- 1. Overview
- 2. Etymology
- 3. Cultural Impact
Net Capital Rule
The Net Capital Rule , formally known as the “uniform net capital rule,” is a pivotal piece of financial regulation established by the U.S. Securities and Exchange Commission (SEC) in 1975. Its primary objective was to directly govern and ensure the financial stability of broker-dealers , thereby safeguarding their ability to meet their financial obligations to customers and other creditors. Broker-dealers, entities engaged in the trading of securities on behalf of clients (brokers) or for their own proprietary accounts (dealers), are inherently exposed to market fluctuations and operational risks. This rule was designed to provide a robust framework to mitigate these risks.
The core of the Net Capital Rule mandates that broker-dealers must value their securities holdings at their current market prices . Crucially, to account for inherent volatility and potential declines in value, a mandatory haircut , or discount, is applied to these market values. The magnitude of this haircut is not arbitrary; it is meticulously determined by the specific risk characteristics associated with each individual security. This process of applying haircuts serves a critical purpose: it allows for the computation of the liquidation value of a broker-dealer’s assets. This liquidation value is then compared against the firm’s total non-subordinated liabilities. The rule demands that a broker-dealer must possess sufficient liquid assets to cover all such liabilities and, importantly, maintain a “cushion” of required liquid assetsâthis is the titular “net capital.” This cushion is the financial buffer intended to ensure the full payment of all obligations owed to customers, even in the event of delays or difficulties in liquidating assets.
2004 Amendment and its Aftermath
A significant amendment to this established rule was voted on by the SEC on April 28, 2004. This amendment was particularly notable for its provision allowing the largest broker-dealersâthose with “tentative net capital” exceeding $5 billionâto seek exemptions from the traditional haircut methodology. Upon receiving SEC approval, these select firms were granted permission to employ sophisticated mathematical models to calculate their security haircuts. This approach was permitted to align with international standards already in use by commercial banks, signaling a move towards harmonized global financial regulation.
The ramifications of this 2004 rule change became a focal point of intense debate and analysis, particularly in the wake of the 2008 financial crisis . Numerous commentators identified the amendment as a substantial contributing factor to the crisis. The argument posited was that by allowing these large investment banksâspecifically Bear Stearns , Goldman Sachs , Lehman Brothers , Merrill Lynch , and Morgan Stanley âto utilize their own models for haircut calculations, the SEC inadvertently permitted them to dramatically increase their leverage , defined as the ratio of their debt to their equity. Financial reports filed by these institutions indeed showed a marked increase in their leverage ratios between 2004 and 2007. However, a closer examination of financial reports filed prior to 2004 revealed that four of these five firms had actually exhibited higher reported leverage ratios in the years preceding the rule change.
Despite the controversy, the 2004 rule change remained in effect. The broker-dealers that obtained SEC approval to use the alternative haircut computation method continued to do so, albeit with modifications that were implemented on January 1, 2010.
The Net Capital Rule and the 2008 Financial Crisis
The period beginning in 2008 saw a surge in commentary linking the 2004 amendment to the SEC’s net capital rule as a primary driver of the 2008 financial crisis by enabling substantial increases in investment bank leverage.
This perspective was notably articulated by Lee A. Pickard, who had served as the Director of the SEC’s Division of Market Regulation when the uniform net capital rule was originally adopted in 1975. In an August 2008 commentary, Pickard argued that prior to the 2004 amendment, broker-dealers were effectively constrained in their debt-to-capital ratios, typically operating at around 12 times their net capital, and at even lower ratios in practice. He contended that had these firms remained subject to the pre-2004 net capital rule, they would have been unable to accumulate such high debt levels without first bolstering their capital bases. A widely cited New York Sun article from September 2008 quoted Pickard directly, identifying the 2004 rule change as the principal reason for the significant losses incurred by investment banks.
Further amplifying this viewpoint, a prominent October 2008 New York Times article, titled “Agency’s ‘04 Rule Let Banks Pile Up New Debt,” asserted that the net capital rule applied to the “brokerage units” of investment banks and that the 2004 change essentially created an “exemption” from an older rule that had capped their debt accumulation. The article suggested that this deregulation unleashed “billions of dollars held in reserve against losses,” leading to a dramatic escalation in investment bank leverage.
In the ensuing period of late 2008 and early 2009, influential scholars such as Alan Blinder , John Coffee , Niall Ferguson , and Joseph Stiglitz echoed these sentiments. They explained that the older net capital rule imposed a leverage limit of approximately 12 or 15 to 1 on investment banks. Following the 2004 rule change, which relaxed or eliminated this restriction, their leverage ratios reportedly soared to 30 or even 40 to 1. This leverage data was often drawn from financial reports filed by the Consolidated Supervised Entity (CSE) Holding Companies with the SEC.
Daniel Gross , writing for Slate , critically observed that the ability of investment banks to go public allowed them to grow larger and subsequently influence the regulatory system to their advantage. He characterized the 2004 SEC rule change, which permitted increased debt on their books, as perhaps the “most disastrous decision of the past decade,” noting it was made at the behest of the CEOs of the “Gang of Five” largest investment banks.
