Recently, the FASB opened up its "Accounting for Financial Instruments and Revisions to the Accounting for Derivative Instruments and Hedging Activities—Financial Instruments (Topic 825) and Derivatives and Hedging (Topic 815)" to public comment: a process which seeks to establish and develop "standards that generally require the increased use of fair value in financial reporting" when it comes to accounting for loss provision on financial assets and trading liabilities "not held for trading" (i.e. as part of the "banking book"). In other words, this is a direct attempt at providing some much needed transparency when it comes to the maze of liabilities that are "held to maturity" and thus, according to current FASB regulations, exempt from Fair Value, and Mark To Market adjustments. It is precisely this toxic hodge podge of bad loans that currently makes up the bulk of bank books, as firms like JPM and Wells Fargo are not required to make these to anything resembling reality, assuming fair value is equivalent to par throughout the life of the loans, and which as a result are materially mispriced. It is this "weakest link" that will inevitably serve as the next financial crisis focal point, once price discovery is forced upon all these "assets." One of the firms providing their input on this critical topic is none other than Goldman Sachs. Since Goldman, unlike the other TBTFs is merely a hedge fund, and is not reliant on warehousing loans (although it sure was good at originating CDOs), the bank sees to gain (but more importantly its competitors have lots to lose), should fair value of of such banking book assets be market to market. Which explains why Goldman's Matthew Schroeder says that: "the current model does not provide investors in large, complex financial institutions with an accurate picture of a company’s financial position and does not foster sound risk management. This latter point is crucial, as poor risk management was at the heart of the financial crisis. The consequences of such decisions can have a significant effect on financial stability." It is not surprising that on this one aspect, Goldman will be all for enhanced transparency: after all, Goldman is the Wall Street firm that has the least reliance on the traditional Wall Street model, as the bulk of its revenues come in the form of daily trading profits, with any risky holding promptly sold off to other more gullible investors.
Yet where the letter gets interesting, is the very detailed explanation by Schroeder, of the "Lend To Play" practice, better known as "relationship loans" involved as part of IPO underwriting management syndicate allocation. In simple terms, according to Goldman, Wall Street IPO managers are only allowed on the IPO team if they commit to providing a loan to the same company, typically on far less than advantageous terms to the underwriter (a process Goldman explains as equivalent to writing a credit default swap on the issuer for the difference between the NPV of the full value of a revolver or a term loan and the periodic commitment fee). And since it is the bank's money that effectively is part of the firm's capital structure, it makes it all too clear that banks have a definitive bias to be extremely bullish in their sell side research on firms which they IPO, as it is their money, albeit higher in the capital structure, that could be impaired should the IPO candidate not trade up as expected.
This was most recently exhibited by GM, which demanded that all firms who are part of its IPO syndicate provide the firm with credit facilities, if not direct debt investments.
This is a glaring conflict of interest, yet the fact that a Wall Street firm has no problem describing this process as a daily occurence makes one wonder just how much fear of regulatory retribution or, heaven forbid, enforcement there is (hint: none at all).
Here is the process described in Goldman's own words:
It is common practice for borrowers to require that a bank participate in the borrower’s revolving credit facility (“revolver”) and/or the borrower’s funded term loan (“term loan”) as a condition of receiving future underwriting business. In many situations the borrower will be very explicit and inform the bank that it will not be permitted to participate in future underwriting business without participating in these “relationship loans.” In some situations, the bank will also be required to participate in a relationship loan in order to participate in future financial advisory assignments (such as a merger transaction) or obtain asset management business from the borrower (e.g., managing its cash balances).
Goldman also makes it clear, that as a result of ancillary kickbacks in the form of other transaction fees, the banks are "incenivzed", despite their will, to price such "relationship loans" not at fair value, but at a price that is disadvantageous to the primary issuer. Of course, once it is out in the open market, secondary trading takes a cue from the primary issuance price, and end buyers result in overpaying for loans that have substantially more endogenous risks. And unlike the primary issuer of the loan, these secondary purchasers do not have the hedge of secondary revenue streams to cover losses associated with a "relationship loan" gone bad.
Borrowers frequently condition future underwriting and advisory business on relationship loans because, without this incentive, they would be unable to obtain these loans on advantageous terms. In other words, borrowers are using the fees and other benefits associated with contemporaneous or future underwriting and advisory transactions to induce banks to make term loan commitments even though the terms of such relationship loans are off-market.
While, Goldman admits, this is not a brand new phenomenon, it has gotten substantially more acute in recent years:
The demand for relationship loans as a condition to participating in future business has become even more prevalent over the past several years. Today, in many initial public offerings (“IPOs”) that involve a company with a significant amount of debt, or the need for a revolving line of credit, the company will often require a loan commitment from any underwriter that wants to have a significant role in the offering. Likewise, for investment grade borrowers who access the commercial paper market, banks will generally be required to commit to the borrower’s back-up credit lines in order to participate in future bond offerings by the issuer. Many companies refuse to execute any business with banks that do not provide credit extension.
