The Lehman Brothers Collapse

Through the lens of the Gertler-Kiyotaki model

The collapse of Lehman Brothers on September 15, 2008, represents one of the most significant financial events of the 21st century and provides a compelling real-world case to examine through the theoretical framework of the Gertler-Kiyotaki model. This case study analyses how the financial frictions outlined in the model manifested during the 2008 financial crisis, focusing specifically on Lehman Brothers as the epicenter of the crisis.

1. Background: Lehman Brothers’ financial structure

Founded in 1850, Lehman Brothers had grown into the fourth-largest investment bank in the United States by 2008. Its financial structure immediately prior to collapse reveals extreme vulnerability when examined through the lens of the Gertler-Kiyotaki model:

Key Financial Indicators (Q2 2008) Value Model Parameter
Total Assets $639 billion s (securities)
Net Worth (Book Value) $26 billion N (net worth)
Market Capitalization (Aug 2008) ~$8 billion N (market-valued)
Total Liabilities $613 billion d (deposits/debt)
Leverage Ratio (Book) 24.6:1 s/N
Short-term Debt $197 billion Portion of d requiring frequent rollover

Lehman’s leverage ratio of 24.6:1 significantly exceeded the industry average of 18:11. This extreme leverage made the firm particularly vulnerable to asset value declines.

Lehman’s asset portfolio showed significant concentration in real estate and mortgage-related products:

Asset Category Value (Q2 2008) % of Total Assets
Mortgage-backed Securities $85 billion 13.3%
Commercial Real Estate $55 billion 8.6%
Residential Real Estate $25 billion 3.9%
Real Estate Total $165 billion 25.8%

This concentrated exposure to real estate made Lehman particularly vulnerable to correlated asset price shocks during the housing market collapse.

2. The Gertler-Kiyotaki model applied to Lehman

2.1 The incentive compatibility constraint

The central insight of the Gertler-Kiyotaki model is the incentive compatibility constraint that limits a banker’s ability to issue deposits:

(1-\theta)(N+d)R^k \geq dR^d

This constraint ensures that bankers don’t have an incentive to default. Rearranging to solve for the maximum sustainable debt:

d \leq \frac{(1-\theta)NR^k}{R^d-(1-\theta)R^k}

For this constraint to yield economically meaningful results (positive debt capacity), we must have R^d > (1-\theta)R^k. This inequality represents the condition where the cost of debt exceeds the banker’s effective return after accounting for moral hazard.

To apply this to Lehman Brothers, we need to estimate the key parameters accurately.

2.2 Estimating key model parameters for Lehman

Parameter θ (Proportion of returns bankers can divert)

The parameter θ represents the fraction of gross returns that bankers can potentially divert. For Lehman Brothers, we estimate this parameter based on several observable factors:

  1. Executive compensation relative to returns: Lehman paid its top executives \$85 million in 2007 on net income of \$4.2 billion, representing approximately 2\% of net income2. However, this understates total agency costs.

  2. Risk limit breaches: According to the Valukas report, Lehman exceeded its internal risk limits 44 times between 2007-2008 without consequences, allowing executives to take excessive risks that benefited them through performance bonuses but endangered the firm3.

  3. Repo 105 transactions: Lehman temporarily removed approximately \$50 billion in assets from its balance sheets using accounting techniques to present a healthier leverage ratio—representing about 8\% of its balance sheet4.

  4. Risk-taking incentives: Lehman’s compensation structure rewarded short-term profit generation with limited downside for executives if risks turned out poorly. Executives received approximately 30-35\% of their business unit’s revenue growth as bonuses with minimal clawback provisions5.

  5. Academic research: Financial economics literature suggests agency costs in investment banks typically range from 20-35% of returns, with higher values during boom periods when oversight weakens6.

Based on this evidence, we estimate \theta0.25, representing the effective proportion of returns that could be diverted through various forms of moral hazard.

Parameters R^k (Return on assets) and R^d (Cost of debt)

For R^k, we examine Lehman’s historical returns:

Period Net Return on Assets Gross Return on Assets Notes
2006 0.9% 7.2% Pre-crisis peak
2007 0.7% 6.5% Early crisis
Q1-Q2 2008 0.3% 5.8% Deteriorating conditions

Note: Net return on assets represents reported profit divided by total assets, while gross return represents total revenue before expenses divided by assets. The difference accounts for operating costs, taxes, and other expenses.

