The Simple Analytics of Monetary Policy: A Post-Crisis Approach

Abstract

The standard workhorse models of monetary policy now commonly in use, both for teaching macro- economics to students and for supporting policymaking within many central banks, are incapable of incorporating the most widely accepted accounts of how the 2007–9 financial crisis occurred and are incapable too of analyzing the actions that monetary policymakers took in response to it. They also offer no point of entry for the frontier research that many economists have subsequently undertaken, especially research revolving around frictions in financial intermediation. The author suggests a simple model that bridges this gap by distinguishing the interest rate that the central bank sets from the interest rate that matters for the spending decisions of households and firms. One version of this model adds to the canonical “new Keynesian” model a fourth equation representing the spread between these two interest rates. An alternate version replaces this reduced-form expression for the spread with explicit supply and demand equations for privately issued credit obligations. The discussion illustrates the use of both versions of the model for analyzing the kind of breakdown in financial intermediation that triggered the 2007–9 crisis as well as “unconventional” central bank actions like large-scale asset purchases and forward guidance on the policy interest rate.

Keywords:

• central bank asset purchases
• central banks
• financial crisis
• forward guidance
• monetary policy
• unconventional monetary policy

JEL code:

• E52

Acknowledgments

This article is based on a paper that was prepared for the session on “After the Crisis: What Did We Learn, and What Should We Teach, about Monetary Policy?” sponsored by the AEA Committee on Economic Education at the AEA meetings in San Diego, January 5, 2013. The author is grateful to numerous colleagues, especially Kenneth Kuttner, for helpful conversations.

Notes

1. See, for example, Brayton and Tinsley (1996), Smets and Wouters (2003), and Christiano, Eichenbaum, and Evans (2005).

2. Bernanke and Blinder (1988), Bernanke and Gertler (1989), Holmstrom and Tirole (1997), and Kiyotaki and Moore (1997) from before the crisis, and Woodford (2010), Gertler and Karadi (2011), Gertler and Kiyotaki (2011), Curdia and Woodford (2011), Choi and Cook (2012), Gilchrist and Zakrajsek (2012), Amano and Shukayev (2012), and Williams (2012), all from after, are just a few of the most prominent examples already in the published literature as of the time of writing. The pipeline of current not-yet-published research inspired at least in part by the crisis is far larger.

3. See, for example, Clarida, Gali, and Gertler (1999). As Roberts (1995) has shown, it is not necessary to assume Calvo's specific form of price inflexibility to motivate the now-standard form of the Phillips curve; staggered contracting, as in Taylor (1979), or quadratic costs of adjustment, as in Rotemberg (1982), deliver an equivalent expression for the behavior of the aggregate price level. Gertler and Leahy (2008) have shown that the familiar “S,s” model of pricing does so as well.

4. See Friedman and Kuttner (2011) on what gives the central bank this control.

5. The introduction of private-sector borrowing and lending (beyond the mere existence of a market in which no one chooses to participate) also means that the model can no longer bear the interpretation of a representative agent: If all private agents were identical, there would be no reason for any one of them to borrow from or lend to any other. But for purposes here, there is no need for a representative-agent model.

6. An alternative way to represent this kind of forward guidance would be to posit that the central bank has adopted a new behavioral pattern to replace its historical behavior as in (2), and the purpose of the public announcement is to inform investors about this change.

7. When investors have constant relative risk aversion and believe that returns on risky assets are normally distributed, asset demands are linear in expected excess returns so that this relationship is linear as well; see Friedman and Roley (1987).

8. Shocks to the credit market, like a decline in borrowers’ collateral value or an erosion of lenders’ capital, may also result in increased credit rationing (i.e., borrowers’ inability to obtain loans despite willingness to pay the interest rate charged to similar borrowers who do receive loans). The model presented here abstracts from such rationing and therefore implicitly treats the spread (ρ − r) as a sufficient statistic for this influence on aggregate demand as well.

9. See, for example, Gagnon et al. (2011), Hancock and Passmore (2011), Krishnamurthy and Vissing-Jorgenson (2011), and Hamilton and Wu (2012).

10. Here it is even more apparent that with private-sector borrowing and lending what results cannot be a representative-agent model.

11. The linearity with respect to expected excess returns again follows under conditions of constant relative risk aversion and normally distributed returns; substitutability coefficient θ₂ in (4^D) is then directly proportional to investors’ total wealth (A in the corresponding reduced-form expression in the four-equation model), and inversely proportional to the coefficient of relative risk aversion and to the variance associated with the risky asset's return. See again Friedman and Roley (1987).

12. For given initial quantities of private-sector obligations outstanding and a given amount held by the central bank, c^S, c^D and c^CB can be equivalently interpreted as either flows or stocks. The market-clearing condition (4^MC) makes clear that the model developed here abstracts from credit rationing; see again note 8.

13. It is also possible to suggest ways in which a reduced value of borrowers’ collateral would depress c^S (i.e., would reduce the demand to borrow even if there were no change in lenders’ willingness to lend to them at a given interest rate). But the effect on c^D is more straightforward, and it is what the literature analyzing the 2007–9 crisis has mostly emphasized.

14. One prominent exception before the 2007–9 crisis was the U.S. Federal Reserve's experiment with a nonborrowed reserves operating procedure during 1979–82.

15. See Friedman and Kuttner (2011) for a detailed discussion of this feature of monetary policy implementation.