SEC Response to Crisis Commentary
In response to the mounting criticism and in connection with an investigation into the SEC’s role in the collapse of Bear Stearns , the SEC’s Division of Trading and Markets issued a rebuttal in late September 2008. They maintained that the arguments conflated leverage at the Bear Stearns holding company, which was not subject to the net capital rule, with leverage at its regulated broker-dealer subsidiaries. Furthermore, the SEC asserted that both before and after the 2004 rule change, the broker-dealers covered by the rule were subject to a net capital requirement equivalent to 2% of customer receivables, rather than an explicit 12-to-1 leverage test.
Erik Sirri, then Director of the SEC’s Division of Trading and Markets, elaborated on this position in an April 2009 speech. He stated that the 2004 rule change did not alter the “basic” net capital rule that contained a leverage limit (though he noted this limit excluded significant broker-dealer debt). He further clarified that an “alternative” net capital rule, established in 1975 and lacking a direct leverage limit, applied to the broker-dealer subsidiaries of the five largest investment banks and other large broker-dealers. Crucially, Sirri argued that neither version of the net capital rule was designed, nor did it operate, to constrain leverage at the investment bank holding company level, where the concentration of leverage and risk often resided in business units outside of the broker-dealer subsidiaries.
A July 2009 report by the Government Accountability Office (GAO) further supported the SEC’s stance. The GAO reported that SEC staff indicated that CSE Brokers did not significantly increase their proprietary positions after adopting the reduced haircuts permitted by the 2004 rule change. They also stated that leverage within these CSE Brokers was primarily driven by customer margin loans, repurchase agreements, and stock lending, which were marked daily and adequately collateralized, thus posing minimal risk. The GAO report also noted that officials from a former CSE Holding Company claimed they did not join the CSE program with the intent of increasing leverage. The GAO did confirm that leverage at CSE Holding Companies was indeed higher at the end of 1998 than at the end of 2006, just prior to the 2008 financial crisis . The SEC, in a comment letter included in the GAO report, reiterated the points made in the 2009 Sirri Speech, characterizing the commentators’ views as a “mischaracterization” of the 2004 rule change and denying it was a major contributor to the crisis. The SEC also noted that “tentative net capital” levels at CSE Brokers remained “relatively stable” or even “increased significantly” after the rule change.
Status of the 2004 Rule Change
It is a widely acknowledged fact that all five of the investment bank holding companies directly affected by the 2004 rule change no longer exist as independent entities or have since converted into bank holding companies. Less commonly noted is that the broker-dealer subsidiaries originally owned by these investment bank holding companies continue to operate under the SEC’s net capital rule, utilizing the alternative net capital computation method established by the 2004 rule change. The key difference now is that these former CSE Brokers, such as Citigroup Global Markets Inc. and JP Morgan Securities Inc., are now part of bank holding companies that are under the consolidated supervision of the Federal Reserve, rather than the SEC’s CSE Program.
In January 2010, SEC Chair Mary Schapiro , in testimony before the Financial Crisis Inquiry Commission (FCIC), stated that SEC staff had informed CSE Brokers that they would be required to apply standardized net capital charges to less liquid mortgage and other asset-backed securities, rather than relying on financial models for these calculations. These new requirements became effective on January 1, 2010. Schapiro also indicated that the SEC was reviewing the broader implications of the 2004 rule change and considering whether the “alternative net capital computation” system should be significantly modified, and, more generally, whether minimum net capital requirements for all broker-dealers should be increased.
Background to and Adoption of the Net Capital Rule
The SEC’s “Uniform Net Capital Rule,” often referred to as the “Basic Method,” was formally adopted in 1975. This adoption followed a period of significant financial market turmoil and a crisis in broker record-keeping that spanned from 1967 to 1970. Concurrently with the adoption of the Basic Method, the SEC also established the “Alternative Net Capital Requirement for Certain Brokers and Dealers,” known as the “Alternative Method.”
The SEC had maintained a net capital rule in some form since 1944. However, prior to 1975, broker-dealers subject to more stringent capital requirements imposed by recognized exchanges, such as the New York Stock Exchange (NYSE), were often exempted from direct SEC oversight in this regard. The 1975 uniform net capital rule retained many elements of the existing SEC rule but incorporated several more rigorous requirements that had been established by the NYSE.
Rule Application: Broker-Dealers, Not Parent Holding Companies
A critical distinction of the Net Capital Rule, both in its Basic and Alternative Methods, is that it applies exclusively to broker-dealers themselves, not to their parent holding companies. The SEC has never imposed a direct net capital requirement on the holding company parent of a broker-dealer. Brokers, by definition, execute securities transactions on behalf of their clients. The Securities Exchange Act of 1934 granted the SEC the authority to regulate the financial condition of broker-dealers specifically to offer customers a degree of assurance that their broker could meet its contractual obligations.
