The latest IMF publications on non-conventional monetary policies affirm their effectiveness. This seems to hold true for all major forms, i.e. government bond purchases, forward guidance, and private asset purchases. However, bond purchases are likely to yield diminishing effects going forward. Their initial stimulus seems to owe much to the signalling of commitment and the repair of broken markets. Also, the economic impact seems to have a time limit. Meanwhile additional credible yield compression becomes ever more difficult as the zero boundary is drawing closer and central banks would struggle to extend commitments to ever longer durations.


Unconventional Monetary Policies – Recent Experiences and Prospects, IMF Policy Paper, May, 16 2013. http://www.imf.org/external/pp/longres.aspx?id=4764
Summary of Informal Discussions with Central Bankers and Other Officials on Unconventional Monetary Policies, IMF Policy Paper, May, 16 2013. http://www.imf.org/external/pp/longres.aspx?id=4766
Unconventional Monetary Policies – Recent Experiences and Prospects, Background Paper, May, 16 2013 http://www.imf.org/external/pp/longres.aspx?id=4765

“Central banks in the United States, United Kingdom, Japan, and euro area adopted a series of unconventional monetary policies with two broad goals. The first was to restore the functioning of financial markets and intermediation. The second was to provide further monetary policy accommodation at the zero lower bound… The first category includes liquidity provision and outright purchases of private and public assets. The second category covers both purchases of government bonds (bond purchases) as well as forward guidance.”

Non-conventional policy I: Government bond purchases

What were the effects of bond purchases on long-term yields?

“[The] preferred approach in the literature to isolate the effect of purchases is event studies. This entails measuring changes in yields in a very narrow time window—typically one day—around an official announcement related to bond purchases…In the U.S., the cumulative effects of bond purchase programs are estimated to be between 90 and 200 bps. Most studies focus on LSAP [Large-Scale Asset Purchases] 1 where the largest effects are found (between 50 bps and 100 bps). In the U.K., cumulative effects range from 45 bps to 160 bps. In Japan, [IMF] staff estimates that purchases of government bonds under the Comprehensive Monetary Easing and Qualitative and Quantitative Monetary Easing policies reduced 10-year yields by a little over 30 bps.”

“In investigating the effectiveness of bond purchases, it is desirable to…explore the bang (change in yields) for the buck (degree of surprise) of announcements…[IMF] staff measure surprises by changes in one year ahead futures on short-term interest rates, as futures to some extent reflect market expectations of policy moves to come…Bigger surprises had bigger effects on yields (including disappointing—positive—surprises which increased yields)…A 25 bps decrease in one year ahead futures would decrease 10-year bond yields, on average, by 25 bps in the U.S. and 20 bps in the U.K. and Japan… The effectiveness of monetary surprises associated with bond purchases appears similar to that of conventional rate cuts prior to the crisis. In other words, changes in bond yields per unit of surprise were similar pre and post-crisis. ”

Initial announcements were especially effective at easing financial conditions, partly by decreasing tail risks of a severe recession….The skewness of inflation forecast distributions decreases noticeably after the first announcements of bond purchases, suggesting that deflation (or very low inflation) tail risks were reduced.”

What were the main transmission channels for bond purchases—signaling, scarcity, or duration?

“A decrease in long-term yields coming through the signaling channel has an effect on GDP growth approximately twice as large as the same shock coming through the portfolio rebalancing channels. The result is consistent with theory. Shocks to long-term rates due to the portfolio rebalancing channel may be expected to be more temporary and reversible, in part due to the volatile market conditions on which this channel relies….[Also] lower premia on riskier long-term bonds induced by portfolio rebalancing might lead firms to buy back shorter-term debt with longer-term issuance, instead of fostering productive investments.”

Have bond purchases been effective at stimulating the economy?

“Both GDP growth and inflation reacted positively and substantially to bond purchases. Effects, though, were short-lived. Yet, they were more persistent for inflation than for output….Most papers find that [in response to a 100bps drop in long-term bond yields] GDP growth increases around 2%-points in the U.S. and U.K, generally lasting around two years…the range [of estimates] is very large (between 0.1%-points and 8 %-points). Effects on inflation are as large as 3.6%-points points, though again within a wide range.”

“How have unconventional measures affected the rest of the world?

Foreign financial spillovers from unconventional measures are largest when policies restore market stability or significantly change the monetary framework. LSAP 1, which met these criteria, led to major global financial market rallies, involving generalized reductions in bond yields, rises in equity prices, and appreciation of foreign currencies vis-à-vis the U.S. dollar. Early announcements of asset purchases in the U.K. had similar exchange rate effects.”

What are the limits of effectiveness of bond purchases?

“Theory suggests that diminishing returns [of bond purchases] will kick in… First, effects through the signaling channel are likely to wane. Bond purchases will lose effectiveness as longer-maturity bond rates approach the zero boundary. In order to continue to lower expected real rates, central banks will have to commit to keep policy rates low for an increasingly longer horizon; but the longer the horizon, the lower the credibility of the commitment.”

“The portfolio rebalancing channels of bond purchases rely on some degree of market segmentation. Thus, they will weaken as health in the financial sector is restored and arbitrage conditions strengthen…Yet, if market functioning remains impaired, the portfolio rebalancing channels should continue to operate, even if the signaling channel wanes.”

“[In the U.S.] consistent with theory, the announcement of LSAP 1 in an environment of acute uncertainty had a visibly bigger impact on bond yields than LSAP 2, a trend that seems to have continued with later programs. The same is true of the effect on tail risks. This suggests that further increases in asset purchases would have incrementally smaller effects.”

Non-conventional policy II: Forward guidance

“Forward guidance entails managing market expectations of future policy, with explicit communication (or other information) on the central bank’s reaction function and economic projections…At the zero boundary…forward guidance can be used to convince markets that the central bank will keep rates low for longer (allow inflation to go higher) than consistent with its usual policy rule..If successful, this will decrease expected future nominal and real rates (flatten the yield curve), and create expectations of a stronger recovery.”

“Studies…suggest that forward guidance has been at least partly effective during the crisis. Swanson and Williams (2012) show that the sensitivity of long-term government bond yields to macroeconomic news declines after explicit forward guidance. Likewise, Woodford (2012) shows how the “extended period of time” language of the U.S. Federal Reserve and the Bank of Canada statements (April 21, 2009) on conditional forward guidance lowered expectations of future interest rates, as measured by Overnight Index Swap (OIS) rates… Forward guidance is less effective when it fails to signal a deviation from the central bank’s ‘normal’ reaction function. That may have been the case in Sweden, speculates Woodford (2012), as markets did not align their interest rate projections with those published by the central bank. Similarly, the BOJ’s forward guidance during 1999–2000 had a negligible effect (Ugai, 2007). ”

Non-conventional policy III: Private asset purchases

Private asset purchases to support financial intermediation show positive results, though mostly in the U.S. In the first phase of LSAP 1, the Fed purchased mortgage-backed securities (MBS) and Agency bonds at a time when the spread over Treasuries was unusually high, in a sign of market dysfunction. LSAP 1 appears to have decreased MBS [mortgage-backed securities] yields by 150 bps (Krishnamurthy and Vissing-Jorgensen, 2010), and mortgage rates by nearly 50 bps (Hancock and Passmore, 2011)…The purchase of corporate bonds, Exchange Traded Funds (ETFs), and Japanese Real Estate Investment Trusts had significant effects at the introduction of the BOJ’s Comprehensive Monetary Easing (Lam, 2011, Ueda, 2012). Yet, the effects appear limited for pre-2010 purchases of ABS in Japan.”