Goldman's Sylvia Ardagna puts Europe's QE in perspective. Just in case someone is still confused why Treasurys around the globe are "no offer" for days on end.
* * *
The size of US QE
The Federal Reserve’s balance sheet increased by about $3.7trn from November 2008, when the central bank announced it would begin buying agency debt and MBS, to October 2014, when QE3 ended. Cumulative purchases of long-term treasury securities amounted to about $2trn. The breakdown by period is as follows: $300bn from November 2008 to June 2010 (QE1), $700bn from November 2010 to June 2011 (QE2) and $810bn from September 2012 to October 2014 (QE3).
Over the same period, the net issuance of US Treasury bonds and notes was a cumulative $7.2trn (of which a $6.5trn increase in marketable debt) and gross issuance of
USTs with a maturity of 2-year and above was $12trn. Hence, since the global financial crisis, the Fed has absorbed a cumulative 28% of net issuance and 16% of gross issuance of USTs, at an average pace of purchases of about $24bn a month.
In terms of the pace of purchases, the second QE2 programme was the largest: in that period the central bank bought about $100bn in USTs per month, equivalent to about 50% of gross issuance and 148% of net issuance. In terms of duration removal, the maturity extension programme (QE3) had the largest effect. When converted into 10-year equivalents, the increase in the Fed UST portfolio was about $1.8trn, and the 10-year equivalent gross issuance over the same period was about $8.3trn, or about 20% of gross issuance.
A EUR500bn Euro area QE would be sizeable relative to expected bond net issuance
Should the ECB announce EUR500bn in government bond purchases to be implemented over a one-year period, as our European Economics team expects, this programme would compare in size to the average monthly purchases of USTs by the Fed during QE3, but it would be significantly larger than the average monthly Fed purchases since the beginning of the global financial crisis. Moreover, given the smaller stock and lower net and gross issuance of government debt in the Euro area than in the US, the programme size would imply a significant absorption of government securities available for private sector investors.
The ECB's stock of eligible Euro area government bonds is EUR7trn (by comparison, the stock of US government securities is about $12trn) and we estimate 2015 net government bond issuance to be around EUR90bn and gross issuance to stand at around EUR800bn (see Macro Rates Monitor, December 19, 2014). The ECB would, hence, buy about 62% of gross issuance of long-term bonds in the Euro area countries and more than five times as much as the net issuance.
Looking across countries, if purchases are determined according to the ECB’s capital key contributions (which is our baseline assumption), the ECB would buy about EUR130bn of German securities, EUR100bn of French, EUR90bn of Italian and EUR60bn of Spanish securities, equal to about 90%, 60%, 43% and 45% of the respective 2015 gross issuance. These numbers are closer to the amount of “supply removal” of the BoJ under the current asset purchase programme. The latter is buying approximately 50% of gross issuance of JGBs and about 90% of gross issuance of 10-year JGBs.