On November 16, 2010, the Federal Court of Canada (“FCA”) heard the Crown’s appeal from a decision by Hogan J. of the Tax Court of Canada (“TCC”). At the trial level, the TCC found in favor of General Electric Capital Canada Inc. (“GEC”) regarding the payment of guarantee fees to its parent, General Electric Capital Corporation (“GEUS”), for guaranteeing GEC’s debt to third parties. GEC paid guarantee fees of over $135 million during the 1996 through 2000 taxation years. The TCC ruling was one of Canada’s most significant cases to date regarding guarantee fees, as well as transfer pricing, and the arm’s length principle. On December 15, 2010 the FCA delivered its judgment, and dismissed the Crown’s appeal with costs.
At issue in this case was the arm's length nature of the payment by GEC to its parent. The Canada Revenue Agency (“CRA”) audited GEC's guarantee fee, and determined that an arm's length person would not have paid for the guarantee as it provided no benefit to GEC. GEC disagreed, asserting that the guarantee reduced its borrowing costs.
In the initial ruling, the Tax Court Judge determined the guarantee fees paid by GEC did not exceed the amount which a person dealing at arm’s length would have paid in like circumstances, thus he vacated the assessments. On appeal, the Crown contended that the TCC Judge committed numerous legal and factual errors.
The Crown’s two central arguments to allow the appeal were premised on 1) that an arm’s length party would not have paid the guarantee fee because no value was provided, and 2) that the behaviour of the Tax Court Judge at trial gave rise to a reasonable apprehension of bias.
After considering these central arguments, the FCA Judge dismissed the Crown’s appeal, determining that the TCC ruling was justified and appropriate.
This case involved an American parent company, GEUS, and its wholly-owned, indirect Canadian subsidiary, GEC. GEC and GEUS did not deal with each other at arm’s length during the years under appeal. During this time, GEC was a financial services company operating in Canada. GEC’s operations essentially consisted of borrowing funds from the capital markets at low cost, and then turning these funds into profits by lending or leasing to other parties. In order to implement this business model, substantial amounts of capital were obtained by GEC by issuing debt in the form of commercial paper and unsecured debentures. The exclusive purchasers of GEC’s debt securities were third parties unrelated to GEC or GEUS.
GEUS guaranteed GEC's debt prior to 1995, however, GEUS only started charging GEC for the guarantees in 1995. Written agreements were entered into by GEC and GEUS concerning the guarantee fee (the “Guarantee”). GEUS agreed to guarantee GEC’s debt securities in return for the payment of a fee equal to 1%, or 100 basis points, per annum of the principal amount of the debt securities outstanding from time to time during a year. The fee was deducted by GEC for the 1996 to 2000 taxation years.
The CRA reassessed GEC because it believed that GEC had obtained no economic benefit from the Guarantee. The CRA was of the opinion that the arm’s length price of the Guarantee should be zero. The TCC was tasked with determining the arm’s length price for the Guarantee.
Standard & Poor’s and Moody’s Investors Service, two credit rating agencies based in the United States, assigned an issuer rating of AAA to GEUS, the highest issuer rating assigned by them. GEC also received the highest credit quality ratings by two Canadian based credit agencies. The consensus between GEC’s representatives and experts is that the AAA investment rating for GEC’s debts would not have been possible without the Guarantee.
At trial several experts explained their findings based on different methodologies. The TCC concluded that the yield curve approach provided the best method to determine the Guarantee fee. This approach is a reflection of the expenses incurred when borrowing money given different maturities and credit ratings. It compares the interest rates that GEC could obtain when borrowing money with the Guarantee versus without the Guarantee.
According to GEC, the yield approach determined that the spread was between 100 and 300 basis points, or 1 to 3%. In general, experts for GEC analyzed the spread between AAA- rated bonds and bonds that are an average of single B and BB. One expert concluded that the overall spread was about 352 basis points between a AAA rating and the B+ to BB- rating for GEC in the absence of an explicit guarantee. Therefore, the value of the Guarantee was determined to be approximately 1.83% based on a BB+ to BBB- credit rating range.
GEC’s experts assigned a B+ to BBB- credit rating to GEC as a stand-alone entity for various reasons. GEC was a profitable entity growing rapidly in a very stable marketplace. However, GEC was thinly capitalized and had a high degree of leverage. GEC’s profitability was also decreasing during the period in question. Despite GEC’s reduced leverage and rapid growth, it was unable to generate profits or increase profits on a continuous basis. Further, GEC was part of an intensely competitive environment. Experts claimed that, in general, when a weak entity is owned by a strong parent, the entity will often receive a stronger rating than it would on a stand-alone basis. Some experts were of the opinion that GEC was only able to borrow the amount of funds it did in the Canadian commercial paper market because of the Guarantee from GEUS.
