1. Alan S. Blinder Princeton University Burt Malkiel Conference April 8, 2011 Rethinking Monetary Policy in a Very Low Inflation Environment

2. References 2 “Quantitative Easing: Entrance and Exit Strategies,” Federal Reserve Bank of St. Louis Review, November/December 2010, pp. 465-479 (the 2010 Homer Jones lecture). “Revisiting Monetary Policy in a Low Inflation and Low Utilization Environment,” Journal of Money, Credit and Banking, forthcoming in a special issue.

3. Things our intellectual fathers… 3 a) taught us b) didn’t tell us

4. Things our intellectual fathers… 4 a) taught us b) didn’t tell us ↑ one of mine

5. And while we’re on the subject… 5

6. Bagehot and recent history “Lend freely, at a penalty rate, against good collateral.”    Most CBs did   Well, not much.  Well, it varied. For the ECB, Bagehot+ was just about enough (until Greece). But the Fed, BoE, and BoJ felt compelled to do more. Why? The zero lower bound (ZLB) on nominal interest rates led to a perceived need for unconventional monetary policy.  ECB never hit the ZLB; neither did the BoE, but it came close. 6

7. Friedman and recent history “Don’t peg the nominal interest rate.” Why not? r = i - π Friedman was thinking mainly about upward instability when π is rising and r is falling. But the argument is symmetric: When i is stuck at zero, and π is falling, r is rising. Again, the ZLB leads to unconventional monetary policy. And, BTW, can also lead to huge fiscal multipliers. 7

8. Keynes and recent history The liquidity trap idea: In a very depressed economy, the central bank might push the short rate all the way down to zero--and still not stimulate the economy. Then monetary policy becomes useless. But fiscal policy becomes powerful. But what if fiscal policy is paralyzed by large deficits and/or public debt? 8

9. Four quick conclusions 1. (obvious) In an environment of very low inflation, we need to worry much more about the ZLB. 2. (less obvious, but should be) If that environment also has low utilization, we may need large doses of expansionary policy.  If fiscal policy is paralyzed, that must be monetary policy. 3. (deduction) Given our starting point today, unconventional monetary policy will be more important than in the past.  So the “crazy aunt” may not be stuffed back in the closet so easily. 4. Maybe we should think more about unconventional monetary policy options. 9

10. Types of unconventional monetary policy 1. Commitment via words (“extended period”) 2. A higher π * (because r = i – π) 3. Lower the interest rate on reserves (no ZLB here) 4. Quantitative easing ( ← will focus on this one) a) Treasury bonds (work on term premia) b) Private-sector assets (work on risk premia) 5. Pegging one or more bond prices  Exactly what our forefathers told us not to do! 6. Supervisory forbearance (if CB is a supervisor) 10

11. The underlying idea Idea: Demand curves for financial assets are not horizontal, so changing relative supplies can change term and/or risk premiums. Rj = r + ρ j Requires imperfect substitutes or “frictions” 11

12. A taxonomy of QE 12 Alter the composition of the central bank’s balance sheet. Increase the size of the central bank’s balance sheet. a) Longer-dated government securities b) Private-sector securities

13. Simplified Fed balance sheet ASSETS LIABILITIES AND NET WORTH Treasury bills Currency Less liquid assetsBank Reserves LoansTreasury deposits Capital 13

14. Specific strategies To shrink term premiums  Buy long-term government bonds... -- and sell T-bills -- by creating new bank reserves  Relies on imperfect arbitrage across yield curve (“preferred habitat” theory)  Related option: commit to keeping the overnight rate low for a long time  Note: This strategy relies on the expectations theory. 14

15. Specific strategies To shrink term premiums  Buy long-term government bonds... -- and sell T-bills -- by creating new bank reserves  Relies on imperfect arbitrage across yield curve (“preferred habitat” theory)  Related option: commit to keeping the overnight rate low for a long time  Note: This strategy relies on the expectations theory. This man taught me those theories. → 15

16. Specific strategies To shrink term premiums  Buy long-term government bonds... -- and sell T-bills -- by creating new bank reserves  Relies on imperfect arbitrage across yield curve (“preferred habitat” theory)  Related option: commit to keeping the overnight rate low for a long time  Note: This strategy relies on the expectations theory. To shrink risk premiums  Buy some risky asset… -- and sell the safe asset -- by creating new bank reserves  Relies on imperfect substitutability (e.g., preferred habitat) 16

17. Swapping assets (QE 0) ASSETS LIABILITIES AND NET WORTH Treasury bills  Currency Less liquid assets  Bank Reserves Loans ↑Treasury deposits Capital 17

18. Early QE did not blow up the Fed’s balance sheet… 18

19. …nor increase bank reserves much 19

20. Lehman changed everything 20

21. Lehman changed everything 21

22. Blowing up the balance sheet (QE I) ASSETS LIABILITIES AND NET WORTH Treasury securities Currency Less liquid assets↑Bank Reserves ↑ Loans ↑Treasury deposits Capital 22

23. CP and MBS purchases: Working on risk premiums R i = r + ρ i  riskless rate 23

24. Did QE I work? 24 Econometric evidence suggests: yes

25. Blowing up the balance sheet (QE II) ASSETS LIABILITIES AND NET WORTH Treasury securities Bills Bonds ↑ Currency Less liquid assetsBank Reserves ↑ LoansTreasury deposits Capital 25

26. Did QE II work? 26 It was supposed to work on term premiums. They have widened. (see next slide)

27. Three yield curves August 2010, November 2010, March 2011 27

28. Did QE II work? 28 It was supposed to work on term premiums. They have widened. (see next slide) But perhaps for other reasons:  brighter outlook for the economy  higher expected inflation  worsening outlook for national debt Too early for econometric evidence. (But you can guess!)

29. Questions for research/thinking 29 Which of the six unconventional monetary policies works best, under what circumstances? If it’s going to be QE, which kind?

30. There is a strong a priori case for using private assets: 30 Treasury market is the broadest, deepest in the world → hardest to move. Any other market is thinner → easier to move. The substitutability between T-bills and T-bonds must be higher than the substitutability between, say, T-bills and MBS.

31. Thank you 31