearly erroneous. Credibility determinations are left to the fact finder, and our review on appeal is strictly limited. The “Tax Court ‘is not bound to accept testimony at face value even when it is uncontroverted if it is improbable, unreasonable or questionable.’ ” Lovell & Hart, Inc. v. Comm’r, 456 F.2d 145, 148 (6th Cir.1972) (quoting Comm’r v. Smith, 285 F.2d 91, 96 (5th Cir.1960)). On the other hand, the Tax Court cannot ignore relevant evidence in making its factual findings and any inferences from those findings.
Here, there is undisputed testimony by Rowe and the FTB lending officer, corroborated by stipulated evidence in the record, that clearly weighs in favor of debt on this factor. Indmar repaid a significant portion of the unpaid advances — $650,-000 — not from profits but by taking on additional debt from FTB. While the interest rate on the FTB loan was lower than 10%, Indmar had to secure the bank loan with inventory, accounts and general intangibles, equipment, and personal guarantees. Thus, Indmar repaid a significant portion of the unsecured stockholder advancements by taking on secured debt from a bank, rather than by taking the funds directly from earnings. This is important evidence that the parties had no expectation that Indmar would repay the advances “solely” from earnings. The Tax Court did not discuss or even cite this evidence in its Roth Steel analysis.
5. The Extent to Which the Advances Were Used to Acquire Capital Assets
Nor did the Tax Court address whether Indmar used the advances for working capital or capital expenditures. “Use of advances to meet the daily operating needs of the corporation, rather than to purchase capital assets, is indicative of bona fide indebtedness.” Roth Steel, 800 F.2d at 632. Richard Rowe testified that Indmar always went to a bank for funds to buy capital equipment. He also testified that all of the advances he made to Indmar were used for working capital, as opposed to capital equipment. This is uncontroverted testimony. The government points, however, to Rowe’s testimony that he ad*783vanced funds even when Indmar did not “need” the funds, and argues that this somehow cuts against his testimony that the advances were used as working capital.
The government’s argument is unpersuasive. We do not find that Rowe’s testimony on this subject was “improbable, unreasonable or questionable,” especially in the absence of the Tax Court addressing this factor in its analysis.5 A review of Indmar’s financial statements shows that it used all of the funds it received in various ways, including working capital and capital equipment expenditures. Thus, Indmar used the advances it received from the Rowes, even if not immediately upon receipt — i.e., Indmar identified a “need” for the advances at some point. There is nothing specific in the record, including Indmar’s financial statements, that suggests the advances went to purchase capital equipment as opposed to being used for working capital. Accordingly, the government’s supposition does not counter Rowe’s testimony, and this factor squarely supports a finding of debt.
6.Sinking Fund
“The failure to establish a sinking fund for repayment is evidence that the advances were capital contributions rather than loans.” Id. The Tax Court was correct to point out that the lack of a sinking fund favors equity. This factor does not, however, deserve significant weight under the circumstances. First, a sinking fund (as a type of reserve) is a form of security for debt, and the Tax Court also counted the general absence of security for the stockholder advances as favoring equity. Second, the presence or absence of a sinking fund is an important consideration when looking at advances made to highly leveraged firms. In that case, the risk of repayment will likely be high on any unsecured loans, so any commercially reasonable lender would require a sinking fund or some other form of security for repayment. Where a company has sound capitalization with outside creditors ready to loan it money (as here), there is less need for a sinking fund. See Bordo Prods., 476 F.2d at 1326.
7. The Remaining Roth Steel Factors
On the remaining Roth Steel factors, the Tax Court determined that one favored equity (lack of security for the advances) and four favored debt (the company had sufficient external financing available to it; the company was adequately capitalized; the advances were not subordinated to all creditors; and the Rowes did not make the advances in proportion to their respective equity holdings). These findings are well-supported in the record.
8. Failure to Pay Dividends
The Tax Court included in its discussion of Roth Steel a factor not actually cited in that case — Indmar’s failure to pay dividends. In support, the Tax Court cited our decision in Jaques v. Commissioner.
The relevance of Jaques to this case is questionable. That case involved the withdrawal of funds by a controlling stockholder from his closely-held corporation. The stockholder argued that the withdrawal itself was a loan. We rejected the argu*784ment, relying in part on the fact that the corporation had never issued a formal dividend, and thus the withdrawal could have been a disguised dividend. Jaques, 935 F.2d at 107-08. The situation here is the exact opposite — the stockholders were advancing money to the corporation (not from), and it is the nature of those advances that we must determine.
Had the Rowes charged Indmar an exorbitant interest rate, the lack of any formal dividends might have been relevant to showing that the payments were not interest payments, but disguised dividends. As this was not the case, see supra Section II.C.l, we do not address further the relevance, if any, of the lack of dividend payments to the debt/equity question presented here.
