ALICE M. BATCHELDER, Circuit Judge.

In this case, we must decide whether the district court was correct in affirming the arbitrator's decision that Plaintiff-Appellant Sherwin-Williams's sale of a wholly-owned subsidiary subjected it to withdrawal liability under section 4212(c) of the Employee Retirement Income Security Act of 1974 ("ERISA"), 29 U.S.C. § 1001 et. seq., as amended by the Multiemployer Pension Plan Amendments Act of 1980 ("MPPAA"), 29 U.S.C. § 1392(c). For the reasons that follow, we affirm the judgment of the district court.

I.

In 1987, Sherwin-Williams, a national manufacturer and distributor of paint and paint-related products, paid $14,204,765.79 to purchase 100% of the stock in Lyons Transportation Lines, Inc., a less than truckload ("LTL") carrier. Apparently, Sherwin-Williams was having trouble competing with local and regional paint manufacturers, who enjoyed the advantage of conveniently located service facilities. Sherwin-Williams planned to use Lyons to cut its own transportation and distribution costs by developing a sort of "rolling warehouse" where inventory could be maintained in trucks so as to be immediately available for delivery to local customers.

The plan was less than successful, however, partly because Lyons incurred massive losses in all but a few months of its operation and partly because the trucking company never adequately met Sherwin-Williams's expectations in terms of developing a rolling warehouse. During each year of ownership, Sherwin-Williams had to subsidize Lyons in order to keep the trucking concern afloat. By 1990, these subsidies were as much as $500,000 each month. Sherwin-Williams recognized that it needed to stop the bleeding, either by turning Lyons into a profitable concern or by getting rid of it.

During 1988 and 1989, Sherwin-Williams received several unsolicited offers to purchase Lyons's stock. At meetings during this period, Sherwin-Williams's executives discussed the possible sale of Lyons to one of the unsolicited bidders, but those who were concerned over the problem of withdrawal liability repeatedly counseled against liquidating or just selling the company. For example, Thomas Stout testified:

And again, it was brought up several times through the year of 1988, the end of the year of `88 . . . and I know it was my understanding at the time, and it was, I'm sure, Sam Hanania's understanding, that our withdrawal liability at that time was 13 or 14 million dollars . . . And we said if you liquidate the company or you sell the company, and that would be initiated, the withdrawal liability is larger than the company, and there would be no benefit from doing it, and those are the kinds of things we talked about. . . . And Sam said, you've got to worry about unfunded liability, it's there, it will raise its head. . . . and Bob [President and General Manager of Sherwin-Williams Transportation Services Division, Robert Kinney] would make the statement, if it happened once it won't happen again. Sherwin-Williams is a large corporation, and we have the wherewithal to fight any litigation that we come up against . . . .

Moreover, Robert Kinney testified that he was aware during 1988 and 1989 that the withdrawal liability associated with Lyons "could be in the neighborhood of 5 to 16 million dollars." In fact, Kinney spearheaded the original move at Sherwin-Williams to acquire a trucking concern in 1986. In a report he prepared specifically for the acquisition of a common motor carrier, Kinney stated that the issue of withdrawal liability was "one of the most significant financial considerations facing the trucking industry today," but also predicted that "the odds for repeal [of the ERISA provisions on withdrawal liability] are better than even." Kinney's report also listed six main strengths and four main weaknesses of acquiring a trucking concern; two of the four weaknesses listed were "union" and "E.R.I.S.A. obligation."

In May 1989, Kinney commissioned a report on possible ways of turning Lyons around to make it profitable. The report predicted that such a turnaround would take at least three years, during which time Sherwin-Williams would be required to continue to subsidize Lyons. In October 1989, the Sherwin-Williams Board of Directors decided against the time and expense required to make Lyons profitable, and, instead, authorized the sale of Lyons.

About seven months later, Kinney received a letter dated January 12, 1990, from J.R.C. Acquisition Corporation ("JRC"), submitting a proposal to acquire Lyons from Sherwin-Williams. JRC had no assets, no financial backing, and no corporate affiliations, but the proposal identified it as a company formed by Jonathan M. Tendler and Robert M. Castello, the principal officers of Common Brothers, Inc., one of the largest distributors of pharmaceuticals in the tri-state New York metropolitan area, with sales in excess of $200 million. Sherwin-Williams checked with Dunn Bradstreet, but that firm had no information regarding JRC in the state of New York. Furthermore, Sherwin-Williams knew that JRC was not a subsidiary of Common Brothers, the identification in the proposal of Tendler and Castello notwithstanding.

