Intercompany financing after the demise of LIBOR

This is a guest post by Harold McClure, a New York City-based independent economist with 26 years of transfer pricing and valuation experience. After reading our recent blog on the Singapore Telecom Australia decision, he was inspired to share his expertise in an in-depth and detailed analysis of 'life after LIBOR' for multinational groups. 

Paul Sutton’s discussion of the intercompany financing arrangement in the Singapore Telecom Australia decision opened with this timely question:

As of 1 January this year, 24 of the 35 LIBOR interest rate benchmarks or ‘settings’ are no longer published. What does this mean for groups which still have intercompany loan agreements, loan notes or cash pooling arrangements that refer to LIBOR? On the face of it, it’s very simple: select an alternative risk-free rate and update the legal documents accordingly.

The London Interbank Offered Rate (LIBOR) is an interbank interest rate that major global banks use as a benchmark for borrowing with each other. LIBOR is often used for setting interest rates for adjustable-rate mortgages and other floating rate loans. LIBOR rates were once quoted for ten currencies but the leading currencies were the U.S. dollar, the British pound, the Japanese yen, the Swiss franc, and the euro. The reporting of LIBOR rates for the Australian dollar, the Canadian dollar, the New Zealand dollar, the Danish krona, and the Swedish krona terminated in August 2013. LIBOR rates existed for various short-term maturities including overnight, weekly, one-month, 3-month, 6-month, and 12-month loans.

The OECD transfer pricing guidance for financial transactions makes a crucial distinction between fixed interest rate loans versus floating rate loans. The demise of LIBOR has no effect on longer term fixed interest rate loans. The OECD guidance does not mention LIBOR but does recognize the use of interbank loans, which it refers to as a risk-free rate. While domestic interbank rates exist that can be used as the alternative to LIBOR, the idea that interbank rates represent risk-free rates requires clarification.

Christian Zimmermann’s July 7, 2016 FRED blog post on the ‘TED spread’ noted:

There are many TEDs, but the TED in FRED is a spread. That is, the spread between the 3-month LIBOR and the 3-month Treasury bill. A little background: LIBOR is the rate banks would charge each other for lending, which can be used to measure economy-wide credit risk. Treasuries are basically the safest assets on the market. So, a large TED spread would indicate a lot of credit risk in the US economy.

His post noted that the TED spread was around 0.3 percent to 0.4 period during the time of his writing, but this spread shows significant variability during periods of financial crisis.

Consider an intercompany loan to a U.S. affiliate on July 1, 2016 where the interest rate is set at the 3-month LIBOR rate plus a 2 percent margin. The 3-month LIBOR rate on this date was 0.65 percent so the interest rate would initially be 2.65 percent. The interest rate on 3-month Treasury bills on this date was only 0.27 percent reflecting a TED spread = 0.38 percent. While the loan margin may have been 2 percent, the appropriate measure of the credit spread implied by this intercompany loan policy is the sum of the loan margin and the TED spread or 2.38 percent.

Two Australian litigations

Several intercompany loans are expressed in terms of a LIBOR rate plus a loan margin. The intercompany issue in Chevron Australia Holdings Pty Ltd. v. Commissioner of Taxation involved a large intercompany loan denominated in Australian dollars issued on June 6, 2003 where the interest rate was based on the one-month LIBOR rate plus a 4.14 percent loan margin. The Australian Tax Office (ATO) not only challenged this high loan margin but also tried to change the currency of denomination of the loan to U.S. dollars. On this date, the one-month LIBOR rate based in Australian dollars was 4.71 percent, while the one-month LIBOR rate based in U.S. dollars was only 1.27 percent. The OECD’s Main Economic Indicators report that during June 2003, the average 3-month Australian interbank rate was 4.67 percent.

The Singtel Australia litigation, however, involved an intercompany loan based on a domestic interbank rate plus a loan margin that was originally only 1 percent. The initial loan was initiated on June 29, 2002. The OECD’s Main Economic Indicators report that during June 2002, the average 3-month Australian interbank rate was 5.07 percent. During this month, the 3-month LIBOR rate based on Australian dollars average 5.02 percent. The difference between LIBOR rates and domestic interbank rates in Australia were generally very modest through August 2013 when the reporting of this LIBOR rate terminated.