One of CRA’s experts used a quantitative approach to measure the creditworthiness of GEC. This approach is generally based on financial market data, such as stock prices, bond prices and CDS spread. He determined that GEC was a core subsidiary of GEUS and should thus receive an AAA rating. At the very least, GEC could have been rated AA if classified as having been strategically important to GEUS at the relevant time, rather than of core importance. As a result, in the opinion of the CRA's expert, the fee would have been between 15 and 24 basis points. He therefore classified the 1% fee as a very high-risk adjusted return on capital.
FCA Decision & Analysis
The arguments raised by the Crown in support of its appeal allege that the Tax Court Judge committed four errors of law. The Crown also requested that the FCA allow the appeal on the basis that an arm’s length party would not have paid the Guarantee fees. The four errors of law alleged by the Crown are as follows:
- The Tax Court Judge erred by removing the explicit guarantee for purposes of identifying the relevant transaction.
- The Tax Court Judge erred in law by failing to consider four relevant characteristics in assessing the value of the explicit guarantee.
- The Tax Court Judge erred by not applying the “reasonableness” check.
- The Tax Court Judge erred by upholding the necessity of the explicit guarantee based on the business judgment of a witness.
In addition, the Crown also alleged that the behaviour of the TCC Judge gave rise to a reasonable apprehension of bias.
The FCA Judge began his analysis by addressing the issue of implicit support. In its decision, the Tax Court Judge reduced the arm’s length price for the Guarantee on account of implicit support. GEC argued that, since transfer pricing principles require that a transaction be examined in a fictitious arm’s length context, implicit support cannot be considered because the concept is rooted in the familial relationship between affiliated companies and is thus a by-product of the non arm’s length relationship. The Crown relied on the concept of implicit support for its proposition that the explicit guarantee was of no value to GEC.
In the recent FCA decision in Glaxosmithkline Inc. v. Canada, 2010 FCA 201 [Glaxo], Glaxo successfully argued that all relevant circumstances must be taken into account, the relevant circumstances being those which an arm’s length purchaser, standing in the shoes of Glaxo, would consider relevant in deciding whether it should pay the price paid by Glaxo for the goods. In light of the Glaxo decision, the FCA Judge in the GE case held at paragraph 59 that in “[a]pplying this test, there is no doubt that the existence of the implicit guarantee is relevant to the inquiry and must be considered in identifying the arm’s length price.” The FCA Judge further stated that in the context of the yield method, implicit support is a factor which an arm’s length person would find relevant in pricing the guarantee.
The FCA Judge then analyzed each of the Crown’s four error of law assertions, and the reasonable apprehension of bias argument. With respect to the first error of law argument, that being that the TCC Judge erred by removing the explicit guarantee, the FCA Judge concluded that this error had no impact on the Tax Court Judge’s finding that a gap existed between the credit rating which the respondent would have obtained with and without the explicit agreement. Moreover, the 1% guarantee fee was within this gap.
The second argument put forth by the Crown was that the TCC Judge failed to consider four relevant characteristics in assessing the value of the explicit guarantee. The FCA Judge determined that the TCC Judge had in fact highlighted the importance of at least three of the four factors and that the Crown’s real complaint was that the TCC Judge should have preferred evidence of one witness over that of another. Having come to this conclusion, the FCA Judge determined that this goes to the weighting of evidence, and that the evidence justifies the decision made by the TCC Judge.
With respect to the third error of law argument, that being that the TCC Judge erred by not applying the “reasonableness” check, the FCA Judge agreed with the TCC Judge in that a “reasonableness” check was not required. To this end, the FCA Judge concluded the “reasonableness” test proposed by the Crown, which would have demonstrated that a 2% guarantee fee would be unreasonable, was not relevant because the fee actually charged and claimed as a deduction was only 1%.
The fourth error of law argument was that the TCC Judge relied on the subjective business judgment of a witness that claimed the explicit guarantee was necessary. The core of the Crown’s argument was that the necessity of the explicit guarantee had to be assessed with reference to objective rather than subjective evidence. The FCA Judge determined that the TCC Judge had considered the witness’ business judgment only after having found, based on objective evidence, that the explicit guarantee was necessary.