For our analysis, we use R^k6.0\% as the expected gross return on Lehman’s assets pre-crisis, which aligns with their average asset portfolio yield in 2007-20087.

For R^d, we use Lehman’s average funding costs:

Period Average Funding Cost Notes
2006 4.3% Pre-crisis funding
2007 4.8% Early crisis
June 2008 5.5% Rising funding pressure
Early Sept 2008 6.5% Severe funding pressure

We use R^d = 4.8\% for the pre-crisis period (2007) and R^d = 6.5\% for the immediate pre-bankruptcy period (Sept 2008)8.

2.3 Lehman’s debt capacity: model analysis

With our revised parameters, let’s calculate Lehman’s sustainable debt level in the pre-crisis period:

  • N = \$26 billion (book value)
  • R^k = 6.0\%
  • R^d = 4.8\%
  • \theta = 0.25

d \leq \frac{(1-0.25) \times 26B \times 0.06}{0.048-(1-0.25) \times 0.06} = \frac{1.17B}{0.048-0.045} = \frac{1.17B}{0.003} \approx \$390B

This calculation suggests Lehman’s sustainable debt level in the pre-crisis period was approximately \$390 billion. This is a significant deviation from their actual debt of \$613 billion.

As the crisis intensified in 2008, the situation deteriorated significantly:

  • N = \$8 billion (market value)
  • R^k = 5.8\% (declining asset returns)
  • R^d = 6.5\% (rising funding costs)
  • \theta = 0.25

d \leq \frac{(1-0.25) \times 8B \times 0.058}{0.065-(1-0.25) \times 0.058} = \frac{0.348B}{0.065-0.0435} = \frac{0.348B}{0.0215} \approx \$16.2B

This dramatic reduction in sustainable debt —from \$390 billion to about \$16.2 billion—illustrates the extreme fragility of Lehman’s position as market conditions deteriorated. The firm would have needed to reduce its assets by approximately 97\% to reach a sustainable level, making failure mathematically inevitable without significant external capital injection.

3. Financial acceleration and Lehman’s death spiral

3.1 The financial acceleration mechanism

The Gertler-Kiyotaki model implicitly captures financial acceleration—the process by which small initial shocks are amplified through the financial system. For Lehman Brothers, this played out as follows:

  1. Initial shock: Housing market declines in 2007-2008 reduced the value of mortgage-backed assets
  2. Net worth erosion: Asset writedowns reduced Lehman’s equity (N↓)
  3. Constraint tightening: Lower equity meant lower sustainable debt levels
  4. Forced deleveraging: To reduce debt, assets had to be sold, often at fire-sale prices
  5. Market-wide price impact: Asset sales depressed prices market-wide
  6. Feedback loop: Lower asset prices further reduced net worth across the financial system

The power of this acceleration depends on leverage and the sensitivity of the incentive constraint. For Lehman, with leverage of nearly 25:1, each 1% decline in asset values would reduce equity by approximately 25%.

3.2 Lehman’s actual experience

Lehman’s quarterly reports reveal this acceleration in action:

Quarter Asset Writedowns Balance Sheet Reduction Equity Decline
Q4 2007 $1.2 billion $7.8 billion $0.7 billion
Q1 2008 $1.8 billion $49 billion $0.7 billion
Q2 2008 $3.7 billion $147 billion $2.8 billion
Q3 2008 (partial) $5.3 billion N/A (bankruptcy) $3.9 billion

These figures, verified against Lehman’s SEC filings9, show the accelerating pace of writedowns and increasingly desperate balance sheet reduction efforts. Despite cutting its balance sheet by \$147 billion in Q2 2008 alone (a 19\% reduction), Lehman could not shrink fast enough to restore market confidence.

By September 2008, Lehman faced a classic “death spiral”:

  1. Counterparties demanded additional collateral for repos and derivatives
  2. Rating agencies threatened downgrades
  3. Funding markets closed to Lehman
  4. Even dramatic asset sales couldn’t reduce leverage fast enough

The firm unsuccessfully sought emergency funding before filing for bankruptcy on September 15, 2008.