Sherwin-Williams also received proposals from other potential purchasers. R.J.M. Associates offered to buy Lyons for $6 million — an offer that was later amended to $8 million. Baytree Investors, Inc., offered to buy all outstanding shares of Lyons's stock for $8.5 million ($2 million cash and the balance as preferred stock, redeemable in 24 months). Arrow Carrier Corporation, a large northeastern motor carrier headquartered in New Jersey, also made a bid.

Despite these other offers, Sherwin-Williams chose to negotiate with JRC. As part of its due diligence, JRC requested a schedule of current pension liabilities as of December 31, 1989, and Lyons sent letters to the Teamsters Fund requesting information on withdrawal liability. On February 22, 1990, JRC and Sherwin-Williams executed a letter of intent outlining a transaction in which JRC would purchase all issued and outstanding stock of Lyons for $7.85 million, contingent on verification that the unfunded ERISA liability of Lyons did not exceed $7 million. Although it approached more than 40 banks and real estate investors, JRC apparently was unable to secure the financing necessary to execute a straight stock-for-cash deal. Stout testified that investors expressed a number of reasons for their reluctance to finance the deal:

[T]he worst thing was that it was into the trucking industry and they weren't interested in lending any more money into that industry. Number two, it was too highly leveraged . . . and the operational losses incurred from the first of the year to that point in time had been too high.

In spite of its knowledge of JRC's difficulties, Sherwin-Williams did not consider the other offers, but instead chose to continue its negotiations with JRC. On May 17, 1990, the parties executed an agreement which called for a down payment of $1.6 million, with the balance of $6.25 million payable on a promissory note over the next six years. The promissory note was secured by mortgages on Lyons's real property. Moreover, Sherwin-Williams agreed to contribute approximately $18 million to the paid-in capital of Lyons in order to eliminate certain intercompany accounts between Lyons and Sherwin-Williams. Finally, the agreement contained a warranty on the part of JRC that it had not encumbered or agreed to encumber or grant any lien on Lyons's assets to anyone other than Sherwin-Williams.

The sale closed on June 1, 1990. Prior to that time, on or about May 25, 1990, before it had the authority to do so, and, so far as this record demonstrates, without the knowledge of Sherwin-Williams, JRC and its principals had executed in the name of Lyons an accounts receivable factoring and security agreement with Allstate Financial Corporation. JRC used the proceeds from the Allstate security agreement to make the $1.6 million cash payment to Sherwin-Williams on June 1, 1990. In short, JRC had no capital or assets of its own with which to subsidize Lyons's monthly losses and it could not obtain financing to do so on the strength of Lyons's assets, because those assets, such as they were, had already been pledged to muster up enough money for the down payment.

Lyons continued to hemorrhage as badly under JRC's control as it had under Sherwin-Williams's. On August 2, 1990, Sherwin-Williams sent Lyons a 30-day termination notice informing Lyons that Sherwin-Williams would no longer accept any freight from Lyons and that from that point on, any line haul or brokerage services would need to be prepaid by Lyons. Then, on September 12, 1990, Sherwin-Williams sent letters to Lyons demanding payments of $74,099.05 for the sublease of certain trailers and $78,125 for the first quarterly payment on the purchase price of Lyons, for which JRC and Lyons were jointly and severally liable.

On October 19, 1990, Lyons and JRC filed Chapter 11 bankruptcy proceedings. Subsequently, the Teamsters Fund sent a letter to Lyons informing it of the possibility of withdrawal liability. On December 11, 1990, the Teamsters Fund sent a letter to Lyons in which it assessed withdrawal liability in the amount of $1,606,605.

Shortly thereafter, the Teamsters Fund sent a letter to Sherwin-Williams demanding payment by Sherwin-Williams of Lyons's withdrawal liability because Sherwin-Williams's principal purpose in selling Lyons's stock was to evade or avoid withdrawal liability under MPPAA. Sherwin-Williams urged the Teamsters Fund to reconsider and also argued that the liability had been incorrectly calculated. After a volley of repeated requests and denials and at least one meeting between the lawyers for the Teamsters Fund and Sherwin-Williams, the Teamsters Fund finally revised its calculation of withdrawal liability downward to $477, 270. The September 18, 1991 letter informing Sherwin-Williams of the revision, however, also stated that the Teamsters Fund had included so-called "gap year interest" in its new calculation.

