Scientific Beta: How low carbon strategies can be mistaken for alpha

16 juli 2021

Noël Amenc, CEO of Scientific Beta, and Felix Goltz, Research Director at Scientific Beta, discuss the findings and implications for investors of their recent research which analyses how low carbon strategies can be mistaken for alpha:

 
Claims that low carbon strategies deliver alpha are common in the industry. Your research comes up with new findings that question this idea. Can you explain your findings?

Felix Goltz: 'Actually, our findings are not at all new. There is a growing academic literature on the performance of low carbon strategies. Just looking at this literature already reveals that industry claims are not supported by the academic evidence.

First, we need an economic rationale for outperformance of low carbon strategies. Economic models predict that green firms should earn lower returns in the long-term than the high emitters. This is because some investors are willing to give up performance to receive non-pecuniary benefits of holding green firms that correspond to their social values. Investing in more sustainable firms should also allow investors to hedge against climate risks. Investors give up some of the expected return for this risk reduction, leading again to lower returns for green firms.

Empirical evidence also contradicts the notion of a positive alpha for holding green firms. There is a recent paper by Patrick Bolton from Columbia University and Marcin Kacperczyk from Imperial College that will appear in the prestigious Journal of Financial Economics. They analyse carbon emissions data and show that brown firms earn higher returns than green firms when using total emissions, while there are no significant return differences when using emissions standardised by revenues. This is exactly what we find. When sorting stocks on emissions standardised by revenue, we cannot detect any return effect beyond exposure to well-known factors. The same finding appears in other papers that use more involved definitions of carbon risk, such as combining emissions data and emission disclosure quality.

Overall, both economic principles and empirical evidence contradict the claim of positive alpha for low carbon strategies that is popular in the industry.'

 
How come providers still make these alpha claims and sometimes are able to back them up by numbers?

Noël Amenc: 'It is possible to show attractive performance numbers, especially when omitting standard risk adjustments or choosing the period well.

Over longer periods, we find that low carbon strategies show some outperformance when ignoring the factor exposures. This is especially due to a tilt to the high profitability factor. This performance benefit however is not specific to low carbon strategies - it just happens to be the case that low carbon strategies load on stocks with high exposure to the profitability factor. When analysing performance, investors need to adjust performance for such factor exposures.

Over shorter periods, low carbon strategies are influenced by changes in fossil fuel prices. We find that returns of low carbon strategies display a negative relation with fossil fuel prices. For example, one of the low carbon strategies we test yielded an annualised return of 9% during times of low oil returns, compared to an unconditional return of only 1.6% per year. Performance during periods of declining fossil fuel prices is thus inflated. As an aside, when promoters of low carbon strategies emphasise their positive performance, they should perhaps also point to the fact that this performance is driven by conditions which hamper the energy transition. Low oil prices are not only associated with high performance of green strategies but also mean low incentives for firms to move away from fossil fuels as an energy source.'

You also analyse portfolio construction methods used in practice and find substantial problems. Can you explain why commonly-used methodologies are not suitable?

Felix Goltz: 'The basic problem of such methodologies is that they use carbon scores as an alpha signal. This is quite convenient because the industry is used to building portfolios based on such signals. You just use a carbon score instead of a value or a momentum score. Everything else stays the same. This is quite convenient but of course it is ineffective when the score you use does not actually work as an alpha score, when it does not provide information on returns.

We find that common methods which use stock-level scores to weight stocks by carbon and factor metrics, while staying close to the cap-weighted index, do not deliver any performance benefits over simple filtering strategies that filter out the worst brown stocks. Despite this lack of performance benefits, these commonly-used methods lead to heavy concentration and face implementation challenges.'

Given these negative conclusions on alpha-seeking strategies, what should investors look at when considering low carbon investing?

Noël Amenc: 'They should consider objectives other than alpha.

In terms of pecuniary objectives, carbon emission metrics can be used as proxies for transition risk exposure. Investors concerned with transition risks may get hedging benefits out of such strategies. In terms of non-pecuniary objectives, many investors emphasise that they want to have an impact on corporate decisions so that firms will emit less carbon.

Addressing these objectives will not come as a by-product of pursuing alpha. It requires careful analysis, dedicated portfolio construction, and further research.

In the end, the pressing issue faced by society and investors is tackling climate change and managing its risks, not generating alpha. And while low carbon alpha appears to be fake, the risks of climate change are real.'