Financial Accelerator

The Financial Accelerator—often referred to as the Bernanke–Gertler–Gilchrist (BGG) Model—is the foundational dynamic general-equilibrium framework in modern macroeconomics, monetary policy, and business and economics theory that explains how credit-market frictions and borrower balance sheets endogenously amplify and propagate real and monetary shocks across the business cycle. Originated by economists Ben S. Bernanke and Mark Gertler in their seminal 1989 American Economic Review paper "Agency Costs, Net Worth, and Business Fluctuations" and embedded into a quantitative Dynamic Stochastic General Equilibrium (DSGE) New Keynesian framework by Ben S. Bernanke, Mark Gertler, and Simon Gilchrist in their landmark 1999 Handbook of Macroeconomics chapter "The Financial Accelerator in a Quantitative Business Cycle Framework", the theory overturned the classical Modigliani–Miller theorem neutrality assumption in business-cycle modeling. By embedding the microeconomic Townsend costly state verification (CSV) debt contract into a macroeconomic model with capital-adjustment costs (Tobin's q) and nominal price rigidity (Rotemberg pricing or Calvo staggered pricing), Bernanke, Gertler, and Gilchrist proved that the External Finance Premium s_t—the wedge between the expected cost of external funds and the opportunity cost of internal funds—is an increasing function of the borrower's leverage ratio Q_t K_t+1 / N_t+1 (s′(·) > 0). Consequently, an adverse productivity or monetary tightening shock that lowers asset prices Q_t and erodes entrepreneurial net worth N_t+1 drives up leverage, widens the external finance premium s_t, contracts capital investment I_t, and triggers a further decline in asset prices Q_t—generating an endogenous credit-market amplification loop and a flight to quality away from small, highly leveraged firms.

The Puzzle of Small Shocks and Large Macroeconomic Fluctuations

A persistent challenge in both Real Business Cycle (RBC) models (Kydland and Prescott, 1982) and frictionless New Keynesian models is explaining how relatively modest, transitory shocks to aggregate productivity or short-term nominal interest rates generate large, persistent, and asymmetric contractions in business fixed investment, employment, and output. In a frictionless neoclassical or IS–LM benchmark where the Modigliani–Miller theorem (1958) holds, corporate balance sheets, internal cash flow, and debt-to-equity ratios are macroeconomic veils: a firm's investment demand depends on the physical marginal product of capital relative to the risk-free real user cost of capital, regardless of whether the firm finances capital purchases out of retained earnings or external debt.

Drawing on Irving Fisher's (1933) debt-deflation hypothesis and Ben Bernanke's (1983) historical analysis of the Great Depression ("Nonmonetary Effects of the Financial Crisis in the Propagation of the Great Depression"), Bernanke and Gertler (1989, 1995) and Bernanke, Gertler, and Gilchrist (1999) argued that information asymmetry and enforcement costs in credit markets create an external finance premium whose countercyclical movement acts as a financial accelerator.

Core Definitions: Borrower Net Worth N_t+1 and the External Finance Premium s_t

Two macroeconomic variables lie at the heart of the BGG financial accelerator:

1. Entrepreneurial Net Worth (N_t+1): The liquid internal funds (cash flow from operations plus the market value of existing physical assets minus outstanding debt obligations) that borrowers can commit as inside equity or collateral toward purchasing capital Q_t K_t+1.
2. The External Finance Premium (s_t): The inevitable agency-cost wedge between the required gross expected return on capital demanded by external lenders, E_t[R_t+1^k], and the gross risk-free opportunity cost of internal funds, R_t+1:

Because higher borrower net worth N_t+1 aligns borrower and lender incentives and reduces expected bankruptcy and auditing deadweight losses, the fundamental theorem of the financial accelerator states that the external finance premium s_t varies inversely with borrower net worth N_t+1 (and positively with the leverage ratio Q_t K_t+1 / N_t+1).

Entrepreneurial Balance Sheets, Capital Purchases, and Idiosyncratic Shocks ω

In the Bernanke–Gertler–Gilchrist (1999) model, the corporate production sector is populated by a continuum of risk-neutral Entrepreneurs indexed by j ∈ [0, 1]. At the end of period t, entrepreneur j has accumulated internal net worth N_t+1^j and purchases physical capital K_t+1^j at the market asset price Q_t (determined by capital-producers facing convex adjustment costs, as in Tobin's q theory) for use in period t + 1 production.

