In the first week of June, volatility in European and US sovereign debt markets picked up again after a two-week hiatus. As with the rise in yields in April and early May, this move again appears to be European-led, and the 10-year Treasury-Bund spread narrowed further on continued Bund underperformance. Broad indices of European equities such as the STOXX Europe 600 continued to underperform the S&P 500, peripheral European debt spreads to Bunds increased late in the week towards recent highs, and the euro five-year forward five-year real swap rate is now at its highest level in six months; overall, it was a poor week for European financial conditions.
Immediate catalysts for the rates sell-off included stronger-than-expected Eurozone May inflation data against a backdrop of what increasingly looks like a bottoming of global inflation (see Exhibit 1 below). An upside surprise to the May US payrolls print on 5 June 2015 also contributed. In addition, comments from European Central Bank (ECB) President Draghi served as an unanticipated (and unfortunate) source of market volatility. His remarks during the post-policy meeting press conference, to the effect that markets should get used to volatility, had an immediate and negative effect on global bond markets. It is not that his comments were inaccurate. There are good reasons to expect heightened volatility when yields are very low, when the Federal Reserve (Fed) appears set to tighten policy later this year, and inflation has ticked up in the Eurozone. But the comments were damaging to markets because they appeared to turn a deaf ear to the fact that quantitative easing (QE) works in part by suppressing fixed income market volatility and thereby easing financial conditions. In Draghi’s defense, he did clarify that the ECB was monitoring market developments and could always reconsider program size and design if monetary conditions tighten. But as with then-Fed-Chairman Ben Bernanke’s relatively balanced comments in the spring of 2013, market participants instead focused on the portion of Draghi’s remarks that spoke to their greater concerns about possible losses on fixed income holdings.
The recent volatility in European sovereign bond markets begs an important question – is Eurozone QE inherently more volatile in its market effects than QE programs elsewhere? There are reasons to believe that it might be, as both the “stock and flow effects”1 of Eurozone purchases may be amplified or more variable over time. With regard to flow effects, neither the ECB nor the national central banks release a calendar of monthly purchases, and indeed there is little information on what drives purchase decisions on a daily or monthly basis. There is also little to no information on purchase methodology. This lack of transparency implies that when purchases occur each day, they have a greater price effect compared to, for example, QE in the United States, where monthly purchases were highly telegraphed. As to stock effects, without information on intended purchase targets across the yield curve, investors are not able to calculate with any certainty the total amount of duration that the Eurosystem will eventually remove from the market. Instead, they are left to rely on ex-post monthly purchase information that is difficult to interpret in the absence of general principles on what drives purchases across each sovereign’s yield curve.
Stock effects are further amplified by the negative-20-basis-point floor on purchase eligibility. As securities, particularly bunds, move in and out of the eligible purchase basket, market expectations for duration removal shift in real time. This dynamic works in a manner that amplifies recent market moves. For example, when yields rise and additional shorter-maturity securities become eligible for purchase, investors reduce their expectations for purchases at longer maturities, which can cause longer-dated yields to rise further.
A final stock effect relates to sovereign issuance. European sovereigns tend to be fairly opportunistic in their debt issuance patterns, which has created uncertainty about whether long-term debt issuance will increase to take advantage of low yield levels.2 Any terming out of issuance can work counter to the objectives of the ECB by adding duration back into the market.
Are these arguments purporting a more volatile QE regime in Europe borne out by the data? Exhibit 2 below compares the standard deviation of daily price returns for 10-year sovereign securities in the 60 days prior to and following the implementation of QE programs in various countries. By this metric, thus far ECB QE does not appear to have led to meaningfully more volatility than similar programs elsewhere. While volatility declined in the United States in the first 60 days following the start of Treasury QE3, it rose following the introduction of QE2 by a roughly similar amount as seen under Eurozone QE. The same holds for the Bank of Japan’s QE. A second metric of volatility, intraday yield changes, leads to broadly similar findings. Thus, after sixty days, it is still too early to draw conclusions about the impact of ECB QE on bond market volatility. But President Draghi can help the ECB’s cause by being careful not to sound indifferent to market volatility. This is all the more important given the reasons highlighted above that ECB QE may ultimately have a more variable effect on markets than QE programs elsewhere.
Exhibit 1: GDP-weighted inflation of 20 countries comprising 85% of World GDP
Source: Bloomberg, IMF, Author’s calculations
Exhibit 2: Standard Deviation of Daily Price Returns (%) on 10-Year Sovereign Securities
“Pre-QE” standard deviation is calculated over the 60 days prior to the introduction of asset purchases. “Post-QE” covers the announcement day and subsequent 60 days.
Source: Bloomberg, Author’s calculations