Holding companies that owned broker-dealers were, in essence, treated as “unregulated” entities. Parties extending credit to these holding companies did so based on their own risk assessments, without the direct regulatory oversight of the company’s financial condition. In practice, the “independent check” on the financial health of these holding companies often fell to rating agencies . To engage in dealer and other credit-sensitive activities, large investment bank holding companies meticulously managed their leverage and overall financial condition to achieve at least an “A” credit rating, which was generally considered essential for such operations. Each of the investment banks that eventually became a CSE Holding Company emphasized the importance of a “net leverage” measure. This metric excluded collateralized customer financing arrangements and other “low risk” assets when calculating “net assets,” presenting a considerably lower leverage ratio than the “gross leverage” ratio derived from total assets and shareholders’ equity. This adjusted ratio was utilized by at least one rating agency in assessing the capital strength of investment banks.
Goals and Tests of the Net Capital Rule
The Self-Liquidation Principle
The SEC has consistently articulated that the fundamental purpose of the net capital rule is to ensure that “every broker-dealer to maintain at all times specified minimum levels of liquid assets, or net capital, sufficient to enable a firm that falls below its minimum requirement to liquidate in an orderly fashion.” The Basic Method endeavors to achieve this by measuring a broker-dealer’s “liquid assets” against the majority of its unsecured indebtedness. These “liquid assets” are intended to serve as a “cushion” to cover the full repayment of that unsecured debt. In contrast, the Alternative Method focuses on measuring “liquid assets” against the obligations owed by customers to the broker-dealer. Here, the “liquid assets” act as a “cushion” for the broker-dealer’s ability to recover the full amounts due from its customers.
Despite these differing endpoints, both the Basic and Alternative Methods commence with the calculation of “net capital.” To arrive at this figure, a broker-dealer first calculates its equity under Generally Accepted Accounting Principles (“GAAP”). This involves marking all securities and other assets to their current market values. Subsequently, this equity calculation is reduced by the value of any “illiquid assets” held by the firm. Any “qualifying subordinated debt” owed by the firm is then added to this adjusted total. This preliminary figure represents the broker-dealer’s “tentative net capital.” Following this calculation, the broker-dealer makes further adjustments by reducing the value of its securities. These reductions are applied as percentage deductions, or “haircuts,” based on the perceived “risk characteristics” of each security. For instance, a stock portfolio might incur a 15% haircut, while a highly stable, less risky U.S. treasury bond with a 30-year maturity might only be subject to a 6% haircut.
In theory, a calculated “net capital” greater than zero signifies that a broker-dealer’s liquid assets could be liquidated to repay all its outstanding obligations, even those not yet due, with the exception of any qualifying subordinated debt which the net capital rule treats as equity. However, both the Basic and Alternative Methods incorporate a critical second step: broker-dealers are required to compute a “cushion of liquid assets in excess of liabilities to cover potential market, credit, and other risks if they should be required to liquidate.” This cushion is also vital for covering ongoing operating expenses during the liquidation process, a factor of particular importance for smaller broker-dealers with minimal absolute dollar amounts of required net capital.
The “cushion” or “buffer” concept is central to understanding the required net capital. This buffer is designed to absorb continuing operating costs during liquidation and any exceptional losses incurred from selling assets that have already been discounted in the net capital calculation. Therefore, the required level of net capital is measured against a significantly more restricted set of liabilities or assets than might be initially assumed by commentators. This measurement is executed in the “second step” of both the Basic and Alternative Methods.
Basic Method: A Partial Unsecured Debt Limit
For its second step, the Basic Method adopts a traditional liability coverage test. This test had long been imposed by the New York Stock Exchange (“NYSE”) and other “self-regulatory” exchanges on their member firms, as well as by the SEC on broker-dealers not affiliated with such exchanges. Under this approach, the SEC mandated that a broker-dealer operating under the Basic Method must maintain “net capital” equivalent to at least 6-2/3% of its “aggregate indebtedness.” This is commonly understood as a 15-to-1 leverage limit, implying that “aggregate indebtedness” could not exceed 15 times the amount of “net capital.” However, the definition of “aggregate indebtedness” excluded certain categories, such as “adequately secured debt,” subordinated debt, and other specified liabilities. Consequently, even the Basic Method did not impose a strict 15-to-1 limit on a broker-dealer’s overall leverage when calculated from a GAAP financial statement.
In practical terms, broker-dealers often rely heavily on secured borrowing, particularly through repurchase agreements. This means their leverage, as calculated from a GAAP balance sheet, would typically be higherâpotentially much higherâthan the ratio of their “aggregate indebtedness” to “net capital,” which is the actual “leverage” ratio tested under the Basic Method. The exclusion of secured debt from the “aggregate indebtedness” test bears a resemblance to, though it is broader than, the rating agency methodology of excluding certain secured debt when computing net leverage, as discussed previously.
