Bank of America and Its High Marginal Cost of Funds

Conclusion: We find that Bank of America suffers a funding disadvantage of as much as 0.30% versus major competitors like Wells Fargo & Co. (WFC). We also find that many other banks among the top 20 banks in the United States can out-compete Bank of America from the perspective of the marginal cost of wholesale funding. Bank of America can only lower its marginal cost of funds by lowering the risk profile of its liabilities with more capital, more profits, and lower costs in other aspects of its business.

The Analysis
Step one in today’s analysis is to report the composite credit spreads of the four “too big to fail” banks underlying the U.S. Dollar Cost of Funds Index ^TM. The big four banks represented 4 of the 8 most heavily traded bond issuers in the United States in this chart of fixed-rate non-call senior debt issues. For information on bond types excluded from the construction of the index, please see the note on the index below. Here is a summary of trading volume on September 8, 2014:

There were 2,604 trades in 414 bond issues representing $698 million in principal of the four reference banks underlying the U.S. Dollar Cost of Funds Index. There is variation in credit quality among the four banks, which we have noted in other posts. For purposes of this note, however, we focus on the composite index alone. The credit spreads fitted to the September 8, 2014 trade data are shown here:

At the common “on the run” maturities, the U.S. Dollar Cost of Funds are summarized here, both in terms of credit spreads and yields.

At the five year point, the composite marginal costs of funds for the big four banks is a credit spread of 0.885% and an all-in yield of 2.605%, compared to a 1.72% yield on the five year Treasury as reported by the U.S. Department of the Treasury.

Comparison of Bank of America Credit Spreads with the
U.S. Dollar Cost of Funds Index, September 8, 2014

Next, we follow the same credit spread calculation procedures used in the U.S. Dollar Cost of Funds Index for Bank of America, using observable bond trades on September 8, 2014. We calculate the difference between the trade-weighted average credit spreads and the U.S. Dollar Cost of Funds credit spreads for the identical maturities for all observable bond trades of Bank of America Corporation with at least $1 million in trading volume. The bonds used are senior non-call fixed rate debt issues reported by TRACE via Market Axess. The results are shown in this table:

On a simple average basis, equally weighting each issue, Bank of America’s average funding differential versus the U.S. Dollar Cost of Funds Index is a premium, or extra credit spread, of 0.134%. This alone is worrisome, but a comparison with other large U.S. banks adds to the worry.

Comparing the Marginal Cost of Wholesale Funds with Other U.S. Banks
First, we look at the spreads on Wells Fargo & Co. (WFC) bond issues and compare them with the U.S. Dollar Cost of Funds Index. Using a simple average of the difference, we find that Wells Fargo & Co. has a marginal cost of funds 0.186% BELOW the U.S. Dollar Cost of Funds Index, an advantage of 0.186% + 0.134% = 0.320% over the marginal cost of funds for Bank of America. Even if we focus only on the two bond issues with more than $15 million in trading volume, where the spread versus the U.S. Dollar Cost of Funds Index is about 0.09%, Wells Fargo & Co. has a 0.22% advantage in the marginal cost of funding versus Bank of America.

We now look at other banks with observable bond trading on September 8, 2014.

Bank of New York Mellon (BK) had credit spreads that were 0.222% below the U.S. Dollar Cost of Funds Index. Fifth Third Bancorp (FITB) had credit spreads just 0.017% above the U.S. Dollar Cost of Funds Index. The KeyCorp (KEY) organization averaged 0.045% below the U.S. Dollar Cost of Funds Index. State Street Corporation (STT) and the U.S. Bancorp (USB) organization were 0.312% and 0.175% below the U.S. Dollar Cost of Funds Index respectively. All of these organizations have a lower marginal cost of wholesale funding below that of Bank of America.

Conclusion
We find that Bank of America suffers a funding disadvantage of as much as 0.30% versus major competitors like Wells Fargo & Co. (WFC). We also find that many other banks among the top 20 banks in the United States can out-compete Bank of America from the perspective of the marginal cost of wholesale funding. Bank of America can only lower its marginal cost of funds by lowering the risk profile of its liabilities with more capital, more profits, and lower costs in other aspects of its business. All of these steps will have the effect of lowering the default probability of Bank of America and the associated credit spreads on its bonds. The often discussed “too big to fail” implicit guarantee of the U.S. government, for whatever reason, is not sufficient to allow Bank of America to be competitive in its marginal cost of funds.

