Short positions can be a meaningful add-on, but should not be an excuse for inaction on the long side of a portfolio. We believe the bar should be set much higher for long-short portfolios, to properly reflect the dual aim of divesting from long positions in heavy emitters and creating additional impact with short positions.
Whether or not short selling could be used to create net-zero portfolios is currently the subject of passionate debate. The issue is that the only way long-only investors can lower the carbon footprints of their portfolios is by selling their holdings in companies with high carbon footprints, either partly or entirely. But since all stocks have positive carbon footprints, it is impossible to reduce the carbon footprint of a long-only portfolio all the way to zero.
Although this might change in future if carbon removal technologies are successfully commercialized, for now it is the reality that investors face. With shorting, however, it becomes remarkably easy to achieve a zero or even negative portfolio carbon footprint: short positions correspond to negative economic ownership, implying that every short position in a stock with a positive carbon footprint lowers the overall carbon footprint of a portfolio.
In a recent Financial Times op-ed,1 the co-head of responsible investing at Man Group argues that counting shorts conflates two very different approaches to assessing climate change within portfolios, referred to in jargon as ‘financial materiality’ and ‘double materiality’.
The former focuses purely on the financial impact of climate change on a company, such as insurance or manufacturing losses stemming from an increase in extreme weather events, or a negative financial impact due to climate regulation. In this context, expressing portfolio exposure to climate change in terms of net long and short exposure is both viable and useful because it shows how economically exposed a portfolio is to transition risk.
With double materiality, fund managers try to understand the financial and non-financial portfolio effects of their investments. Here the author argues that focusing purely on the financial impact ignores the fundamental mission of the net-zero push to reduce real-world carbon emissions in the atmosphere.
Although shorting a high carbon-emitting company may provide financial de-risking benefits, it does not reduce the tons of CO2 released into the atmosphere. Hence, netting out short and long exposures does not alter the impact a portfolio has on the environment. Short selling is therefore not equivalent to offsetting, since a short position is neither a carbon offset nor a sink to store carbon.
This stance is countered by Cliff Asness,2 who starts by acknowledging that short positions should indeed not be regarded as a carbon offset. He adds that divesting from long positions in stocks with high footprints also does not magically remove CO2 from the atmosphere. But he argues that the mechanism through which financial markets ultimately have real-world impact is by affecting firms’ cost of capital. Companies only engage in projects with a positive net present value, and the higher the cost of capital, the fewer projects meet this hurdle.
If many investors sell their long positions in a certain stock, it tends to raise the firm’s cost of capital. This effect does not stop when the position size has reached zero. By further reducing the position in a stock, i.e. by going short, the cost of capital tends to be increased further. This view is supported by many academic studies on short selling. Also, corporate executives typically do not hide their displeasure at being targeted by short sellers, with Elon Musk being a well-known example of a CEO who has been particularly vocal on this matter.
We should first note that the entire argument for decarbonization via shorting critically hinges on the assumption that short positions reduce portfolio carbon footprints because they equate to negative economic ownership. However, carbon accounting is still an evolving practice, and there are varying opinions on this.
For instance, some argue that short positions and derivatives should be ignored when computing portfolio carbon footprints. But for the purposes of advancing this discussion, we will proceed to assume that longs and shorts both count. This also ensures that all portfolio positions are factored in and all resultant carbon footprints are accounted for, i.e. that no carbon emissions disappear into thin air.
That said, it is clear that neither selling long positions nor taking short positions physically removes CO2 from the atmosphere. We also agree with the notion that the real-world impact of these actions is achieved via the cost of capital channel.
Divestment by long investors and outright short positions taken by long-short investors tend to increase the cost of capital experienced by firms. So in that sense, both approaches are effective and mutually supportive. But if a portfolio technically establishes a carbon footprint of zero due to its short positions, should it automatically qualify as being compliant with a net-zero investment philosophy?
We argue not. To understand this, we begin by noting that the carbon footprint data is extremely skewed. A small number of firms have very high carbon footprints, while the vast majority of companies have very low carbon footprints that are almost negligible by comparison.
As a consequence, investors only require a handful of short positions in heavy emitters to drastically reduce the carbon footprint of their portfolio. Portfolio carbon footprints of zero or less are therefore well within reach through a few targeted short positions.
However, the problem with a portfolio that achieves net zero through shorting is that it can still contain many long positions in heavy emitters – provided these are offset by short positions in other, or even heavier emitters.
For instance, a long position in a thermal coal producer (firm A) combined with a short position in a similar thermal coal producer (firm B) may add up to zero emissions on balance from a carbon accounting perspective. But does the short position in firm B justify holding on to a long position in firm A? Is it not better to divest the long position in A regardless of whether B is shorted?
In our view, net-zero investing should always start with decarbonizing the long portfolio. The next step can then entail the addition of short positions in certain heavy emitters. However, these shorts should not be used as justification to retain long positions in other heavy emitters, even if technically their carbon footprints cancel each other out. In other words, short positions can be a meaningful add-on, but should not become an excuse for postponing action on the long side of a portfolio.
As implied above, the carbon footprints of long-only portfolios and long-short portfolios are not directly comparable. For long-only portfolios, carbon footprint reductions of up to 50% are generally attainable without too many complications, but more ambitious targets are progressively more challenging to achieve – see for example our work on decarbonizing the value factor.3 Moreover, zero carbon footprints for long-only portfolios are impossible as no firm is emissions free at present.
For long-short portfolios, on the other hand, establishing a zero footprint is almost trivial by comparison. Thus, the bar should be set much higher for such portfolios, to properly reflect the dual aim of divesting from long positions in heavy emitters and creating additional impact with short positions.
In other words, long-short portfolios should not be evaluated solely on bottom-line netted footprints, but the long and short legs of the portfolio should each be aligned with a net-zero ambition. This relates to the double materiality principle and is consistent with the stance taken by the Institutional Investors Group on Climate Change (IIGCC) in a recent discussion paper.4
In sum, long-only and long-short portfolios require distinct assessment frameworks that recognize their fundamentally different scopes and opportunities for decarbonization.
1 Mitchell, J., February 2022, “Short-selling does not count as a carbon offset”, Financial Times.
2 Asness, C., February 2022, “Shorting counts”, AQR article.
3 See for example: Blitz, D., and Hoogteijling, T., April 2022, “Carbon-tax-adjusted value”, Journal of Portfolio Management.
4 IGCC, May 2022, “Incorporating derivatives & hedge funds into the net zero investment framework”, IGCC discussion paper.