Alternative Method: A Customer Receivable Limit
The Alternative Method was an optional compliance route for broker-dealers. To utilize it, a broker-dealer had to compute its “aggregate debit items” owed by customers in accordance with the “customer reserve formula,” which was established by the SEC’s 1972 rules governing the segregation of customer assets. Under its second step, a broker-dealer employing the Alternative Method was required to maintain “net capital” equal to at least 4% of its “aggregate debit items” owed by customers. This methodology operated on the assumption that funds received from customer receivables would, in conjunction with the net capital cushion of “liquid assets,” be used to satisfy the broker-dealer’s obligations to its customers and cover administrative costs during a liquidation. It is important to note, however, that the net capital held in the form of “liquid assets” was not required to be separately escrowed specifically for the exclusive benefit of customers.
The Alternative Method did not directly regulate the overall leverage of a broker-dealer. As the SEC itself explained when adopting the net capital rule, the Alternative Method “indicates to other creditors with whom the broker or dealer may deal what portion of its liquid assets in excess of that required to protect customers is available to meet other commitments of the broker or dealer.”
In 1982, the net capital requirement under the Alternative Method was reduced to 2% of customer indebtedness. This effectively meant that customer receivables could not exceed 50 times the broker-dealer’s net capital. However, neither the 2% nor the earlier 4% requirement translated into an overall leverage limit of 25-to-1 or 50-to-1 on a broker-dealer’s total assets. The capital requirement at these percentages was strictly tied to customer assets, specifically amounts owed by customers to the broker-dealer.
Early Warning Requirements
The 2009 Sirri Speech highlighted the significance of the “early warning requirements” under the net capital rule, describing them as the “effective limits” for the CSE Brokers. Beyond the $5 billion tentative net capital reporting threshold established for CSE Brokers in the 2004 rule change, the SEC mandated that broker-dealers provide notification to the SEC if their net capital fell below a specified level that was higher than the minimum requirement that would trigger a broker-dealer liquidation. Upon receiving such a notice, a broker-dealer would be subjected to closer regulatory scrutiny and would be prohibited from making capital distributions. Such distributions were also forbidden if they would themselves trigger an early warning requirement.
Each broker-dealer was obligated to issue an “early warning” if its net capital declined to less than 120% of the broker-dealer’s absolute minimum net capital requirement. For broker-dealers operating under the Basic Method, an “early warning” was triggered if their “aggregate indebtedness” exceeded 12 times their “net capital.” The 2009 Sirri Speech suggests that this particular requirement is the likely origin of the widely disseminated assertion that broker-dealers were subject to a 12-to-1 leverage limit. For broker-dealers utilizing the Alternative Method, including all CSE Brokers, an “early warning” was required if their “aggregate debit items” exceeded 20 times their net capital, meaning their net capital fell below 5% of aggregate debit items.
Post-1975 Experience Under the Net Capital Rule
The implementation of the uniform net capital rule in 1975 coincided with a period of profound transformation within the brokerage industry. The simultaneous elimination of fixed commissions is frequently cited as a factor that eroded broker profitability, compelling firms to place a greater emphasis on “proprietary trading” and other “principal transactions.”
Broker-Dealer Leverage Increased Post-1975
Following the enactment of the SEC’s uniform net capital rule, the reported overall leverage at broker-dealers experienced an increase. In a 1980 release proposing amendments to the “haircuts” used for net capital computations, the SEC noted that aggregate broker-dealer leverage had risen from debt-to-equity ratios of 7.44 and 7.45 to 1 in 1974 and 1975, respectively, to a ratio of 17.95 to 1 by 1979. The GAO later found that by 1991, the average leverage ratio for thirteen large broker-dealers it studied had reached 27 to 1, with the average among nine of those firms being even higher.
Large Broker-Dealers Favored the Alternative Method
The same 1980 SEC release highlighted a clear divergence in the application of the net capital rule. The majority of broker-dealers continued to use the Basic Method. However, large broker-dealers, which increasingly held the substantial majority of customer balances, opted for the Alternative Method. Reflecting this trend, all of the major broker-dealers owned by investment bank holding companies that would later become CSE Brokers after the 2004 rule change utilized the Alternative Method.
There were recognized barriers to adopting the Alternative Method. Firstly, the absolute minimum net capital requirement of $250,000 under the Alternative Method could be lower under the Basic Method if a broker-dealer deliberately limited its customer-facing activities. This effectively deterred many smaller broker-dealers from switching to the Alternative Method.
Secondly, to adopt the Alternative Method, a broker-dealer was required to compute its “aggregate debit balances” owed by customers according to the “customer reserve formula” stipulated in SEC Rule 15c3-3. Many smaller broker-dealers found it more convenient to comply with one of the three exemptions from Rule 15c3-3 requirements rather than invest in the operational infrastructure necessary for “fully computing” compliance with the customer reserve formula. These exemptions, however, imposed significant restrictions on a broker-dealer’s capacity to handle customer funds and securities.
Neither of these considerations was applicable to the CSE Brokers. Their net capital was measured in the billions of dollars, not hundreds of thousands. Furthermore, they were actively engaged in customer brokerage activities that necessitated the full computation of the customer reserve formula under Rule 15c3-3. Consequently, they had long been using the Alternative Method for many years prior to the establishment of the CSE Program. This choice generally resulted in a reduced net capital requirement for them.