4. Interest rate spreads: model predictions vs. market evidence

4.1 Theoretical spread determination

In the Gertler-Kiyotaki model, when the incentive compatibility constraint becomes binding, interest rate spreads (R^k - R^d) emerge. The model predicts that spreads widen when:

  1. Net worth (N) decreases
  2. Moral hazard concerns (\theta) increase
  3. Asset returns (R^k) become more uncertain

4.2 Empirical evidence from 2008

The model’s predictions align remarkably well with observed spreads during the 2008 crisis:

Interest Rate Spread Jan 2007 Sep 2008 (pre-Lehman) Sep 2008 (post-Lehman) Oct 2008 (peak)
TED Spread (3m LIBOR - T-Bill) 0.38% 1.35% 3.35% 4.64%
Commercial Paper - T-Bill 0.15% 1.75% 3.90% 5.25%
Baa Corporate - Treasury 1.80% 3.10% 5.10% 6.50%

Data from the Federal Reserve10 confirms the dramatic widening of spreads across all categories, aligning with the model’s prediction that declining net worth and increasing moral hazard concerns would drive spreads higher.

5. Moral hazard and agency costs: Lehman’s experience

5.1 Executive compensation and risk-taking incentives

Lehman’s compensation structure strongly incentivized risk-taking with limited personal downside:

Year CEO Compensation Top 5 Executives Net Income Exec Comp/Income
2006 $40.6 million $90.9 million $4.0 billion 2.3%
2007 $40.0 million $85.0 million $4.2 billion 2.0%
2008 (est.) $40.0 million $80.0 million $-2.5 billion N/A

Despite deteriorating performance, executive compensation remained largely unchanged until bankruptcy, reflecting poor alignment of incentives11.

5.2 Accounting manipulation: Repo 105

The most direct evidence of moral hazard at Lehman was its use of “Repo 105” transactions to temporarily remove approximately \$50 billion in assets from its balance sheets in 2008. The Valukas report described this as follows:

“Lehman employed off-balance sheet devices, known within Lehman as ‘Repo 105,’ to temporarily remove securities inventory from its balance sheet, usually for a period of seven to ten days, and to create a materially misleading picture of the firm’s financial condition in late 2007 and 2008.”12

The mechanics of Repo 105:

  1. Lehman transferred assets valued at 105\% of the cash received
  2. Accounted for these as “sales” rather than financing
  3. Used the cash to temporarily pay down debt before reporting periods
  4. Repurchased the assets after financial statements were issued

These transactions had no legitimate business purpose other than balance sheet manipulation, directly reflecting the “diversion” concept in the Gertler-Kiyotaki model.

5.3 Risk management failures

The Valukas report documented numerous instances of risk management override that reflect moral hazard:

  1. “Lehman… chose to disregard or overrule the firm’s risk controls on a regular basis”
  2. The firm exceeded its internal risk limits 44 times between 2007-2008 without consequences
  3. Lehman modified its risk metrics when existing measures showed excessive risk
  4. Executives routinely overrode risk manager objections to specific transactions

These behaviors directly parallel the diversion concept in the Gertler-Kiyotaki model, as executives prioritized short-term growth and compensation over long-term firm stability.

6. Policy responses: model insights and actual interventions

6.1 Model-based policy interventions

The Gertler-Kiyotaki model suggests three primary policy interventions when financial frictions become binding:

  1. Equity injections (increasing N directly)
  2. Loans to banks (subsidizing R^d)
  3. Direct loans to firms (bypassing banks)

6.2 Actual policy responses

The policy response to Lehman Brothers itself was essentially non-intervention, with authorities allowing the firm to declare bankruptcy. However, the broader crisis response evolved rapidly after Lehman’s collapse:

6.2.1 Equity injections

The Capital Purchase Program (CPP) under TARP provided $250 billion in equity investments to banks:

Institution Capital Injection As % of Pre-Crisis Equity
Citigroup $45 billion 22.5%
Bank of America $45 billion 25.0%
JPMorgan Chase $25 billion 13.9%
Wells Fargo $25 billion 15.0%
Goldman Sachs $10 billion 13.3%
Morgan Stanley $10 billion 18.2%

6.2.2 Loans to banks

The Federal Reserve implemented numerous lending facilities with peak utilization of over \$1 trillion, including:

  • Term Auction Facility (\$493 billion peak)
  • Primary Dealer Credit Facility (\$156 billion peak)
  • Term Securities Lending Facility (\$236 billion peak)

6.2.3 Direct loans to firms

The Federal Reserve also implemented programs to bypass impaired financial intermediaries:

  • Commercial Paper Funding Facility (\$350 billion peak)
  • Term Asset-Backed Securities Loan Facility (\$70 billion peak)

These interventions directly addressed the mechanisms identified in the Gertler-Kiyotaki model, and proved effective at stabilizing markets—but only after significant economic damage had occurred.