To finance the total capital outlay Q_t K_t+1^j, the entrepreneur commits all internal net worth N_t+1^j and borrows the remaining external debt B_t+1^j from a competitive financial intermediary (which raises deposits from households at the gross riskless real interest rate R_t+1):

In period t + 1, the aggregate economy realizes a gross return to capital R_t+1^k (reflecting marginal rental income plus depreciated resale value Q_t+1). Simultaneously, each individual entrepreneur j draws an i.i.d. idiosyncratic productivity multiplier ω^j ∈ (0, ∞) from a continuous lognormal cumulative distribution function F(ω) with density f(ω) and unit mean E[ω] = 1, so entrepreneur j's realized ex-post gross project return is ω^j R_t+1^k Q_t K_t+1^j.

The Optimal CSV Debt Contract and Default Cutoff ω̄_t+1

Following Townsend costly state verification (1979) and Gale and Hellwig (1985), entrepreneur j observes their idiosyncratic realization ω^j costlessly, whereas the financial intermediary can observe ω^j only by initiating bankruptcy verification at a proportional auditing cost μ ∈ (0, 1) of the firm's realized gross assets: μ ω^j R_t+1^k Q_t K_t+1^j (calibrated by BGG to μ = 0.12).

The unique optimal incentive-compatible contract is a standard debt contract specified by a default cutoff threshold ω̄_t+1^j (corresponding to a gross non-default loan rate Z_t+1^j such that ω̄_t+1^j R_t+1^k Q_t K_t+1^j = Z_t+1^j B_t+1^j):

• Solvent States (ω^j ≥ ω̄_t+1^j): The entrepreneur pays the promised debt obligation ω̄_t+1^j R_t+1^k Q_t K_t+1^j without auditing and retains the equity residual (ω^j − ω̄_t+1^j) R_t+1^k Q_t K_t+1^j.
• Default / Bankruptcy States (ω^j < ω̄_t+1^j): The entrepreneur defaults, the lender audits the firm at proportional cost μ, and the lender collects the net liquidation recovery (1 − μ) ω^j R_t+1^k Q_t K_t+1^j, leaving the entrepreneur with 0.

Constant-Returns Aggregation and the Leverage Function s(Q_t K_t+1 / N_t+1)

Define two functions of the default cutoff ω̄ spanning 100% of gross project return:

1. Expected Gross Entrepreneurial Share (1 − Γ(ω̄)): where Γ(ω̄) ≡ ∫₀^ω̄ ω f(ω) dω + ω̄ ∫_ω̄^∞ f(ω) dω is the gross share going to the lender before auditing costs (with Γ′(ω̄) = 1 − F(ω̄) > 0 and Γ″(ω̄) = −f(ω̄) < 0).
2. Expected Proportional Auditing Loss μ G(ω̄): where G(ω̄) ≡ ∫₀^ω̄ ω f(ω) dω is the expected value of ω conditional on default multiplied by F(ω̄), so the lender's expected net share of total project return is Γ(ω̄) − μ G(ω̄).

Because the financial intermediary diversifies idiosyncratic risk ω^j across the continuum of entrepreneurs, the lender's break-even participation constraint in each aggregate state R_t+1^k is:

Dividing both sides by R_t+1 N_t+1^j and solving the entrepreneur's optimal contracting choice over (K_t+1^j, ω̄_t+1^j) yields BGG's celebrated Scale-Invariant Aggregation Result: every entrepreneur j chooses the exact same leverage ratio k_t ≡ Q_t K_t+1^j / N_t+1^j regardless of their individual net worth N_t+1^j, so the aggregate capital-to-net-worth ratio satisfies:

Bernanke–Gertler–Gilchrist (1999) External Finance Premium Equation:
: s_t ≡ E_t [ R_t+1^k ] / R_t+1 = s( Q_t K_t+1 / N_t+1 ),      with    s(1) = 1   and   s′(·) > 0

Log-linearizing this relation around the non-stochastic steady state yields the core structural equation of the financial accelerator (where lowercase variables denote log-deviations from steady state and ν > 0 is the elasticity of the external finance premium with respect to leverage):

Evolution of Aggregate Net Worth N_t+1 and Asset-Price Feedback (Q_t)