2004 Change to the Net Capital Rule
In 2004, the SEC enacted an amendment to the net capital rule that offered a significant concession to broker-dealers possessing at least $5 billion in “tentative net capital.” These firms were permitted to apply for an “exemption” from the established method of calculating “haircuts.” Instead, they could compute their net capital using historical data-driven mathematical models and scenario testing, methodologies already authorized for commercial banks under the “Basel Standards.” This regulatory shift has been colloquially referred to as the “Bear Stearns exemption” by some observers, like Barry Ritholtz . This “exemption” ultimately encompassed not only Bear Stearns but also the four other major investment banksâLehman Brothers, Merrill Lynch, Morgan Stanley, and Goldman Sachsâas well as two commercial bank firms, Citigroup and JP Morgan Chase. It has been suggested that Henry Paulson , then CEO of Goldman Sachs , played a leading role in advocating for this rule change.
The SEC’s expectation was that this amendment would lead to a substantial increase in the computed net capital for these broker-dealers. This, in turn, would theoretically allow the parent holding companies to redeploy this “excess” net capital into other business ventures. To mitigate potential risks associated with this increased capital mobility, the SEC instituted a new minimum net capital requirement of $500 million (and $1 billion in “tentative net capital”) for these specialized brokers. More critically, they were required to provide the SEC with an “early warning” notice if their “tentative net capital” dropped below the $5 billion threshold. Prior to this amendment, the minimum net capital requirement for such large firms was a mere $250,000, with an early warning trigger of $300,000. However, the effective minimums for these large firms were significantly higher due to the requirement to maintain net capital equal to 2% of aggregate customer debits, with an early warning at 5% of aggregate debit balances. The 2004 amendment also allowed these CSE Brokers to include “assets for which there is no ready market” in their “tentative net capital” calculations, provided the SEC approved their use of mathematical models to determine the haircuts for such positions.
Issues Addressed by the Rule Change
The 2004 amendment to the net capital rule was a response to two primary concerns. First, the European Union (“EU”) had adopted a Financial Conglomerate Directive in 2002, set to take effect on January 1, 2005. This directive mandated supplemental supervision for unregulated financial holding companies (involved in banking, insurance, or securities) that controlled regulated entities, such as broker-dealers. To avoid this supplemental supervision, non-EU holding companies could seek an exemption if their home country’s regulatory framework was deemed “equivalent” by an EU member state.
The second issue revolved around the applicability of international capital standards, specifically those established by the Basel Committee on Banking Supervision (the “Basel Standards”), to U.S. broker-dealers. The SEC had previously issued a “concept release” in 1997 exploring the potential application of these standards to the net capital rule.
In the United States, a comprehensive system of consolidated supervision for investment bank holding companies did not exist. Instead, the SEC supervised their regulated broker-dealer subsidiaries and other regulated entities like investment advisors. The Gramm-Leach-Bliley Act , which had dismantled key provisions of the GlassâSteagall Act that separated commercial and investment banking, introduced an optional framework for investment bank firms to register with the SEC as “Supervised Investment Bank Holding Companies” (SIBHCs). In contrast, commercial bank holding companies had long been subject to consolidated supervision by the Federal Reserve.
To address the impending EU deadline for consolidated supervision and to align with international capital standards, the SEC issued two proposals in 2003, which were finalized in 2004. The first, the “SIBHC Program,” established rules for companies owning broker-dealers but not banks to register as investment bank holding companies. The second, the “CSE Program,” introduced a new alternative net capital computation method for qualifying broker-dealers (“CSE Brokers”) whose holding companies (“CSE Holding Companies”) elected to become “Consolidated Supervised Entities.” The SEC estimated that establishing a separate European sub-holding company to comply with the EU directive would cost each CSE Holding Company approximately $8 million annually if equivalent consolidated supervision were not established domestically. Both programs aimed to monitor various risksâmarket, credit, liquidity, and operationalâwithin these investment bank holding companies. The CSE Program uniquely allowed CSE Brokers to compute their net capital based on the Basel Standards.
Delayed Use of Rule Change by CSE Brokers
Ultimately, five major investment bank holding companiesâThe Bear Stearns Companies Inc., The Goldman Sachs Group, Inc., Lehman Brothers Holdings Inc., Merrill Lynch & Co. Inc., and Morgan Stanleyâentered the CSE Program. While these firms were not bank holding companies subject to Federal Reserve supervision, they were ineligible for the SIBHC Program due to their ownership of certain types of banks. Two commercial bank holding companies, Citigroup Inc. and JP Morgan Chase & Co., also joined the CSE Program, with their consolidated supervision remaining under the Federal Reserve.
The broker-dealer subsidiary of Merrill Lynch was the first to adopt the new net capital computation method, effective January 1, 2005. A Goldman Sachs broker-dealer subsidiary followed suit between March 23 and May 27, 2005. The broker-dealer subsidiaries of Bear Stearns, Lehman Brothers, and Morgan Stanley all transitioned to the new method on December 1, 2005, marking the beginning of their respective 2006 fiscal years.