The issues in the Singtel Australia litigation had to do more with the various amendments to the original intercompany agreement. The second amendment significantly increased the loan margin. The third amendment changed the nature of the intercompany loan from a floating rate loan to a fixed rate loan with a high interest rate.

A Norwegian litigation and a challenge from the Swedish National Tax Authority

Ford purchased Volvo in early 1999 for approximately $6.5 million establishing Ford VHC as its Swedish affiliate. Ford extended an intercompany loan for over 75 percent of the acquisition price with the loan denominated in Swedish krona. The initial intercompany loan between Ford and Ford VHC was set at the Stockholm Interbank Offered Rate plus 1 percent. This interbank rate is a domestic interbank rate. The reporting of LIBOR based on Swedish krona did not commence until January 2006.

While this loan was initially a floating interest rate loan with a maturity of 30 years, the loan was refinanced in 2000 with a fixed interest rate that exceeded the initial interest rate. During 2000, interest rates on long-term Swedish bonds averaged 5.37 percent, while the average interest rate on Swedish three-month Treasury bills was only 3.95 percent. In other words, the term structure was such that interest rates on long-term debt were approximately 1.4 percent higher than interest rates on short-term debt. During 2000, the Stockholm Interbank Offered Rate average 4.06 percent implying a credit spread that was approximately 0.1 percent. In other words, the credit spread implied by the original loan was 1.1 percent. Since the Swedish National Tax Authority did not object to the intercompany interest rate policy of the original loan, one could reasonably assume that this tax authority was willing to accept the proposition that the credit worthiness of Ford VHC was such that a 1.1 percent credit spread was reasonable.

The Swedish National Tax Authority argued that the terms of the second loan were unfavorable and an independent borrower would not have chosen to break off the first loan and enter into the second one. The government's contention that no borrower would have chosen to renegotiate a floating rate loan is not necessarily true. Borrowers may prefer the certainty of fixed interest rates over floating rates even if the current fixed interest rate is somewhat higher than current short-term rates especially if interest rates are expected to increase over time. Another way to evaluate this issue would be to add the interest rate on government bonds of the same maturity as the new intercompany loan to a reasonable credit spread. If the appropriate credit spread were indeed 1.1 percent and the interest rate on the corresponding government bonds were 5.4 percent, then this evaluation would suggest that an intercompany interest rate equal to 6.5 percent was consistent with arm’s length pricing even if the prior loan had set the floating rate loan closer to 5 percent. The fact that the Swedish borrower chose to lock their rate to a fixed rate equal to 6.5 percent as opposed to being subject to a variable rate loan is consistent with U.S. homeowners preferring to borrow with a conventional mortgage with a 3.5 percent fixed rate over 30 years instead of taking out a variable rate mortgage with the current rate being only 2.5 percent. This argument considers the strong possibility that short-term rates will rise in the future, which stands in contrast to the ATO argument about Australian interest rate in the spring of 2009 in its case involving Singtel Australia switching from a floating rate policy to a fixed interest rate policy.

Norway vs. Exxonmobil Production Norway Inc. involved a 20 billion Norwegian Krona 10-year intercompany loan from Exxonmobil Production Norway Inc. to Exxon Mobile Delaware Holdings Inc. (EMDHI) on November 16, 2009. The interest rate was set at 3-months Nibor plus a margin of 30 basis points.

Nibor is the Norwegian Interbank Offered Rate, which is a collective term for Norwegian money market rates at different maturities. Nibor is intended to reflect the interest rate level a bank requires for unsecured money market lending in Norwegian krone to another bank. LIBOR rates have never been reported for the Norwegian krone. On November 16, 2009, the 3-month Nibor rate was 2.06 percent, while the 3-month U.S LIBOR rate was only 0.27 percent. The difference between these two interbank rates reflected the fact the interest rate on 3-month Treasury bills in Norway were approximately 1.8 percent higher than the interest rate on 3-month Treasury bills in the U.S. Both of the interbank rates exceeded the corresponding Treasury bill rates by approximately 0.2 percent.