Sherwin-Williams paid the assessed amount to the Teamsters Fund, but preserved its objections for arbitration, which it later initiated on July 9, 1991. Sherwin-Williams sought arbitration of two issues: whether a principal purpose of its sale of Lyons's stock was to evade or avoid withdrawal liability and whether the Teamsters Fund was correct to include "gap year interest" in its calculation of liability.

The arbitration hearing, held in October 1993, lasted four days and included the admission into evidence of both testimony and exhibits. On February 7, 1994, the arbitrator issued a 63-page decision finding in favor of the Teamsters Fund on the issue of whether a principal purpose of Sherwin-Williams's sale of Lyons was evasion or avoidance of withdrawal liability, and in favor of Sherwin-Williams on the issue of "gap year interest."

On March 8, 1994, Sherwin-Williams filed an action, pursuant to MPPAA, 29 U.S.C. § 1401(b)(2), in the United States District Court for the Northern District of Ohio, requesting that the court vacate the arbitrator's decision. On February 29, 1996, the magistrate judge issued a report and recommendation in which he recommended that the arbitrator's decision be enforced on both issues. Sherwin-Williams timely filed its objections, but on March 31, 1997, the district court issued a memorandum of opinion and order adopting the findings and recommendations of the magistrate judge.

Sherwin-Williams now appeals the district court's decision.

II.

The MPPAA provides the procedures for both calculating and assessing liability of an employer withdrawing from a multiemployer pension plan. Among other things, MPPAA provides that withdrawal liability is to be assessed without regard to any transaction entered into by the employer if a principal purpose of that transaction was to avoid or evade that liability. 29 U.S.C. § 1392(c).

The MPPAA provides for resolution of disputes over both calculation and assessment of withdrawal liability by means of arbitration, 29 U.S.C. § 1401(a)(1), and for limited judicial review of the arbitrator's decision resolving such a dispute. 29 U.S.C. § 1401(b)(2). Further, the MPPAA defines the statutory presumptions and limited standard of review to be used in the resolution of a dispute over withdrawal liability. Section 1401(a)(3)(A) provides that a determination of withdrawal liability made by a plan sponsor, such as the Teamsters Fund in the present case, "is presumed correct unless the party contesting the determination shows by a preponderance of the evidence that the determination was unreasonable or clearly erroneous." 29 U.S.C. § 1401(a)(3)(A). The Supreme Court has explained that the presumption established by § 1401(a)(3)(A) requires an employer challenging a factual determination made by the plan sponsor to disprove that factual determination by a preponderance of the evidence. Concrete Pipe Prods. of Cal., Inc. v. Construction Laborers Pension Trust for S. Cal., 508 U.S. 602, 629 (1993). The Court defined the preponderance standard as requiring the arbitrator "to believe that the existence of a fact is more probable than its nonexistence before [he] may find in favor of the party who has the burden to persuade the [arbitrator] of the fact's existence." Id. at 622.

The judicial review provided pursuant to § 1401(b)(2) is extremely narrow, because "[i]n any proceeding under subsection (b) . . . there shall be a presumption, rebuttable only by a clear preponderance of the evidence, that the findings of fact made by the arbitrator were correct." 29 U.S.C. § 1401(c) (emphasis added). Moreover, reviewing courts must give great deference to an arbitrator's determination because of the MPPAA's strong policy favoring arbitration of withdrawal liability disputes. Mason Dixon Tank Lines, Inc. v. Central States Pension Fund, 852 F.2d 156, 163-64 (6th Cir. 1988) (stating that arbitration is the "preferred method" for resolving withdrawal liability disputes and that "under the MPPAA `arbitration reigns supreme'").

As we have held, "The series of presumptions prescribed by the Multiemployer Act were intended by Congress to `ensure the enforceability of employer liability. In the absence of these presumptions, employers could effectively nullify their obligation by refusing to pay and forcing the plan sponsor to prove every element involved in making an actuarial determination.'" Board of Trustees, Mich. United Food Commercial Workers Unions v. Eberhard Foods, Inc., 831 F.2d 1258, 1260 (6th Cir. 1987) (quoting H.R. Rep. No. 869, pt. I, 96th Cong., 2d Sess. 1, 86, reprinted in 1980 U.S. Code Cong. Admin. News 2918, 2954). Furthermore, deference to the findings of the arbitrator is proper because the arbitrators chosen to resolve the complicated issue of withdrawal liability often have relevant expertise in the field of pension law which can contribute significantly to the accuracy of a decision. See, e.g., Chicago Truck Drivers Pension Fund v. Louis Zahn Drug Co., 890 F.2d 1405 (7th Cir. 1989).