Possible Effects of the Rule Change
By permitting CSE Brokers to calculate their net capital using Basel Standards, the SEC anticipated a significant reduction, estimated at roughly 40%, in the “haircuts” applied to their assets. This was before accounting for the newly introduced $5 billion “tentative net capital” early warning requirement. However, the SEC also acknowledged that it was uncertain whether this would lead to a decrease in actual capital levels, as broker-dealers typically maintained net capital well above the required minimums. This practice was partly driven by the requirement under the Alternative Method for broker-dealers to report when their net capital fell below 5% of “aggregate customer debit balances,” at which point they were prohibited from distributing excess capital to their owners. A 1998 GAO report had already noted that excess net capital in large broker-dealers significantly surpassed even this “early warning” threshold, largely due to the demands of counterparties for engaging in business. This reliance on counterparty demands for excess capital had been a long-standing practice. A 1987 study by the Federal Reserve Bank of New York found that, at the close of 1986, sixteen diversified broker-dealers reported average net capital levels that were 7.3 times greater than their required net capital. The same study observed that “Market pressures, rather than regulations, determine how much excess net capital securities firms need to compete.” Nevertheless, to safeguard against substantial reductions in CSE Broker net capital, the SEC implemented the additional “early warning” requirement mandating notification if “tentative net capital” fell below $5 billion.
Consequently, the 2004 amendment to the Alternative Method introduced the possibility that increased net capital computations, based on the same assets (and potentially including additional “less liquid” securities), could weaken customer protections during a CSE Broker’s liquidation. It also raised the concern that capital could be withdrawn from CSE Brokers and channeled into the non-broker/dealer operations of their CSE Holding Companies. Crucially, the amendment did not alter the fundamental test against which net capital was measured, a test that had never directly limited the overall leverage of either a broker-dealer or its parent holding company.
The Collapses of Bear Stearns and Lehman Brothers
Bear Stearns Bear Stearns was the first CSE Holding Company to face collapse. It was ultimately “saved” through a brokered merger with JP Morgan Chase & Co. (“JP Morgan”), announced in March 2008. The Federal Reserve Bank of New York (“FRBNY”) facilitated this merger by providing a $29 billion loan to a special purpose entity (Maiden Lane LLC) which acquired various Bear Stearns assets valued at approximately $30 billion as of March 14, 2008. The broker-dealer operations of Bear Stearns were subsequently absorbed by JP Morgan. Reports indicate that the net capital position of the Bear Stearns CSE Broker (Bear Stearns & Co. Inc.) was adequate, and it appears it would have remained so even under the pre-2004 “haircuts.”
Lehman Brothers Lehman Brothers Holdings Inc. filed for bankruptcy on September 15, 2008. Its CSE Broker, Lehman Brothers Inc. (“LBI”), was not part of the bankruptcy filing and continued to operate until its customer accounts and other assets were acquired by Barclays Capital Inc. As a condition of this acquisition, the Securities Investor Protection Corporation (“SIPC”) initiated a liquidation proceeding for LBI on September 19, 2008, to facilitate the transfer of customer accounts and resolve outstanding claims. While there was controversy surrounding two overnight loans to LBI guaranteed by the FRBNY, these loans appear to have funded intraday or overnight exposures related to customer transaction settlements. The collateral supporting these loans, primarily customer settlement payments, reportedly repaid the loans on the same day or the following day. Reports suggest that the “troubled assets” held by Lehman were primarily commercial real estate assets.
Capital Withdrawals from CSE Brokers and Leverage at CSE Holding Companies
Although the CSE Brokers of Bear Stearns and Lehman Brothers may have maintained solvency and liquidity after the 2004 net capital rule change, it has been argued that the amendment facilitated a significant expansion of the non-broker/dealer operations of Bear, Lehman, and the other CSE Holding Companies. This occurred because these entities were allegedly able to extract excess net capital from their broker-dealers and deploy it to acquire substantial positions in “risky assets.”
Limited Evidence of Capital Withdrawals from CSE Brokers Publicly available studies directly comparing capital levels at CSE Brokers before and after the adoption of the 2004 rule change are scarce. As previously noted, the SEC has stated that “tentative net capital” levels at these firms remained stable, or in some instances, increased following the rule change. However, CSE Brokers were permitted to include certain “less liquid” securities in their net capital computations that had been excluded prior to 2004.
Lehman Brothers Holdings Inc. was the only CSE Holding Company that provided separate financial information for its CSE Broker (LBI) through consolidating financial statements in its Form 10-K filings, primarily because LBI had publicly issued subordinated debt. These filings indicate an increase in LBI’s reported shareholders’ equity after it became a CSE Broker. Goldman Sachs maintains an archive of its CSE Broker’s (Goldman Sachs & Co.) periodic Consolidated Statement of Financial Condition reports on its website. These reports show an increase in total partners’ capital from $4.211 billion at the end of fiscal year 2004 (before becoming a CSE Broker) to $6.248 billion by the end of fiscal year 2007.
