What leads to this counter-intuitive outcome? In a nutshell, it results from the inadequacy of the data and methodologies generally used to measure a company’s carbon emissions.
The carbon emissions of a company and its products fall into three categories. The first – known as Scope 1 – consists of direct emissions from activities that a company controls, such as the combustion of fuel on site, its vehicle fleet, air conditioning and so forth. Scope 2 consists of indirect emissions from energy purchased and used, while Scope 3 means all other indirect emissions coming from sources the company does not own or control. This latter category includes emissions produced in the company’s supply chain (upstream) as well as those arising from the use of its products (downstream).
Understandably, Scope 3 emissions are notoriously hard to measure. Many companies, both upstream and downstream, and many factors contribute to these carbon emissions, including the country in which operations take place, the production processes involved and proximity to raw materials. Furthermore, there is no internationally agreed standard for measuring these emissions.
It is therefore not surprising that most companies choose to limit their carbon data disclosures to Scopes 1 and 2. Accordingly, the providers of low-carbon benchmarks – which form the basis of low-carbon ETFs and index funds – use only Scope 1 and 2 emissions in their calculations.
The big problem with ignoring Scope 3 emissions is that, depending on the company, they could be multiples of Scope 1 and 2 emissions put together. For example, according to Morgan Stanley, the vast majority of an oil company’s emissions stem from the use of the hydrocarbon products it sells. MS estimates that for European oil majors, Scope 3 emissions are eight times larger than Scope 1 and 2 emissions combined.
By itself, the unavailability of Scope 3 data would be reason enough to question the validity of low-carbon portfolios. Unfortunately, there is an even bigger problem, related to the methodology used by low-carbon benchmark providers: they fail to consider the carbon emissions avoided through the use of a company’s products. This is the basis of the claim made at the start of this article, i.e. that low-carbon portfolios may end up excluding some of the most important providers of solutions to the emissions problem.
The most obvious example is clean energy.
Solar and wind energy are two of the most important ways of mitigating carbon emissions. However, making solar panels or wind turbines is not an emission-free process.
From a carbon point of view, the only way it makes sense to manufacture them is by considering the huge amount of emissions they avoid during their useful lives. If emissions avoidance is not considered, these companies will be perceived as high emitters.
Take the example of Xinyi Solar Holdings, a Chinese company that is among the world’s leading solar glass manufacturers. It also owns and operates solar farms. According to MSCI, its carbon emission intensity is 1,627 tons of CO2 per USD million of sales. The figure for Royal Dutch Shell is 218. So for investors trying to reduce the carbon footprint of their portfolios based on this data alone, it makes perfect sense to buy Shell and sell Xinyi Solar.
Except that of course carbon emissions produced by using Shell products are estimated to be around 8 times the figure considered by MSCI. And assuming that the 2.5GW of solar farms that Xinyi Solar operates replace coal power plants, this would avoid nearly 2.1m tonnes of CO2 emissions annually. That is, 2,143 tonnes avoided per USD million of sales every year (as a bonus, 62,000 tons of SO2 emissions would also be avoided). Thus, when product-related emissions are considered, Shell’s carbon intensity is estimated at around 1,962 versus negative 516 for Xinyi Solar. In other words, as common sense would suggest, Shell is a big source of emissions while Xinyi Solar offers a solution.
Xinyi Solar is not unique in this respect.
Global leaders in many sectors that play a key role in solving the world’s environmental problems – such as renewable energy equipment, waste management, water treatment, utilities and recycling – emit large amounts of carbon. On the other hand, many companies whose products damage the environment have low carbon emissions according to the measures used by the industry.
It is no wonder, then, that a large provider’s low-carbon ETF is full of names like Coca Cola, Philip Morris, Caterpillar, Valero Energy, Vale, Rio Tinto, Halliburton and Raytheon. In fact, the biggest clean energy company in the portfolio is Vestas, with a tiny weighting of 0.05% (probably because a large proportion of Vestas’ manufacturing is subcontracted, which means that its own emissions are low). If an investor chooses a low-carbon fund with the intention of helping the environment, this portfolio is unlikely to be fit for purpose.
This is probably less of an issue for institutional investors, which will look beyond carbon data alone when selecting funds. It is a bigger problem for retail investors, who lack the resources to do this. They rely on fund selectors to be gatekeepers. A worrying trend in this respect is for fund selectors to adopt cut-off points for portfolios’ carbon intensity, thus excluding funds with higher carbon intensity. For the reasons explained above, this could be doing a disservice to their clients.
Until better data collection, monitoring and measurement methods emerge, the finance industry should not just rely on reported carbon data. It should develop more sophisticated and holistic methods to measure the environmental impact of portfolios.
Portfolio Manager Emerging Equities