The arbitrator was required to assess the facts of this case in light of the rebuttable presumption in favor of the Teamsters Fund, and the district court, in turn, was required to review the arbitrator's findings of fact for clear error. We must also review the arbitrator's findings of fact for clear error, which means that we will only disturb the arbitrator's findings if, after reviewing the entire record, we are left with the definite and firm conviction that a mistake has been committed. Anderson v. Bessemer City, 470 U.S. 564, 570 (1985). In addition, the arbitrator's application of the appropriate burden of evidence is a question of law which we review de novo.

A. Findings of Fact

Sherwin-Williams argues first that the arbitrator committed clear error when he determined that a principal purpose of Sherwin-Williams's decision to sell Lyons was to evade or avoid withdrawal liability. This argument is essential to its appeal, since if this finding is erroneous, there is no basis upon which to assess Lyons's withdrawal liability against Sherwin-Williams. According to Sherwin-Williams, the crucial issue, not addressed by the arbitrator or the district court, is not whether or not Lyons was actually economically viable, but whether Sherwin-Williams believed that Lyons was economically viable. Sherwin-Williams is mistaken.

Lyons's economic viability was relevant to the arbitrator's decision, because the arbitrator found that

[w]hile the solvency or general fiscal status of Lyons at June 1, 1990 is highly relevant, it is not determinative. The issue presented for resolution is whether Sherwin-Williams sold Lyons to JRC with a principal purpose to evade or avoid the withdrawal liability under the Fund if Lyons were to make a complete or partial withdrawal from the Fund. Solvency obviously bears on the issue. . . . The less `healthy and viable the corporate entity' at the sale date, the more likely, on a circumstantial basis, that ERISA Section 4212(c) motivations were involved.

Sherwin-Williams argues, however, that if it believed Lyons was economically viable, evading or avoiding withdrawal liability could not have been a principal purpose of the transaction because in that event, it would have had no reason to expect Lyons and JRC to withdraw from the Fund.

The arbitrator's analysis makes it clear that he looked to the economically doomed status of Lyons as circumstantial evidence that Sherwin-Williams must have known that Lyons was going to fail the minute Sherwin-Williams cut it loose. It is true that some of Sherwin-Williams's employees testified that they fully expected JRC to be able to turn Lyons into a profitable concern, but it is equally clear that the arbitrator did not find such testimony credible in light of the facts.

Whether Lyons was insolvent on June 1, 1990 or merely on its financial deathbed is irrelevant. No one can deny it was fiscally moribund. Sherwin-Williams knew that without a deep pocket to subvend its monthly losses, Lyons could not survive.

In other words, Sherwin-Williams must have known that Lyons was going to fail. JRC did not even have enough investors to finance a straight cash-for-stock deal, much less to subsidize Lyons to the tune of $500,000 per month for the three years that Sherwin-Williams predicted it would take to return the company to profitability. Indeed, JRC was forced to pledge Lyons's accounts receivable just to obtain the money needed for the down payment. From where, exactly, did Sherwin-Williams expect JRC and Lyons to obtain the remaining $6.25 million?

The Fund supplied an answer, which the arbitrator believed, and which Sherwin-Williams failed to rebut. Namely, Sherwin-Williams never expected to receive the remaining $6.25 million of the purchase price. Sherwin-Williams originally paid over $14 million to purchase Lyons, but four years later it was accepting an offer of $7.85 million, only $1.6 million of which it could ever realistically expect to see. Such a transaction makes little sense unless Sherwin-Williams sought to avoid, by means of this transaction, its potential employer withdrawal liability, which it had at various times estimated to be somewhere between $5 million and $16 million.

Thus, Sherwin-Williams's argument that the arbitrator failed to ask whether Sherwin-Williams believed Lyons was economically viable is without merit. The arbitrator looked at the evidence, concluded that Sherwin-Williams must have known that Lyons was going under, and consequently viewed Sherwin-Williams's protestations to the contrary as lacking credibility. The Supreme Court has noted that "[w]hen findings are based on determinations regarding the credibility of witnesses, Rule 52(a) demands even greater deference to the trial court's findings." Anderson, 470 U.S. at 575. Given the cornucopia of evidence demonstrating Lyons's failing health, we cannot say that the arbitrator was wrong to doubt Sherwin-Williams's claims of ignorance as to the financial doom awaiting Lyons.

