How are green bonds priced versus non-green bonds on the secondary market? In this analysis, we break down pricing by sector, credit rating, country and currency and we examine whether differences exist in terms of bonds’ green credentials.
There has been considerable research on the pricing of green bonds in the primary market. Until now, though, there has been no complete, in-depth research on pricing of green bonds in the secondary market. To the best of our knowledge, this FocusPoint is the first paper that focuses exclusively on green bond pricing in the secondary market across the full spectrum of global green bonds. It answers several key questions. How much is the green bond’s premium? How has it evolved over time?
Average monthly yield difference
Source: NN Investment Partners
Green bonds are bond instruments whose proceeds are used to finance projects beneficial to the environment. They are otherwise identical to unsecured issues. A bond’s “green” label depends on the type of projects it funds, not on the issuer’s green credentials.
Since the 2014 publication of the first Green Bond Principles, the market for green bonds has grown to over EUR 270 billion. There are several reasons for the strong demand for green bonds from investors. First of all, green bond issuers voluntarily comply with disclosure agreements and performance indicators that meet investors’ requirements for greater transparency. Most green bonds have an independent second-party opinion, which helps investors to better understand how the issuer would allocate the use of proceeds. Green bond issuers have also shown a commitment to the integration of finance and sustainability teams to mitigate climate-related risks. The transition to a carbon-neutral economy is likely to impact the private sectors. Issuers that take steps to face the climate-related risks could have an advantage.
The methodology of our analysis
We defined our green bond universe using the bonds included in Bloomberg MSCI Global Green Bond Index as a starting point. In our view this is the most credible and most widely used green bond index. The index includes only investment grade and liquid green bonds; it does not include social and sustainability bonds. We have collected monthly data from December 2014 to November 2017. The market before this period was too fragmented and too small to use in the analysis. Our sample includes 133 unique labelled green bonds issued by 59 entities from 16 countries and 7 supranational organisations.
For the non-green bond group we have taken the bonds from green bond issuers in the Bloomberg Global Aggregate, Bloomberg Euro Aggregate, Bloomberg Canadian Aggregate and Bloomberg Australian Aggregate indices. We included green bonds in our analysis only if we had five or more non-green bonds to be able to interpolate the yield curves. To quantify differences we have interpolated the yield curve on a monthly basis per issuer and have assumed that the relationship between maturity and yield is not necessarily linear. The expected (interpolated) yield and the yield of the labelled green bond are compared for every month, resulting in a difference between the interpolated yield curve of the issuer versus the yield of the green bond. A negative difference means the interpolated yield is higher than the yield of the green bond with the same maturity and seniority. The opposite holds when the yield difference is positive. We used simple average to calculate the average yield difference per month and per segment. On average every green issue is matched with 14 non-green bonds of the same issuer with similar seniority. Our full sample includes 2,417 data points of green bonds (36 months, 133 unique green bonds) and in our view gives a reliable estimate how much green bonds deviate from the issuer’s curve. We believe our model displays a good balanced between simplicity and a goodness-of-fit that minimises statistical error.
The results of our analysis
The average yield on green bonds in the sample was lower than the interpolated yield of non-green bonds. The difference between the observed and interpolated yields for the entire sample was -0.011%, meaning that on average and over time, a green bond yield is 1.1 basis points (bps) lower than a non-green bond yield. Still, some 37% of the green issues had a yield above the interpolated curve.
Furthermore, we divided our sample in two equal periods of 18 months. The first sub-sample, December 2014 to May 2016, had 67 unique green bonds and the second sub-sample, June 2016 to November 2017, had 126. On average the yield of green bonds in the first sub-sample was 2.3bps lower than the issuer’s interpolated yield, compared with 0.7bps in the second sub-sample. The figure summarizes the number of available issues per month and the average difference between the yield on the green bond and the interpolated yield of the issuer’s curve.
Two factors may explain the difference between yields for green and non-green bonds. One is a possible mismatch between supply and demand. The difference in yields may be a result of growing interest from investors and a limited number of issues with a green feature. The growth of the green bond market increases investors’ choices and could explain why green bonds have become less expensive over time compared to their issuer’s curve.
Another explanation is that in certain market circumstances, green bonds may be less volatile than their peers. In periods of risk aversion, green bonds tend to be more stable, due to more buy-and-hold investors holding the bonds in their portfolios. The bond’s lower volatility compensates the investor for its lower yield. The reasoning here is that the investor in green bonds has a long-term horizon and does not trade actively, hence reducing the price volatility.
The green bonds universe was at its smallest at the beginning of the sample period, which means there were relatively few eligible bonds to be included in our analysis. This increases the weighting of each green bond in the earlier periods when calculating the average yield difference versus the non-green bond curve. For instance, there are only 45 green issues in our November 2015 sample, giving each bond a weight of 1/45, while in November 2017 there are 116 green bonds, each of which has a weight of only 1/116. The volatility of the yield difference time series also decreases as the sample size grows.
Conclusions from our analysis
Investor demand for green bonds has increased dramatically in the past three years. The market’s growth has made it possible to analyse pricing of these bonds relative to non-green bonds. Based on our analysis of the greater part of the secondary market for green bonds, we can draw the following conclusions:
- On average green bonds yields are only slightly below the interpolated yield of non-green bonds, with a difference of -0.011%, or -1.1bps.
- As a result of the strong growth of the green bond market, the sample of green bonds increased and their yields moved closer to the interpolated yield of non-green bonds. The difference in the second sub-sample (from June 2016 to November 2017) was only -0.007%, or -0.7bps.
- Roughly 37% of the green bonds have a yield higher than their interpolated non-green curve.
- Yield differences across segments (country, currency, sector, credit rating, use of proceeds) vary a lot. More importantly, within segments, the gap between maximum and minimum yield difference could be as high as 20bps. This makes active management an important factor that can add value.
John Buckley, Senior Financial Editor NN Investment Partners