ECON6002 · Macroeconomic Analysis
Financial Frictions & Macroprudential Policy
Financial frictions and macroprudential policy form Topic 10 of ECON6002 Macroeconomic Analysis at the University of Sydney — the capstone of the New Keynesian block. The standard NK model (Topics 7–9) assumes frictionless finance: one representative household, Modigliani–Miller, and no gap between the borrowing rate and the saving rate, so there is no credit spread and nothing for financial policy to do. Following Suh (2014) this chapter splits households into patient savers (β) and impatient borrowers (β_b < β), inserts financial intermediaries whose cost makes the borrowing rate exceed the saving rate, and lets that spread widen with credit — a financial accelerator. Because a second friction adds a second objective, the Tinbergen count says you need a second instrument: macroprudential policy (φ_L) for financial stability while the policy rate keeps targeting inflation.
What this chapter covers
- 011. Why standard NK is not enough — representative agent, Modigliani–Miller, no credit spread, no intermediaries, no systemic risk
- 022. The Suh (2014) extension — patient savers (β) vs impatient borrowers (β_b < β) plus a banking sector
- 033. Two-type Euler equations — savers face R^s, borrowers face R^b > R^s, so aggregate demand becomes financially sensitive
- 044. The credit spread — R̂ᵇ = R̂ˢ + ζℓ̂: level term (1+ω+φ_L) sets the steady-state spread, elastic term ζℓ̂ makes it cyclical
- 055. The financial accelerator — credit up → spread up → borrower rate up → demand down → feeds back onto credit
- 066. Macroprudential policy (φ_L) — countercyclical LTV/DTI caps, capital buffers and reserve requirements that lean against credit booms
- 077. The separation principle — policy rate for inflation, macropru for financial stability; Taylor principle φ_π > 1 still required
- 088. Two frictions, two instruments — the Tinbergen argument and Suh's result that macropru stabilises credit sharply at small inflation cost
Log-linearise the credit-spread equation
- +1Take logs of the level equation: log R^b = log(1 + ω + φ_L) + log Ξ₀ + ζ·log ℓ + log R^s. The product becomes a sum and the exponent ζ comes down as a coefficient.
- +1Identify the constants. Both log(1 + ω + φ_L) and log Ξ₀ take the same value every period — they are level shifters, not time-varying.
- +1Subtract the steady state. The two constants cancel in the deviation, leaving the log-linear spread equation R̂ᵇ = R̂ˢ + ζ·ℓ̂.
- +1Read off the spread: R̂ᵇ − R̂ˢ = ζ·ℓ̂. The credit spread rises with the volume of real credit, with elasticity ζ.
- +1(c) Evaluate: ζ·ℓ̂ = 0.5 × 8% = +4%. An 8% credit boom widens the cyclical spread by about 4%.
- +1(b) Why R^b > R^s: intermediaries charge a markup over the deposit rate to cover the intermediation cost ω (the constant part of the spread), and the marginal cost of lending rises with the loan book (ζ > 0), so the spread also widens with credit.
- +1Where φ_L shows up: it sits inside the level term (1 + ω + φ_L), which dropped out of the log-deviation. So macropru moves the steady-state spread, not the cyclical coefficient ζ.
- +1State the assumption: borrowers must be impatient (β_b < β), otherwise no one borrows in equilibrium, the two Euler equations collapse to one, and the spread vanishes.
Key terms
- Financial friction
- Any imperfection — an intermediation cost, a collateral constraint, a credit spread — that breaks the frictionless-finance assumption of the standard NK model. It is what gives finance an independent role in the macroeconomy.
- Credit spread
- The borrowing rate minus the saving rate, R^b − R^s. It is exactly zero in the standard NK model (one representative household, Modigliani–Miller) and strictly positive here; in data it is large and strongly countercyclical, blowing out in recessions.
- Financial accelerator
- The feedback loop by which credit conditions amplify shocks: a rise in credit widens the spread, which raises the borrower's real rate and cuts demand and net worth, feeding back onto credit. In the model it is carried by the elastic term ζℓ̂.
- Macroprudential policy (φ_L)
- A countercyclical tool (φ_L > 0) targeting systemic risk and excessive leverage — proxying loan-to-value (LTV) and debt-to-income (DTI) caps, countercyclical capital buffers and reserve requirements. It leans against credit booms without moving the policy rate.
