The Hidden Cost of QE: How ECB Bond Buys Raised Bund Yields by 30bps

Over the full program, that adds up to a 30-basis-point reduction. Since the average safety premium across the four countries was 66 basis points during the sample period, the cumulative effect was nearly half of that average. The yields on German Bunds, Swiss government bonds, and their peers were 30 basis points higher than they would have been without the safety premium channel.

The effect was not limited to the euro area. Denmark, Sweden, and Switzerland all saw their safety premiums decline even though the ECB was buying bonds only from euro-area countries. That international spillover means an increase in safe assets in one region lowers the price of safety globally.

For traders, the mechanism matters whenever a central bank buys risky assets with reserves. The Fed's pandemic-era purchases included mortgage-backed securities and corporate bonds, not just Treasuries. That likely created a similar offset for U.S. Treasury yields, though the paper does not estimate it directly. The same logic applies to the Bank of Japan's purchases of ETFs and the Bank of England's corporate bond buying.

The practical implication: do not assume a given amount of QE will produce a proportional decline in safe yields. The composition of purchases matters. Buying risky assets expands the safe asset pool and compresses safety premiums, which works against the yield-lowering goal. The FRBSF paper provides a framework for estimating that offset: roughly 1 basis point per 1% of GDP in purchases, concentrated in the safest bonds.

The next time a central bank announces asset purchases, watch what it buys. If it is loading up on peripheral debt, corporate bonds, or other risky securities, the safety premium effect will mute the impact on Bunds, Treasuries, or JGBs. The 30-basis-point offset from the ECB's program is a concrete benchmark for that calculation.