Merrill Lynch reported in its 2005 Form 10-K that it had withdrawn $2.5 billion in capital from its CSE Broker during that year. The same filing noted that the 2004 rule change was intended to “reduce regulatory capital costs” and that its CSE Broker subsidiary anticipated further reductions in its excess net capital. Merrill Lynch also reported consolidated year-end shareholders’ equity of $31.4 billion in 2004, increasing to $35.6 billion in 2005.
The remaining three CSE Brokers that reported net capital levelsâBear Stearns, Lehman Brothers, and Morgan Stanleyâall showed significant increases in net capital within their CSE Holding Companies’ Form 10-Q Reports for the initial reporting period under the new computation method (the first fiscal quarter of 2006). This would be consistent with the reduced haircuts permitted by the new method. However, subsequent decreases were also reported, potentially indicating capital withdrawals. It is noteworthy that these firms had also experienced fluctuations in net capital levels, including significant decreases, in periods preceding their adoption of the new method. None of these three firms, nor Merrill Lynch after 2005, reported capital withdrawals from their CSE Brokers in their Form 10-Q or 10-K filings.
CSE Holding Company Leverage Increased Post-2004, But Was Higher in the 1990s
Studies examining Form 10-K and Form 10-Q filings by CSE Holding Companies reveal an increase in their overall reported year-end leverage between 2003 and 2007, and a rise in their reported fiscal quarter leverage from August 2006 through February 2008. However, a review of fiscal year-end balance sheet data from 1993 through 2002 for these same firms indicates that four of the five companies reported higher leverage ratios in one or more years prior to 2000 than they did in any year between 2004 and 2007. Morgan Stanley was the sole exception, reporting higher leverage in 2007 than in any preceding year since 1993. Specifically, Lehman Brothers reported a debt-to-equity ratio of 38.2 to 1 in 1993, Merrill Lynch 34.2 to 1 in 1997, and Bear Stearns and Goldman Sachs 35 to 1 and 31.6 to 1, respectively, in 1998. These figures all predate the 2004 rule change. Even Morgan Stanley, which reported lower leverage after its 1997 merger with Dean Witter Reynolds , had reported a debt-to-equity leverage of 29.1 to 1 at the end of fiscal year 1996.
The higher leverage reported by these four firms during the 1990s was not an isolated anomaly. Form 10-Q filings from three of those firms show an even more pronounced difference between fiscal quarter-end debt-to-equity ratios in the 1990s compared to the post-2004 period. Each of the eight 10-Qs filed by Bear Stearns between its first fiscal quarter of 1997 and the second quarter of 1999 shows a debt-to-equity ratio higher than the highest ratio (32.5 to 1) reported by Bear Stearns in any Form 10-Q filed after 2004. Lehman Brothers reported a debt-to-equity ratio of 54.2 to 1 in its earliest available Form 10-Q (first fiscal quarter of 1994). All nine of Lehman’s 10-Qs filed between 1997 and 1999 show higher debt-to-equity ratios than any of its post-2004 10-Qs. Merrill Lynch’s highest reported debt-to-equity ratio in a post-2004 Form 10-Q is 27.5 to 1; however, all seven of its 10-Qs filed from the first fiscal quarter of 1997 through the first quarter of 1999 show a higher ratio. Goldman Sachs, having become a public company only in 1999, did not file Form 10-Q reports during this earlier period. These pronounced quarterly leverage spikes in the 1990s, relative to the post-2004 period, are visually represented in a 2008 staff report from the Federal Reserve Bank of New York.
The fact that the investment banks that later became CSE Holding Companies exhibited leverage levels significantly exceeding 15 to 1 in the 1990s was noted even before 2008. The President’s Working Group on Financial Markets, in its April 1999 report on hedge funds and the collapse of Long-Term Capital Management (LTCM), pointed out that these five largest investment banks averaged 27-to-1 leverage at the end of 1998, even after absorbing the LTCM portfolio. Popular accounts, such as Frank Partnoy’s 2004 book Infectious Greed, also highlighted this finding. In 1999, the GAO provided leverage statistics for the four largest investment banks in its LTCM report, showing Goldman Sachs and Merrill Lynch both exceeding the LTCM leverage ratio (at 34-to-1 and 30-to-1, respectively), Lehman equaling LTCM’s 28-to-1 leverage, and Morgan Stanley trailing at 22-to-1. A 2009 GAO report reiterated these findings and charted leverage from 1998 to 2007 for four of the CSE Holding Companies, demonstrating that three of these firms had higher leverage at the end of fiscal year 1998 than they did at the end of fiscal year 2006, prior to the onset of the crisis.
