- An entity’s carbon intensity is an input into Harbour’s ESG assessment of a borrower, alongside an appraisal of behaviours and performance in other environmental, social and governance (ESG) aspects.
- Carbon footprint is calculated as the weighted average of a portfolio’s ownership share of each holding company’s carbon output. This is a useful metric for equity fund investors, however fixed interest investors don’t have an ‘ownership’ stake in a company and therefore don’t typically receive an equivalent appraisal of their fund.
- Harbour’s solution to measuring carbon performance is by providing Weighted Average Carbon Intensity – a metric that is comparable across portfolios and, unlike some metrics, does not suffer from portfolio size distortions. Weighted Average Carbon Intensity measures the portfolio’s weighted average of each borrowing entity’s carbon output per million dollars of revenue.
- All of Harbour’s funds now include Weighted Average Carbon Intensity performance in their monthly Fund Fact Sheets which are available on our website.
Harbour has been publishing the carbon performance of its equity funds in factsheets and monthly reports since early 2021. These metrics are useful for equity fund investors and are becoming the norm. However, data challenges mean very few investors in fixed interest funds globally receive an equivalent appraisal of their fund[1]. This note explains how Harbour has approached such data challenges in publishing a carbon footprint for all its fixed interest funds.
How does Harbour use carbon data?
We are primarily looking at the appropriateness of the carbon intensity compared to peers in the same industry, rather than an outright level. This is an acknowledgment that producing goods creates more carbon than providing services. An entity’s carbon intensity is an input into Harbour’s ESG assessment of a borrower, alongside an appraisal of behaviours and performance in other environmental, social and governance aspects.
Standard approach: Carbon footprint
Convention, including that recommended by the TCFD (Taskforce on Climate-related Financial Disclosures) calculates carbon footprint as the weighted average of a portfolio’s ownership share of each holding company’s carbon output. In other words, a 1% ownership of a company would attribute 1% of that company’s emissions to the portfolio. For those mathematically minded:
As with the data provided in Harbour’s equity reports, the metric is typically divided by $m invested to provide a fair comparison across fund sizes.
The hitch for fixed interest investors is they don’t have an ‘ownership’ stake in a company (bold in the formula above). If fixed interest investors were to also use market capitalisation as a denominator, double counting would occur because the carbon output is already fully apportioned to equity investors. For this reason, most entities that own portfolios of both equity and debt, including, for example, the NZ Super Fund, measure their carbon footprint across their equity investments only[2].
Substituting market capitalisation with enterprise value (the value of a company’s debt and equity) as a denominator would avoid double counting. This may be a more sensible approach, however it is our judgment that bucking convention might create more confusion than it adds value[3].
Our solution: Weighted Average Carbon Intensity
Harbour’s solution is to provide Weighted Average Carbon Intensity - a metric that is comparable across portfolios and, unlike some metrics, does not suffer from portfolio size distortions. Weighted Average Carbon Intensity measures the portfolio’s weighted average of each borrowing entity’s carbon output per million dollars of revenue.
All of Harbour’s funds now include Weighted Average Carbon Intensity performance in their monthly Fund Fact Sheets which are available on our website.
Data availability
The portion of companies reporting carbon data is improving rapidly. Indeed, reporting coverage now extends to 78% of the Harbour Australasian Equity Fund. This will increase as climate-related disclosures legislation comes into force in New Zealand. For now, data providers such as ISS ESG, whom Harbour subscribes to, can use sophisticated models to estimate the remaining non-reporting entities that they cover.
However, carbon reporting from New Zealand fixed interest issuers has been less encouraging. Prominent among the non-reporting entities is the New Zealand Government (who passed the climate-related disclosures legislation, but exempted itself). At a 64% weighting in the Bloomberg Composite Bond Index, we believe omitting such a large borrower and grossing up the remaining borrowers compromises the integrity of a portfolio-level metric. We have been engaging with the Debt Management Office for an extended period on this issue and have confidence government-wide reporting is in the making. In the interim, we have created a proxy for the Government’s carbon footprint. To calculate this proxy, we have referred to the footprint of other nations and normalised for spend by activity, for example, normalising for the fact that countries that spend more on defence tend to have a higher footprint.
Limits to data
We note carbon accounting remains in its infancy and summary measures will suffer logical inconsistencies. Switching from the Dunedin City Treasury to the Local Government Funding Agency (LGFA) provides an example. The LGFA measures the carbon output of its operations as a bank for various local governments. As a large entity by dollars of borrowing, but a small entity by administrative staff, it has a low carbon intensity. The Dunedin City Treasury has recently announced it will borrow via the LGFA. This change would result in a fund reporting lower carbon outputs when there has been no real-world difference. This in part explains why we have chosen to use our proxy for the New Zealand Government for all local government borrowing.
Until Scope 3 data are more readily available, our reporting captures scope 1 and 2 emissions[4].
Why might the portfolio be more carbon intense than the benchmark?
While we are using carbon data at an issuer level, we do not manage the portfolio to target an overall carbon intensity. In portfolios benchmarked against the Bloomberg Composite Bond Index, an issuance-weighted mix of government and corporate securities, we typically hold less government securities as the NZ Government stock’s weight has risen above what we believe is optimal for investors following quantitative easing. Given the Government is largely an entity that provides services such as education, it has a lighter than average intensity and any bias away impacts the portfolio’s carbon intensity relative to the benchmark. We believe capital can make a difference if it is funnelled to those who are efficient at what they do.
[1] As at January 2023 we are yet to identify any peers publishing these data.
[2] Stamped-Carbon-Footprint-Report.pdf (nzsuperfund.nz)
[3] Calculating the enterprise value of some entities, such as local governments, is also problematic.
[4] For more information, including definitions, please refer to this explanatory publication.