To prevent entrepreneurs from accumulating infinite internal wealth over time and outgrowing the credit constraint, BGG assume that each entrepreneur survives into the next period with probability γ_e ∈ (0, 1) (calibrated to γ_e = 0.9728 per quarter, implying an average horizon of ~36 quarters) and exits with probability 1 − γ_e. Aggregate entrepreneurial net worth N_t+1 evolves according to:

where W_t^e is small entrepreneurial wage income and V_t is aggregate entrepreneurial equity realized in period t:

Notice why V_t (and therefore N_t+1) is a leveraged multiple of the realized return to capital R_t^k (and asset price Q_t): because debt obligations R_t B_t are predetermined in nominal/real units, a 1% unexpected drop in the market price of capital Q_t reduces R_t^k by roughly 1%, which slashes leveraged entrepreneurial net worth N_t+1 by ( Q K / N ) × 1% (e.g., by 2% when the capital-to-net-worth ratio is 2:1)!

Step-by-Step Transmission of a Monetary Policy Tightening or Adverse Shock

When the central bank raises the nominal policy rate (a monetary policy tightening shock) or an adverse productivity shock hits the economy, the BGG Financial Accelerator Loop operates through five mutually reinforcing stages:

1. Initial Impact on Asset Demand and Price Q_t: Higher real interest rates R_t+1 and lower cash flows depress investment demand I_t, causing the price of installed capital Q_t to fall along the upward-sloping supply curve of capital goods (Tobin's q).
2. Leveraged Erosion of Borrower Net Worth N_t+1: The drop in asset price Q_t and realized capital return R_t^k—combined with Fisher debt-deflation if debt contracts are fixed in nominal terms and inflation falls—wipes out a large fraction of entrepreneurial net worth N_t+1.
3. Jump in Leverage and the External Finance Premium s_t: Because net worth N_t+1 falls much faster than capital Q_t K_t+1, the aggregate leverage ratio Q_t K_t+1 / N_t+1 spikes upward, raising the default probability F(ω̄_t+1) and driving up the external finance premium s_t = E_t[R_t+1^k] / R_t+1.
4. Contraction in Capital Investment I_t: Facing a higher required return on capital E_t[R_t+1^k] = s_t R_t+1, entrepreneurs cut back sharply on capital purchases K_t+1 and investment I_t.
5. Second-Round Multiplier Feedback onto Q_t: The sharp drop in investment demand I_t pushes the current asset price Q_t down even further, which further erodes current net worth N_t+1 in a simultaneous fixed-point amplification loop!

In BGG's calibrated New Keynesian simulations, the financial accelerator increases the peak output response to a monetary policy shock by roughly 50% and nearly doubles the persistence (half-life) of the recession compared to an identical New Keynesian model with ν = 0 (no financial accelerator).

Why Small, Bank-Dependent Firms Bear the Brunt of Monetary Contractions

Beyond aggregate business-cycle amplification, the BGG financial accelerator generates a sharp cross-sectional prediction known as the Flight to Quality (Bernanke, Gertler, and Gilchrist, 1996, Review of Economics and Statistics). Consider an economy with two groups of firms:

• High-Net-Worth / Large Public Corporations: Possess strong balance sheets, audited financial statements, investment-grade commercial paper access, and low leverage, placing them on the flat region of the external finance premium curve (s_t ≈ 1).
• Low-Net-Worth / Small Bank-Dependent Firms: Face severe informational opacity, high proportional auditing costs μ, and high leverage Q_t K_t+1 / N_t+1, placing them on the steep, convex region of s(·).

When a monetary tightening or recession hits and erodes borrower net worth across the economy, the external finance premium s_t rises disproportionately for small, low-net-worth borrowers, prompting lenders to reallocate credit toward safe, high-grade borrowers (Flight to Quality). Using quarterly panel data from the U.S. Census Bureau's Quarterly Financial Report (QFR) for manufacturing corporations (following Gertler and Gilchrist, 1994), Bernanke, Gertler, and Gilchrist demonstrated that following a Federal Reserve monetary tightening, small manufacturing firms experience a substantially steeper contraction in short-term bank borrowing, inventory accumulation, and sales than large manufacturing firms, accounting for a disproportionate share of the aggregate manufacturing downturn.