- Patient vs impatient households
- The two agent types in Suh (2014): patient savers with a high discount factor β who lend, and impatient borrowers with β_b < β who borrow. The impatience assumption is what makes lending happen in equilibrium and a spread emerge.
- Separation principle
- The assignment rule that the policy rate targets inflation while macroprudential policy targets financial stability. The two instruments are largely independent, and the Taylor principle φ_π > 1 is still required for a determinate equilibrium.
- Lean against the wind
- Using the monetary-policy rate to curb a credit boom. Suh (2014) and Svensson (2017) are sceptical: leaning with the policy rate over-tightens demand and misses the inflation target, so the credit boom is better addressed with a separate macroprudential instrument.
Financial Frictions & Macroprudential Policy FAQ
Can AI help me with financial frictions and macroprudential policy?
Yes — ask Sia to walk through any financial-frictions or macroprudential-policy problem or concept step by step, the way University of Sydney tests it. It is an AI tutor that explains the derivation — log-linearising the credit spread, the two-type Euler equations, the Tinbergen two-instruments argument — so you learn to reproduce it in the closed-book final, rather than an answer service.
Why does the standard New Keynesian model have no credit spread?
Because it has a single representative household that effectively lends to itself, so one interest rate clears every market and firm financing is irrelevant (Modigliani–Miller). With no borrowers-versus-savers distinction and no intermediaries, the borrowing and saving rates coincide (R^b = R^s). Suh's whole point is to break that: add intermediation and R^b > R^s, a spread that moves over the cycle.
Why do borrowers need β_b < β?
For lending to occur in equilibrium someone must want to borrow. The impatient household discounts the future more heavily (β_b < β), so at the same rate it brings consumption forward and runs down net worth while the patient household saves. Without β_b < β the two Euler equations collapse into one and the spread disappears — so always state this assumption.
Does macroprudential policy replace monetary policy?
No — it complements it. The separation principle assigns the policy rate to inflation and the macroprudential tool φ_L to financial stability. The Taylor principle φ_π > 1 is still needed for determinacy, and macropru cannot substitute for inflation stabilisation. Suh's result is that stronger macropru dampens the credit cycle sharply at only a small cost to inflation stabilisation.
What is the most common Topic 10 exam trap?
Writing R^b = R^s — that is the frictionless case the whole topic is built to fix. Two more: forgetting to state β_b < β, and putting φ_L in the cyclical coefficient ζ when it actually sits in the level term (1 + ω + φ_L), so it moves the steady-state spread, not ζ. Getting these straight is where the reasoning marks are.
Is Topic 10 on the ECON6002 final exam?
Yes. It is the last of the examinable Topics 7–10 on the 55% closed-book final, which provides a standard log-linearised NK formula sheet — so marks come from log-linearising, naming the mechanism (why R^b > R^s) and stating assumptions, not from recall. A True/False/Uncertain part on the 'one instrument, two targets' question is a signature item. Practice with Tutorial 10, and confirm the format against your own unit outline.
Studying with AI? Sia — free AI economics tutor works through ECON6002 step by step.
Exam move
Treat Topic 10 as one equation plus one argument. The equation is the credit spread R̂ᵇ = R̂ˢ + ζℓ̂: be able to log-linearise it from the level form, show the constants cancel, and separate the level shifter (1 + ω + φ_L), which sets the steady-state spread and is where intermediation cost and macropru bite, from the elastic term ζℓ̂, which is the cyclical financial accelerator. The argument is Tinbergen: two frictions (nominal rigidity plus the credit friction) create two objectives, so one instrument cannot hit both — the policy rate over-tightens if it leans against credit, so you add a second, macroprudential instrument φ_L and keep the policy rate for inflation (the separation principle, with φ_π > 1 still required). Drill the standard True/False/Uncertain — 'the policy rate alone can stabilise both inflation and credit' is False — and rehearse the five-line answer: name the standard-NK limitation, add the Suh ingredients (savers vs borrowers, intermediaries, the spread), explain the financial accelerator, invoke Tinbergen for the second instrument, then state the separation principle and Suh's small-cost result (supporting Svensson). Always state β_b < β, never write R^b = R^s, and remember macropru lives in the level, not in ζ.