The Norwegian tax authorities successfully argued for a higher loan margin. Both the taxpayer and the tax authority agreed that EMDHI had an AA+ credit rating with the sole issue being how to translate this credit rating into a credit spread. The Norwegian version of the TED spread during this period was approximately 0.2 percent so the 0.3 percent loan margin in the intercompany contract represented a credit spread equal to 0.5 percent.

The Norwegian tax authority, however, presented evidence that suggested that this loan margin be at least 0.54 percent, that is, a credit spread near 0.75 percent. An English translation of the court decision noted:

The Appeals Board used market data indices – from the analysts Bloomberg , Reuters, JP Morgan (LIBOR USD), JP Morgan (EURIBOR) and Barclays Capital. The state has presented the following corrected numbers for the Court of Appeal:

• Bloomberg: 68 basis points
• Reuters: 54 basis points
• JP Morgan (USD): 81 basis points
• JP Morgan (EUR) 64 basis points
• Barclays: 77 basis points (88 basis points)

The indices express the estimated margins. Estimates are based on real transactions, but the data from these are adjusted and adjusted to different maturities and ratings to form normalized margin curves. The market transactions on which the indices are based can deviate considerably from the point estimate.

The taxpayer objected to the use of point estimates, insisting that a range based on individual transactions must be constructed:

It is stated that the arm's length principle presupposes a comparison with specific transactions. Some loans in the market are shown with the corresponding or lower interest rates, which are given approximately at the same time and to companies with a rating between A and AAA. An overview of nine loans has been presented, where the margins are between 20 and 65 basis points. EPNI has adjusted the margins for different credit ratings and calculated adjusted margin, assuming AA + ratings to be between 7 and 46.5 basis points.

The court disagreed, relying on the testimony of Espen Henriksen:

Any statistical-economic estimate will be associated with a standard deviation. Close to any loan, there may be distinctive circumstances that are not captured by maturity and credit rating, but still appreciated by the market. The curve that shows the relationship between margin and credit rating (and thus the margin or a given credit rating) is estimated to represent the systematic relationship between maturity and credit rating and interest rate margin. When certain issues do not fall right on the curve, this may be due to particular circumstances beyond credit rating and maturity by considering these issues that the market participants take into account once they have appreciated them. 

The loan information from Reuters cited by the court decision would have suggested an intermediate position where the arm’s length intercompany rate would be the 3-month Norwegian interbank rate plus 55 percent. Since the difference between the 3-month NIBOR rate and the 3-month Norwegian Treasury bill rate was 0.2 percent, this evidence suggested that the appropriate credit spread should have been 0.75 percent.

Concluding comments

A standard model for evaluating whether an intercompany interest rate is arm’s length can be seen to have two components – the intercompany contract and the credit rating of the related party borrower. Properly articulated intercompany contracts stipulate: the date of the loan; the currency of denomination; the term of the loan; and the interest rate. The first three items allow the analyst to determine the market interest rate of the corresponding government bond. This intercompany interest rate minus the market interest rate of the corresponding government bond can be seen as the credit spread implied by the intercompany loan contract.

While the evaluation of fixed rate intercompany loans is a straightforward application of the standard model, floating rate loans based on interbank rates such as LIBOR involve loan margins rather than a fixed interest rate. The loan margin would be equivalent to the credit spread only if the base rate were the same as the Treasury bill rate for the relevant currency. If not, the TED spread must be added to the loan margin to appropriately measure the credit spread for intercompany loans denominated in U.S. dollars. For intercompany loans denominated in other currencies, a similar spread between the relevant interbank rate and the interest rate on that nation’s corresponding Treasury bill rate should be calculated and then added to the loan margin. Domestic interbank rates have historically had market rates similar to LIBOR rates. The transition to an alternative base rates after the demise of LIBOR should continue to evaluate the relationship between the selected alternative base rate and the corresponding Treasury bill rate in order to properly measure the credit spread implied by the intercompany loan contract.