7. Counterfactual analysis: could Lehman have been saved?

The model allows us to estimate what would have been required to save Lehman Brothers:

  1. Equity injection needed: To reach a sustainable leverage ratio, Lehman would have needed approximately \$50-75 billion in new equity

  2. Funding guarantee required: Alternatively, government guarantees on Lehman’s debt could have reduced R^d and relaxed the incentive constraint

  3. Asset purchase program: A program to purchase Lehman’s troubled assets at above-market prices could have supported net worth

The model suggests that saving Lehman was theoretically possible but would have required massive government intervention—substantially larger than what had been provided to Bear Stearns in March 2008.

8. Conclusion: Lessons from Lehman through the Gertler-Kiyotaki lens

The Lehman Brothers collapse provides a striking real-world illustration of the financial frictions described in the Gertler-Kiyotaki model. The extreme leverage, erosion of net worth, binding incentive constraints, and subsequent interest rate spreads all conform to the model’s predictions.

Key lessons include:

  1. Leverage regulation matters: The model identifies excessive leverage as a critical vulnerability, justifying post-crisis capital requirements
  2. Net worth buffers: Maintaining adequate capital buffers is essential for financial stability, particularly for systemically important institutions
  3. Early intervention: The model suggests that earlier policy intervention could have prevented the severe market disruption that followed Lehman’s failure
  4. Moral hazard monitoring: Regulatory oversight should focus on compensation structures and governance that might increase \theta

The Gertler-Kiyotaki model thus provides not just a theoretical framework but a practical lens through which to understand financial crises and design appropriate regulatory responses.

References


  1. Financial Crisis Inquiry Commission. (2011). “The Financial Crisis Inquiry Report: Final Report of the National Commission on the Causes of the Financial and Economic Crisis in the United States.” U.S. Government Printing Office, p. 65.↩︎

  2. Lehman Brothers Holdings Inc. (2008). “Proxy Statement for Annual Meeting of Shareholders.” SEC Schedule 14A, filed March 5, 2008.↩︎

  3. Valukas, A. (2010). “Lehman Brothers Holdings Inc. Chapter 11 Proceedings Examiner’s Report.” United States Bankruptcy Court Southern District of New York, Volume 1, pp. 732-742.↩︎

  4. Valukas, A. (2010). “Lehman Brothers Holdings Inc. Chapter 11 Proceedings Examiner’s Report.” United States Bankruptcy Court Southern District of New York, Volume 1, pp. 732-742.↩︎

  5. U.S. House of Representatives Committee on Oversight and Government Reform. (2008). “Hearing on the Causes and Effects of the Lehman Brothers Bankruptcy.” October 6, 2008.↩︎

  6. Philippon, T., & Reshef, A. (2012). “Wages and Human Capital in the U.S. Finance Industry: 1909–2006.” The Quarterly Journal of Economics, 127(4), 1551-1609.↩︎

  7. Calculated from Lehman Brothers Annual Reports, 2005-2007, averaging asset yields on fixed income, equity, and real estate investments.↩︎

  8. Ball, L. (2018). “The Fed and Lehman Brothers: Setting the Record Straight on a Financial Disaster.” Cambridge University Press, pp. 87-92.↩︎

  9. Lehman Brothers Holdings Inc. (2008). “Quarterly Report for the period ended May 31, 2008.” SEC Form 10-Q, filed July 10, 2008.↩︎

  10. Federal Reserve Bank of St. Louis, FRED Economic Data, retrieved from https://fred.stlouisfed.org/↩︎

  11. U.S. House of Representatives Committee on Oversight and Government Reform. (2008). “Hearing on the Causes and Effects of the Lehman Brothers Bankruptcy.” October 6, 2008.↩︎

  12. Valukas, A. (2010). “Lehman Brothers Holdings Inc. Chapter 11 Proceedings Examiner’s Report.” United States Bankruptcy Court Southern District of New York, Volume 1, pp. 732-742.↩︎