Next, Sherwin-Williams argues that the arbitrator's findings of facts were erroneous because they were based on flawed logic. Although Sherwin-Williams takes a scattershot approach to introducing the various flaws which it says undermined the validity of the arbitrator's factual findings, we can organize the arguments into two distinct claims: (1) Sherwin-Williams's choice of how to structure the sale of Lyons does not necessarily support the logical inference that a principal purpose of the sale was to evade or avoid withdrawal liability; and (2) the arbitrator should have believed the testimony of the various Sherwin-Williams employees who stated that they were motivated by legitimate considerations other than the evasion or avoidance of withdrawal liability.

According to Sherwin-Williams, the arbitrator wrongly concluded that Sherwin-Williams disposed of Lyons by selling stock because any other alternative would have resulted in withdrawal liability. This, says Sherwin-Williams, is not true. An asset sale structured to comply with 29 U.S.C. § 1384 would not have subjected it to withdrawal liability. We observe, however, that Sherwin-Williams presented no evidence that such an asset sale was a viable alternative. Indeed, Sherwin-Williams went to great pains to explain that the LTL trucking market in the late 1980s was glutted and that Sherwin-Williams would not have been able to obtain a fair price for Lyons's assets on such a market. According to Kinney, "There were 900 plus motor carriers out of business, terminals everywhere, equipment for sale, auctions daily. I think to have even considered an asset sale without the customer base would have been a waste of time." The burden was on Sherwin-Williams to prove that avoidance was not a principal purpose of the transaction. Pointing to the existence of a statute, without proving its applicability to these particular circumstances, is not enough to carry that burden.

Next, Sherwin-Williams argues that the sale of Lyons was authorized on October 20, 1989, before Sherwin-Williams was aware of the withdrawal liability. The arbitrator conceded as much when he stated that there was "no evidence sufficient to support an affirmative intent to violate ERISA § 4212(c) at this point in time." However, the arbitrator never found that avoiding withdrawal liability was the only principal purpose of selling Lyons. No one disputes that Lyons represented a massive drain on Sherwin-Williams's cash-flow and that Sherwin-Williams wanted to stop the bleeding by getting rid of Lyons.

This does not alter the fact, however, that Sherwin-Williams was presented with a number of options for disposing of Lyons, and that it chose the only one that would not subject it to withdrawal liability. It is undisputed that Sherwin-Williams chose the method of disposing of Lyons after it became aware of the likely dimensions of the potential withdrawal liability. Thus, the decision of the board of directors on October 20, 1989 does nothing to undermine the logical validity of the arbitrator's factual findings.

In addition to attacking the logical inferences drawn from the evidence by the arbitrator, Sherwin-Williams also urges that the testimony proves that Sherwin-Williams had other legitimate motives, and that the existence of these motives contradicts the arbitrator's finding that evading or avoiding withdrawal liability was a principal purpose of the transaction. Again, Sherwin-Williams's argument succeeds only if we read the MPPAA to require that avoidance or evasion of withdrawal liability be "the principal purpose" rather than "a principal purpose" of the sale.

As the arbitrator correctly pointed out, the language of the MPPAA makes it clear that an employer can have more than one principal purpose in conducting a transaction. This is especially true where, as here, one principal purpose can be said to motivate the decision about whether to sell the company at all, while another principal purpose can be said to motivate the decision about how to sell the company. In the present case, the arbitrator found that Sherwin-Williams's "number one, first reason" for deciding to sell Lyons may have been to eliminate the negative cash flow, but that a principal purpose in selling Lyons to a shell corporation with no assets or corporate affiliations in a highly-leveraged cash-for-stock swap was to evade or avoid withdrawal liability.

After reviewing the exhibits and testimony, as well as the arbitrator's thorough and well-reasoned analysis, we are not left with the firm conviction that a mistake has been made. Without such a firm conviction, we cannot hold the arbitrator's decision to be clearly erroneous.

B. Due Process

Next, Sherwin-Williams argues that the arbitrator must have been impermissibly biased against Sherwin-Williams, as evidenced by the final statements from the arbitrator's decision:

[Withdrawal liability] may be considered — and if the truth be known, is always considered (or, at least, should always be known and considered) — by competent corporate executives and benefits counsel. . . . [O]nce you realize that Sherwin-Williams is a Fortune 200 company, with knowledgeable experts and personnel and a $7 million issue, you know it was a factor and, in the overall context, "a principal purpose" to be rid of the Title IV liability.