Net Capital Rule in Financial Crisis Explanations Since 2008
Since 2008, the assertion that the 2004 rule change enabled a dramatic increase in leverage at CSE Holding Companies has become deeply ingrained in the extensive literature analyzing the 2008 financial crisis and in congressional testimony. In their book This Time is Different, Professors Carmen M. Reinhart and Kenneth S. Rogoff characterized the 2004 rule change as the SEC’s decision “to allow investment banks to triple their leverage ratios.” Robert Pozen, in Too Big to Save, described the “fundamental problem” for CSE Holding Companies as stemming primarily from the SEC’s decision to permit more than a doubling of their leverage ratio, compounded by the SEC’s ineffective implementation of consolidated supervision. Jane DâArista testified before the House Financial Services Committee in October 2009, stating that the 2004 rule change represented a “relaxation of the leverage limit for investment banks from $12 to $30 per $1 of capital.” The issue even entered military theory, with Gautam Mukunda and Major General William J. Troy writing in the US Army War College Quarterly that the 2004 rule change allowed investment banks “to increase their leverage to as high as 40 to 1” when previously “the SEC had limited investment banks to a leverage ratio of 12 to 1.”
In reviewing similar statements made by former SEC chief economist Susan Woodward and scholars such as Professors Alan Blinder, John Coffee, and Joseph Stiglitz, Professor Andrew Lo and Mark Mueller noted that 1999 and 2009 GAO reports documented that investment bank leverage ratios had actually been higher before 2000 than after the 2004 rule change. They further observed that “these leverage numbers were in the public domain and easily available through company annual reports and quarterly SEC filings.” Lo and Mueller cite this as an example of how “sophisticated and informed individuals can be so easily misled on a relatively simple and empirically verifiable issue,” underscoring the need for careful analysis, especially during times of crisis when urgency might lead to premature and inaccurate conclusions. They posit that mental errors, contributing both to the crisis and to current misunderstandings of it, arise from “mental models of reality that are not complete or entirely accurate” but which lead individuals to accept information that “confirms our preconceptions” without critical scrutiny.
Lo and Mueller point out that the New York Times has not issued a correction to its 2008 article. At least two scholars who initially claimed pre-2004 leverage was limited to 12-to-1 have since implicitly corrected themselves. Niall Ferguson, in a JulyâAugust 2009 Harvard Business Review article, acknowledged information indicating that from 1993 to 2002, Bear Stearns, Goldman Sachs, Merrill Lynch, and Morgan Stanley reported average leverage ratios of 26-to-1 (with Bear Stearns averaging 32-to-1 during those years). In his 2010 book, Freefall, Joseph Stiglitz noted that by 2002, leverage at large investment banks had reached as high as 29-to-1. He also directed readers to the 2009 Sirri Speech and the 2008 New York Sun article for alternative perspectives on the role of the 2004 rule change in investment bank difficulties.
Net Capital Rule in the Financial Crisis Inquiry Commission Report
The Financial Crisis Inquiry Report (“FCIC Report”), released in January 2011, presented the majority’s perspective on the 2004 net capital rule change’s impact on leverage. It stated: “Leverage at the investment banks increased from 2004 to 2007, growth that some critics have blamed on the SEC’s change in the net capital rules…In fact, leverage had been higher at the five investment banks in the late 1990s, then dropped before increasing over the life of the CSE programâa history that suggests that the program was not solely responsible for the changes.”
While the FCIC Report identified leverage, particularly at investment banks, as a significant factor in the financial crisis, it did not pinpoint the 2004 net capital rule change as the sole cause or enabler of that leverage. The report also noted an 18-month delay before all CSE Holding Companies qualified for the CSE Program. While the FCIC did not draw specific conclusions from this delay beyond a general critique of the SEC’s execution of the program, the delay itself effectively removes the 2004 rule change from being a factor in any leverage increases observed in 2004. Furthermore, since all CSE Holding Companies only joined the program at the commencement of their 2006 fiscal years, the delay precludes the 2004 rule change from influencing leverage increases in 2005 reported by Bear Stearns, Lehman Brothers, or Morgan Stanley.
The FCIC Report was highly critical of the SEC’s implementation of the CSE Program, rejecting former SEC Chairman Christopher Cox ’s assertion that the program was “fatally flawed” due to its voluntary nature. The report concluded that the SEC failed to exercise its powers under the CSE Program to restrict “risky activities” of CSE Holding Companies and to “require them to hold adequate capital and liquidity for their activities”âpowers that the SEC did not possess prior to 2004 and only gained through the CSE Program.
Despite these findings, persistent confusion regarding the application of the net capital rule to CSE Holding Companies surfaced within the FCIC Report. After correctly explaining that the net capital rule applied to broker-dealers, not their parent holding companies, the report misquoted the 2009 Sirri Speech, stating that “investment bank net capital was stable, or in some cases increased, after the rule change.” A more serious indication of ongoing misunderstanding occurred on the day the FCIC released its report (January 27, 2011), when economist Robert Hall addressed an MIT symposium. While discussing the background to the financial crisis, he stated: “I think that the one most important failure of regulation is easy to identify. In 2004 the SEC removed its capital requirements on investment banks… So the reason that Lehman was able to do what it did, which proved so destructive, was that it had no limits on the amount of leverage it could adopt.”