Appendix

Background on the U.S. Dollar Cost of Funds Index^TM
Measuring the marginal cost of funds for a financial institution is a critical calculation from both a management perspective and a regulatory perspective. From a management perspective, if one’s own firm faces a funding disadvantage versus a group of banks who are “too big to fail,” that has dramatic implications for corporate strategy. From a regulatory perspective, favoring a group of financial institutions with an implicit guarantee of survival can create an anti-competitive subsidy as an unintended consequence. In this section, we explain the U.S. Dollar Cost of Funds Index^TM from Kamakura Corporation which makes quantification of funding advantages and disadvantages a practical daily reality.

Background on the U.S. Dollar Cost of Funds Index
On a typical day in the U.S. corporate bond market, there are significant bond trades in fewer than 20 of the 6,730 banks insured by the U.S. Federal Deposit Insurance Corporation. Moreover, there have been trades in only 14 U.S. banking legal entities in the credit default swap market since July 2010. That means for the overwhelming majority of financial institutions, an arm’s length benchmark for marginal funding costs is the only alternative since there is literally no trading in their own name. For many years, the interest rate swap curve was used as this benchmark for transfer pricing, asset and liability management, and internal probability assessment by the majority of large banks. With the unveiling of the Libor scandal and the associated prolonged period of negative interest rate swap spreads to the U.S. Treasury yield curve, the swap curve has lost credibility as a reliable benchmark for the marginal cost of funds.

For this reason, a new benchmark is needed that has these characteristics:

• It should realistically measure the marginal cost of new funds to the financial institutions with the largest trading volume.
• It should be based on public information that is readily verifiable by any investor or other market participant.
• It should be based on actual trades, not estimates of trading costs.
• It should be a calculation that can be done without the input of information or the cooperation of any of the financial institutions named in Libor-related lawsuits.
• It should be based on such a large volume of trades that manipulation of the benchmark would be difficult, if not impossible.
• It should be a composite of data relating to a number of financial institutions so that no one institution’s cost of funds is used as the benchmark.
• The calculation methodology should be transparent and replicatable by any third party.
• The calculation agent should not be a legal entity or other organization directly or indirectly controlled by the group of financial institutions named in Libor-related lawsuits.

The U.S. Dollar Cost of Funds Index^TM from Kamakura Corporation
Kamakura Corporation has created just such an index, labeled the “U.S. Dollar Cost of Funds Index.^TM” The U.S. Dollar Cost of Funds Index ^TM measures the trade-weighted cost of funds for the largest deposit-taking U.S. bank holding companies. The index is a credit spread, measured in percent and updated daily, over the matched maturity U.S. Treasury yield on the same day. The current bank holding companies used in determining the index are Bank of America Corporation (BAC), Citigroup Inc. (C), JPMorgan Chase & Co. (JPM), and Wells Fargo & Company (WFC). The index is an independent market-based alternative to the Libor-swap curve that has traditionally been used by many banks as an estimate of their marginal cost of funds. Kamakura Corporation is the calculation agent, and the underlying bond price data is provided by TRACE and the U.S. Department of the Treasury. The National Association of Securities Dealers launched the TRACE ( Trade Reporting and Compliance Engine) in July 2002 in order to increase price transparency in the U.S. corporate debt market. The system captures information on secondary market transactions in publicly traded securities (investment grade, high yield and convertible corporate debt) representing all over-the-counter market activity in these bonds. For each of the reference names used in the calculation of the index, Kamakura Corporation assembles bond data as follows:

• All bond trades of that reference name are gathered from TRACE.
• All bonds which are not fixed rate debt issues are dropped.
• All bonds which are not “non-call” (other than “make whole” calls) are dropped
• All bonds which do not pay interest semi-annually are dropped.
• All “survivorship option” bonds are dropped.
• All bonds with a changing interest rate or principal amount tied to any index are dropped. This eliminates structured products.

The result is a large number of straight non-call fixed rate senior debt issues which traded that day. This contrasts with the Libor calculation procedure in which there are zero trades for $0 on 0 instruments.

Assembling the Composite Index
The composite credit spread is then calculated as follows:

1. The traded-weighted average yield is calculated for each of the underlying bonds.
2. The matched-maturity U.S. Treasury yield is interpolated from the “on the run”
   maturities published by the U.S. Department of the Treasury.
3. The trade-weighted credit spread for each issue is calculated by subtracting the interpolated matched maturity Treasury yield from the trade-weighted average yield.
4. Since the spread varies by years to maturity, a curve is fit to all observable bonds using a cubic function of the years to maturity t on each issue.

   Credit spread = a + bt + ct² + dt³

   The coefficients a, b, c, and d are calculated using ordinary least squares

   The terms, trading volume, and pricing for each underlying bond, the matched maturity U.S. Treasury yields, the coefficients of the index function for that day, and the “on the run” values of the index are provided by Kamakura Corporation to index clients.

Copyright ©2014 Donald van Deventer