To illustrate this phenomenon, if one looks at a list of recent and pending IPOs involving companies with significant bank debt, there will generally be symmetry between the lead underwriters for the IPO and the composition of the banks in the company’s credit facility and term loan. In many transactions, the fees generated by each bank for underwriting the IPO generally will be proportionate with their lending commitments.
The “lend to play” practice is equally prevalent in the case of public companies that demand relationship loans from banks in order to participate as an underwriter in their future debt offerings. Many public companies will generally not invite a bank to serve as an underwriter without receiving a significant lending commitment from such bank.
Where it gets even more interesting is that Goldman acknowledges that the lender is fully aware of the explicit mispriced risks in the "contract" or issuance price of the loan.
In these relationship loan situations, it is generally clear, based on discussions with the borrower and the terms of the arrangements, that the bank making the loan has reason to believe there is a significant difference between the contract price and the fair value of such loan or loan commitment and there is reliable evidence to support this difference. In addition, in these situations there will clearly be “other elements” present, notably the expectation of other fees the bank will earn by participating in such a relationship loan. In other words, there is a clear link between banks that are willing to extend credit on off-market terms in order to capture the future fees related to various underwriting and other financial advisory activities.
In the following Appendix, Goldman present a real life example of precisely this phenomenon in action, and derives the precise impairment associated with the misrepresentation of the fair value of the loan on the books of the holders. In other words, in a perfect world, lenders of "relationship loans" would be forced to immediately write down their investment materially, with the loss amortizing to par over the life of the note. This makes a lot of intuitive sense, as this impairment is effectively the counterparty risk associated with agreeing to a long-term credit instrument with an entity that may very likely not survive through the duration of the original loan. Yet this is something that never happens in the banking community, and in fact, the opposite is prevalent, where banks misprice loans far to the upside in order to pad their underwater capitalization ratios (see Repo 105-like scams committed by virtually every single bank with a loan book).
Setting aside the fact that loans, both those held on books, and traded in the secondary market, are by implication largely mispriced (although one could make the argument that sophisticated investors should be able to adjust for this syndicator arbitrage... of course one could also make that claim of CDO purchasers in 2006 and 2007), the bigger question is just how major the conflict of interest is to the firms that serve as both lender and IPO underwriter: does one realistically see the possibility of a bank issuing anything less than a Buy or a Strong Buy in a name in which it was forced off the bat to put in debt capital, and whose equity buffer could be largely impaired if the same firm's Sell rating were to decimate the equity market cap? Perhaps, once the SEC is done dismantling HFT, it can take a look at the practice of "Lend to Play" which could promptly become the biggest threat to investor wealth, as more and more companies are going public in anticipation of a market peak.
And going back to the original topic of transparency, Goldman takes the following stab at those firms who will fight tooth and nail to block increased transparency into the banking book model:
Some constituents do not agree that fair values for financial assets are relevant because they have access to stable sources of funds (core deposits) and believe they will maintain those sources in difficult funding environments. While that is often the case, funding sources can change unexpectedly and, when they do, the consequences are usually negative.
We, for one, are not holding our breath on the FASB doing anything that forces banks to disclose the sorry state of their "held to maturity" books. Any inkling of that occuring would result in an immediate hit to the tune of at least $250 billion and possibly far more: this is an amount which with the recent weakness in the XLF, banks would simply not be able to sustain, endless blatherings to the contrary by Dick Bove aside. We are far more concerned by the implications of the just disclosed "cast study" by Goldman, which exposes yet another perspective in the endless conflict of interest game, so shrewdly played by Wall Street each and every day.
For those curious, here is a real world example of the Lend To Play phenomenon presented by Goldman:
Appendix: Real World Examples of the “Lend to Play” Phenomenon
Example #1: Banks Required to Commit to Credit Facility to Participate in Underwriting and IPO
The following is a typical example of a transaction involving an IPO where the company going public required the banks competing for the IPO business to commit to a credit facility in order to participate as an underwriter in the IPO. We have not used actual names in this example, have simplified some facts and rounded off some of the figures in order to avoid disclosing any non-public information.
Company X is a private company owned by a group of financial sponsors (“Sponsors”). Company X solicited a number of large banks to serve as underwriters in its IPO. The solicitation process included a formal “request for proposal” (RFP) which included a number of specific questions and requirements that the banks had to address during their “pitch.” One of the requirements in the RFP was for each bank had to commit to a 3 year extension of an existing revolving credit facility (“Revolver”) that was set to mature in 2.5 years time in order to have a significant role in the IPO. There were minimum commitment sizes based on the title awarded to each bank, and a bank that did not make this minimum commitment would not be considered a candidate for a significant role in the IPO regardless of its qualification.