Post-2008 Extensions: Risk Shocks and Financial Intermediary Net Worth

Following the 2007–2009 Global Financial Crisis, Lawrence Christiano, Roberto Motto, and Massimo Rostagno (2014, American Economic Review) embedded the BGG financial accelerator into an estimated medium-scale DSGE model of the U.S. and Euro-Area economies and showed that Risk Shocks—time-varying cross-sectional dispersion σ_ω,t in the idiosyncratic productivity distribution F_t(ω)—are the single most important driver of business-cycle fluctuations in investment and credit spreads. Simultaneously, Mark Gertler and Peter Karadi (2011) extended the BGG balance-sheet mechanism from non-financial corporate borrowers to financial intermediaries (banks) themselves, providing the workhorse analytical model for central-bank quantitative easing and emergency credit facilities. For his foundational research on the macroeconomic role of credit frictions and banking crises, Ben S. Bernanke was awarded the 2022 Nobel Memorial Prize in Economic Sciences alongside Douglas W. Diamond and Philip H. Dybvig.

How does the Bernanke–Gertler–Gilchrist financial accelerator differ from Kiyotaki–Moore (1997) credit cycles?

While both models show how falling asset prices Q_t interact with borrower balance sheets to amplify business cycles, they rely on different microeconomic contracting foundations. The BGG Financial Accelerator embeds Townsend costly state verification, where default occurs in equilibrium with positive probability F(ω̄_t+1) and leverage drives up the price of external credit (the external finance premium s_t). In contrast, Nobuhiro Kiyotaki and John Moore's (1997) Credit Cycles model embeds Hart–Moore incomplete contracts, where lenders cannot verify cash flows at all and can only seize durable physical collateral (land K_t+1), imposing a hard quantity ceiling on borrowing (R_t+1 B_t ≤ E_t[Q_t+1] K_t+1) with zero equilibrium default and a risk-free loan interest rate.

Why can't entrepreneurs in the BGG model simply issue outside equity to avoid costly debt bankruptcy?

Under the Townsend costly state verification friction at the core of the BGG model, a firm's realized idiosyncratic return ω^j R_t+1^k Q_t K_t+1^j is privately observed by the entrepreneur. If the entrepreneur issued outside equity promising to pay outside shareholders a proportional fraction of realized profits, the entrepreneur would have an incentive to underreport ω^j unless outside shareholders audited the firm's books at deadweight cost μ in every state of nature. The standard debt contract is uniquely optimal precisely because it restricts costly auditing to the lower tail of default states (ω^j < ω̄_t+1).

How does nominal debt (Fisher debt deflation) interact with the financial accelerator in the BGG model?

In the quantitative BGG (1999) model, debt contracts are denominated in nominal currency units rather than indexed to the aggregate price level P_t. When an adverse aggregate demand or monetary tightening shock causes unexpected disinflation or deflation (a drop in P_t below expected E_t−1[P_t]), the real value of entrepreneurs' predetermined debt liabilities rises while the real value of their assets falls, compounding the drop in real entrepreneurial net worth N_t+1 and further amplifying the spike in the external finance premium s_t.

1. Bernanke, B., & Gertler, M. (1989). "Agency Costs, Net Worth, and Business Fluctuations." American Economic Review, 79(1), 14–31. https://www.jstor.org/stable/1804770
2. Bernanke, B. S., Gertler, M., & Gilchrist, S. (1999). "The Financial Accelerator in a Quantitative Business Cycle Framework." In J. B. Taylor & M. Woodford (Eds.), Handbook of Macroeconomics (Vol. 1C, Ch. 21, pp. 1341–1393). Amsterdam: Elsevier. https://doi.org/10.1016/S1574-0048(99)10034-X10034-X)
3. Bernanke, B., Gertler, M., & Gilchrist, S. (1996). "The Financial Accelerator and the Flight to Quality." Review of Economics and Statistics, 78(1), 1–15. https://doi.org/10.2307/2109844
4. Kiyotaki, N., & Moore, J. (1997). "Credit Cycles." Journal of Political Economy, 105(2), 211–248. https://doi.org/10.1086/262072
5. Christiano, L. J., Motto, R., & Rostagno, M. (2014). "Risk Shocks." American Economic Review, 104(1), 27–65. https://doi.org/10.1257/aer.104.1.27