The Benchmark Greenium
Abstract
In this paper, we study the benchmark greenium, the frictionless premium of sovereignrisk-free green securities relative to otherwise identical non-green securities. Identifying this type of greenium is important for two reasons. First, since sovereign green bonds are not project-specific, in the absence of market frictions, our greenium will capture the shadow value of wide environmental concerns. Second, because bids in the primary auction market depend largely on resale prices in the secondary market, the benchmark greenium provides a signal about future savings from the issuance of green securities.
Exploiting the unique ``twin'' structure of German government green and conventional securities, we use a dynamic term structure model to estimate a frictionless sovereign risk-free greenium, distinct from the yield spread between the green security and its conventional twin (the green spread). The model purifies the green spread from confounding and idiosyncratic factors unrelated to environmental concerns.
We have three main findings. First, the model-implied benchmark greenium differs substantially from the observed green spread: it tends to be significantly larger; at times it
widens while the green spread narrows; and its term structure is mostly flat, rather than
downward-sloping. Second, proxies of confounding and idiosyncratic risk factors, such as
stock market prices, measures of flight-to-quality, and liquidity, do not affect the model-
implied greenium, but do correlate with the green spread. Conversely, the benchmark greenium correlates with shocks to environmental concerns, such as the economic damages from environmental disasters. Interestingly, once we control for confounding and idiosyncratic risk factors, the green spread does not correlate significantly with shocks to environmental concerns. Third, the difference between the expected return on green and conventional bonds, the expected green excess return, varies with the investment horizon and investors’ information set: it is positive at issuance and turns negative after the German floods.
Our findings are important not only because they indicate that green spreads should not
be taken at their face value to price environmental concerns, but also due to their policy
implications. First, the fact that the benchmark greenium is not reverting to zero over
time signals that green security issuance can provide interest cost savings to governments. However, to access these savings, government finance agencies need to minimize market
frictions. Second, a persistent benchmark greenium justifies the inclusion of green assets by central banks in the conduct of conventional and unconventional monetary policy, since it indicates investors’ preference for green investments, and eligibility for central bank
operations reduces liquidity risks and other frictions affecting these assets, especially in
periods of crisis. Third, a risk-free benchmark greenium will allow for more efficient pricing
of private green securities, strengthening the market for green investments.