The reasonable apprehension of bias argument rests upon the Crown’s allegation that the TCC Judge developed his own theory of the case. The FCA Judge concluded there was no established reasonable apprehension of bias by the TCC Judge.
Having tackled each error of law argument, and further determining that there was no reasonable apprehension of bias, the FCA Judge dismissed the Crown’s appeal with costs.
Secondary Issues Involving Guarantee Fees
In light of the FCA decision, and the public attention it and its predecessor (TCC) decision received, it will be safe to assume that tax advisors, when developing their transfer pricing strategies for their clients, will be more inclined to pursue the possibility of arm's length guarantee fees where the facts and circumstances support it. Such strategies may, of course, ultimately depend on whether there is a desire to report profits in the guarantor's state of residence. Regardless, as there now could be a surge of activity on this front, one should also be aware of certain recent secondary issues pertaining to guarantee fees.
Paragraph 4 of Article XXII of the Canada-United States Tax Convention (1980) (“Treaty”) was recently added, as a consequence of the Fifth Protocol, to ensure that compensation derived by a resident of a Contracting State in respect of the provisions of a guarantee of indebtedness shall be taxable only in that State. The only exception for this rule will be where the compensation is considered business profits attributable to a permanent establishment situated in the other contracting state. The Technical Explanation to paragraph 4 of article XXII also clarifies that guarantee fees with respect to the debt of related parties is ordinarily not an independent economic undertaking that would generate business profits. In brief, new Article XXII(4) will apply to ensure that there is no withholding taxes in the source state on most guarantee payments.
Prior to the Fifth Protocol, both Canada and the U.S. had different approaches to the taxation of guarantee fees. For example, in many situations paragraph 214(15)(b) of the Income Tax Act (“Act”) would apply on southbound guarantee fees to deem the payment of a guarantee fee by a Canadian resident to a U.S. resident to be “interest” for purposes of Part XIII of the Act. The deemed interest would then be subject to the reduced withholding rates under Article XI of the Treaty. For northbound guarantee fees, the U.S. did not deem such payments to be “interest”, but instead, imposed a 30-percent tax on “fixed or determinable annual or periodical” income received from sources within the U.S. by a foreign corporation, but only to the extent the amount so received was not effectively connected with the conduct of a trade or business within the U.S. Where the conditions for the 30% withholding tax were met, depending on the circumstances, the U.S. viewed the northbound guarantee fees as being subject to Article XXII. The result was that the U.S. maintained full source state taxation rights. Now there is some degree of symmetry between the two countries’ tax treatment of outbound guarantee fees, and taxpayers can be assured of no withholding tax implications.
Finally, in light of arbitration, which was also introduced in the Fifth Protocol, there is now complete protection for taxpayers against exposure to double taxation in the event the guarantee fee is subsequently challenged by the CRA or IRS. Between arbitration, and the fact that there is no longer withholding tax implications in the source state, this will further encourage tax advisors to give due consideration to transfer pricing strategies involving guarantee fees.
The FCA upheld the TCC’s ruling by agreeing to uphold the arm's length principle and the concept of implicit support, and set the price of the transaction (the Guarantee) between GEC and GEUS at a price that two independent parties would have agreed to. By reaching this conclusion, the FCA has instilled and bolstered the fundamental principles of transfer pricing. At the very heart of transfer pricing is the notion that intercompany prices between related parties be set as if the parties were independent of one another. This principle is enshrined in the OECD guidelines on transfer pricing, which the CRA generally adheres to. The CRA adjustment that led GEC to court is troubling on a number of levels, especially given the fact that certain transfer pricing principles, namely the idea that prices be set to represent arm’s length standards, were ignored by the CRA.
The CRA's position is unfortunately another example of where the Agency did not follow its own principles or past decisions, and decided to take a tough stance based on the pure quantum of the adjustment. The CRA must be more consistent in the manner in which they approach transfer pricing, particularly in respect to guarantee fees, considering it has accepted guarantee fees in the past on both an inbound and outbound basis. It is important to note that the size of the fee should not matter, as long as such fees meet the commercial and economic realities of the given transaction. In the final analysis, any fee negotiated should be based on sound economic theory, and principles that meet the economic reality of the transaction. The FCA’s ruling confirms these sound transfer pricing principles.
The Transfer Pricing and Competent Authority Group would also like to thank Student-at-Law Jonathan Greenwald for his assistance on this article.
 The IRS recently lost a court decision on this issue where it was determined by the court that the guarantee fee paid by a U.S. resident was not a U.S. source - see Container Corp. v. Comm'r, 134 T.C. No. 5 (2010) (United States Tax Court).