D. The Tax Court Committed Clear Error
To summarize, eight of the eleven Roth Steel factors favor debt. The three remaining factors suggest the advances were equity, but, as we explained above, two of the factors — the absence of a fixed maturity date and schedule of payments and the absence of a sinking fund — deserve little weight under the facts of this case. Moreover, the non -Roth Steel factor relied upop by the Tax Court — Indmar’s failure to pay dividends — has questionable relevance to our inquiry. The only factor weighing in favor of equity with any real significance— the lack of security — does not outweigh all of the other factors in favor of debt.6
Accordingly) the trial evidence, when reviewed as a whole, conclusively shows that the' Rowes’ advances to Indmar were bona fide loans. The Tax Court committed clear error in finding otherwise.
III. CONCLUSION
For the foregoing reasons, we reverse the Tax Court’s determination that the stockholders’ advances were equity contributions. We find that the advances exhibited clear, objective indicia of bona fide debt. Accordingly, we also reverse the Tax Court’s assessment of accuracy-related penalties.
. This position was not, however, uniformly embraced by all judges of this circuit. For example, in Livemois Trust v. Commissioner, the lower court found that the stockholder advances were equity contributions. On appeal, the decision was affirmed under the clearly erroneous standard. 433 F.2d 879, 883 (6th Cir.1970). In his concurring opinion, Judge McCree agreed that the lower court decision should be affirmed, but disagreed with the majority's deferential review:
My only difficulty with the opinion of the majority is its holding that we are bound by the "clearly erroneous” standard of review of Commissioner of Internal Revenue v. Duberstein, 363 U.S. 278, 80 S.Ct. 1190, 4 L.Ed.2d 1218 (1960). I believe that this is a case like Austin Village, Inc. v. United States, 432 F.2d 741 (6th Cir.1970), rev’g 296 F.Supp. 382 (N.D.Ohio 1968), and Union Planters National Bank v. United States, 426 F.2d 115, 117 (6th Cir.1970), where the objective indicia of the transaction are determinative of its legal consequences.... Thus the court in both Barthold [v. Commissioner, 404 F.2d 119, 122 (6th Cir.1968)] and this case is examining the objective manifestations of the transactions in question. The existence vel non of such indicia is of course a question of fact. But the determination of the legal effect of the existence of these indicia is, in my view, a question of law, and we should not limit our review of a District Court’s determination of such questions by the Duberstein standard.
Id. at 883 (McCree, J., concurring).
.The government took pains during trial to show that Indmar could have received a lower interest rate from FTB than its stockholders. Had the 10% rate been exorbitant, this would have been a useful line of inquiry. Under the circumstances here, the rate was not exorbitant, and whether Indmar could have received a better rate through FTB is of no import. To show bona fide debt, a taxpayer does not need to prove that it received the financially optimal rate, just a commercially reasonable one.
. The Tax Court did count in favor of debt the fact that Indmar consistently reported the payments as interest expense on its federal taxes.
. The Tax Court was troubled by the treatment of the advances in the company's records as both demand debt and long-term debt. While we share the Tax Court’s concerns, this was not the issue before the Tax Court or us on appeal. Importantly, regardless of whether it classified the payments as demand debt or long-term debt, the company at all times *781identified the advances as some form of debt. We leave it to the Tennessee authorities to determine whether Indmar owes any state taxes or penalties as a result of its reporting and accounting practices.
. As the Tax Court did not address this factor, it made no credibility determinations with respect to Richard Rowe's testimony relating to the use of the advancements. At one point in its decision the Tax Court did find that Rowe's testimony was "contradictory, inconsistent, and unconvincing" and that the parties "manipulated facts,” but this was in specific reference to its discussion about the inconsistent treatment of the advances as demand debt and long-term debt. See Ind-mar, 2005 T.C.M. LEXIS 31, at *12-13; see also supra note 4.
. Judge Moore in her dissent suggests that we give '‘minimal deference” to the conclusions of the Tax Court. We respectfully disagree. Much of our analysis is expressly predicated on the Tax Court’s findings of fact. See, e.g., supra §§ II.C.l & 2 (relying on the Tax Court's findings of a fixed 10% interest rate and regular interest payments, and documentation of advances from 1993-2000); 3 & 6 (accepting the Tax Court's findings of no fixed maturity date or schedule of payments, and no sinking fund, but concluding that these factors do not deserve significant weight under the circumstances); and 7 (accepting in full the Tax Court's findings on five of the Roth Steel factors). On the remaining two Roth Steel factors, we point to clear, uncontroverted evidence in the record favoring debt, evidence that the Tax Court unfortunately did not discuss in its Roth Steel analysis. On the issue of witness credibility, it is apparent from our analysis that we take issue not with any credibility determinations the Tax Court may have made as to Rowe's testimony on certain topics, but with the its failure to address evidence.