The expected size of the IPO was $5 billion, although the actual size could have been larger or smaller depending upon a variety of factors, including market conditions at the time of the offering. The total fees payable to the underwriters in the IPO was estimated to be approximately $175 million (or 3.5% of deal size). In addition, the underwriters in the IPO would be in an advantageous position to serve as the underwriters for any future follow-on equity offerings as the Sponsors sell down their retained stake in Company X. These underwriters would also be likely to lead future bond offerings for Company X and would be well positioned to earn other advisory fees (e.g., for merger transactions). Thus, the total fee potential for the underwriters is significant, but they can only participate in earning those fees if they also commit to the Revolver.
Company X has significant outstanding debt and has a non-investment grade credit rating from the major credit rating agencies.
The total size of the extended Revolver is $2 billion and substantially all of this commitment will be sourced from banks that are underwriters in the IPO.
Key Terms of the Extended Revolver
The key terms of the Revolver are as follows:
- Maturity: 5.5 years
- Upfront fees: 0.0% of commitment amount
- Annual fees for undrawn amounts: 0.5%
- Interest rate on drawn amounts: floating rate equal to one-month LIBOR plus a spread of 2.75% (L + 275)
- Security: any amount drawn on the revolver will be secured by most of Company X’s assets
- Covenants: the Revolver will contain both “incurrence” and “maintenance” covenants. The covenants are generally less restrictive than the covenants that would likely have been demanded if the Revolver was sourced from third party lenders that were not induced to participate in the Revolver by the opportunity to earn underwriting fees associated with the IPO and other transactions.
Fair Value of the Revolver
In this situation, there is reliable evidence that the fair value of the Revolver is significantly less than the contract price (i.e., the Revolver commitment represents a net liability at the contract date). However, the banks participating in the Revolver are willing to participate because of “other factors,” specifically the other fees they expect to earn in connection with the IPO and subsequent underwriting and advisory transactions. (As discussed above, under current accounting rules, participants that hold the Revolver in their “banking book” would not properly record the pricing discount thereby overstating their investment banking revenues.)
There are several forms of reliable evidence to calculate the fair value of the Revolver at the contract date.
Method One: Comparing Undrawn Fee to Credit Spread on Funded Loan
One method to estimate the fair value of the Revolver involves comparing the interest rate charged by the lenders on amounts borrowed under the Revolver relative to the annual fee on undrawn amounts. A lender under the Revolver is exposed to Company X’s credit risk even if the Revolver is undrawn because an undrawn commitment represents an obligation to lend to Company X through the maturity day of the Revolver. In other words, the undrawn Revolver is similar to a bank writing a credit default swap on Company X’s credit since the bank has been exposed to Company X’s credit risk without actually funding a loan.
Given that the underlying index (i.e., LIBOR) represents an approximation of a risk-free rate, one can think of the interest rate on drawn amounts to represent a credit spread of 2.75%. However, the lender under the Revolver is only receiving an undrawn fee of 0.5% to take credit risk on undrawn amounts which equates to at least 2.25% per year less than an arm’s-length rate. Over five and a half years, this represents approximately 8.8% on a present value basis net of all fees.
Method Two: Comparing Undrawn Fee to Cost of Credit Default Swaps
A second method that can be used to calculate the fair value of the Revolver is to look at the cost of purchasing insurance against a default by Company X. Given that an unfunded Revolver is akin to writing credit default protection, a lender can hedge its exposure by purchasing insurance in the form of loan credit default swaps (LCDS). This would put the lender in a “neutral” position since any loss incurred by the bank on the Revolver as a result of a default (relating to amounts drawn prior to default) would be offset by the “gain” on the LCDS that it purchased.
At the time the Revolver was entered into, Company X had senior secured LCDS contracts that provided holders with protection on debt obligations with the same level of seniority as the Revolver. The cost of purchasing a 5-year LCDS was approximately 2.6% per year. Accordingly, a bank participating in the Revolver that wanted to hedge its risk could purchase LCDS protection in an amount equal to its Revolver commitment.
The cost of purchasing this protection for 5.5 years (on a present value basis) would be approximately 10.2% of the Revolver commitment whereas the upfront and ongoing fees associated with the Revolver commitment total 2.0% on a PV basis. Accordingly, under this hedging approach, the fair value of a Revolver commitment illustrates an upfront cost of approximately 8.2% of the notional commitment.
Based on the above methodologies, the fair value of the Revolver is approximately 91-92% of the notional commitment amount. This would imply an upfront loss of approximately $8-9 million for a bank that agrees to a $100mm commitment. Under current accounting rules, banks that hold their Revolver commitments in their “banking book” would not properly record the pricing discount even though the Revolver commitment was made on these borrower-friendly terms with the expectation that such lending would create other opportunities to generate fees for the lending bank.
Full Goldman Sachs response.