Sherwin-Williams argues that this statement proves that the arbitrator was predisposed to rule against Sherwin-Williams and to reject the testimony of Sherwin-Williams's witnesses.

This "statement" is pieced together out of the final four pages of the arbitrator's 63-page decision and opinion, and included in those final four pages is the arbitrator's strong statement of his finding that Sherwin-Williams's first purpose in selling Lyons was to stop the huge monthly losses. Even considering this patchwork quotation in isolation, we find in it no indication of bias against Sherwin-Williams. The arbitrator observed that it would be impossible for a company such as Sherwin-Williams not to take withdrawal liability into consideration. In fact, the record is replete with evidence that Sherwin-Williams's employees discussed the issue frequently while deciding what to do about Lyons. The arbitrator opined that "in the overall context" the evasion or avoidance of withdrawal liability was a principal factor. The arbitrator's task is to make precisely that kind of determination; i.e., to determine what, in the overall context, were the employer's principal purposes for undertaking the transaction at issue. The arbitrator's statement is not evidence of bias, but rather the articulation of a conclusion with which Sherwin-Williams disagrees.

Sherwin-Williams's remaining attempts to impugn the arbitrator's objectivity are meritless. In his 63-page opinion, the arbitrator thoroughly analyzed the arguments, evidence and testimony offered by Sherwin-Williams. The arbitrator did eventually rule against Sherwin-Williams, but he was not biased in doing so. If anything, the arbitrator gave Sherwin-Williams an advantage by failing to give the Teamsters Fund the benefit of the required statutory presumption.

C. Statutory Presumptions

Finally, Sherwin-Williams argues that the arbitrator failed to use the appropriate standard in analyzing this case. We agree. As we explained above, the MPPAA requires an arbitrator to presume the correctness of a plan sponsor's determinations with regard to withdrawal liability, including the determination of whether a principal purpose for the withdrawal was to evade or avoid liability, made under § 1392(c). The employer bears the burden of rebutting this presumption and must prove, by a preponderance of the evidence, that the plan sponsor's determination was incorrect.

Thus, according to the MPPAA as interpreted by the Supreme Court in Concrete Pipe, 508 U.S. at 629, the arbitrator in this case should have started with the presumption that the Teamsters Fund correctly determined that a principal purpose behind Sherwin-Williams's sale of Lyons to JRC was to evade or avoid liability under the MPPAA. The arbitrator should have examined all of the evidence to determine whether a preponderance of that evidence rebutted the presumption. This is not what the arbitrator did.

Instead, the arbitrator held,

From a careful review of the evidence, it is clear beyond any doubt that each party has introduced a significant volume of credible testimonial and documentary evidence to support its conclusion. While agreeing with the holdings in Concrete Pipe and the arguments derived therefrom by the Fund and Sherwin-Williams, this dispute can and will be resolved by the evidence presented at the hearing and without reliance on the statutory presumptions. To couch this conclusion in the language of the precedential opinions, the documentary and testimonial evidence introduced by [Sherwin-Williams] was sufficient to disprove the Fund's factual determination by [sic] preponderance, i.e., the existence of a fact is more probable than its nonexistence, if accepted. The same, however, is true of the evidence actually introduced by the Fund.

Although we agree with Sherwin-Williams that the arbitrator failed to apply the statutory presumption in favor of the Teamsters Fund, we fail to see how this in any way harmed Sherwin-Williams. The arbitrator examined the evidence and made his legal determinations without giving the Teamsters Fund the benefit of a rebuttable presumption in its favor. The standard the arbitrator actually used was more favorable to Sherwin-Williams than the law allows, and the arbitrator still found against Sherwin-Williams. Sherwin-Williams cannot expect to find any more success under a less favorable standard.

The district court found that the arbitrator's findings of fact are not clearly erroneous. We have reviewed the record in this case and conclude that the district court is correct. Applying the statutory presumption of correctness in favor of the arbitrator's findings, as we must do in our de novo review of the arbitrator's legal conclusions, we hold that Sherwin-Williams failed to rebut that presumption by a preponderance of the evidence. For this reason, we conclude that the arbitrator's mistake of law with respect to the application of the presumption in favor of the Teamsters Fund was harmless error.

III. Conclusion

For the reasons discussed above, we AFFIRM the decision of the district court affirming in its entirety the decision and opinion of the arbitrator.