Abstract Climate change is likely to impact financial stability through different channels. We aim at assessing if those channels are already operating, and how they may operate in the near future. The data and assessments we provide would have to be refined and expanded in order to assess causal relationships, but we nevertheless give a broad picture of the issues at stake. We first focus on physical risks. Climate change, for instance is expected to translate into an increased number of natural catastrophes and weather events that will, if on a sufficient scale, translate into financial instability. We focus on the case of hurricanes in the USA and their impact. We then focus on transition risks. Transition risks should happen when environmental transition policies are implemented. If such a transition is too fast or too abrupt, it may trigger financial instability. Such risks may also materialise when policy announcements are made, since financial agents react to collective anticipations. We investigate the rationale behind this risk, and show the impact on average stock value of big polluting firms of the Paris Agreement. We then explain why climate-related financial risks will remain difficult to assess due to the intrinsic uncertainties of all climate phenomenons. This leads us to the conclusion that climate change might be considered as a “systemic risk” for the financial system. We conclude our analysis with a review of potential public policies that a central bank might want to undertake. While a “green quantitative easing” would imply a change in central banks’ mandate, we argue that climate stress-tests and disclosure of risks should become increasingly important in the future.
"J’ai compris lentement la raison profonde de cette impuissance : on veut décrire un drame qui n’a pas existé." Antoine de St-Exupéry, Terre des hommes
On November 8 2018, Benoit Coeuré, Member of the Executive Board of the European Central Bank, addressed for the first time the responsibility of the ECB concerning climate change. He concluded his speech by saying: “The best the ECB can do is to concentrate its efforts on creating the right conditions for supporting the flow of capital into sustainable sectors. This way, it will remain faithful to its primary objective of price stability (...), while supporting environmental objectives”1. This seems to be a turning point, with a clear recognition of the link between climate events and monetary policy. Benoit Coeuré is thus following Mark Carney, Governor of the Bank of England, who explained in 2015 the challenges posed by climate change for central bankers. Businesses’, politicians’ and administrative authorities’ short-termism leaves the climate ticking bomb to the next generation. This is what Mark Carney calls “The Tragedy of the Horizon”2. Nonetheless, this does not come without risks, which have yet to be identified. Therefore, regarding central banking, the first question to ask is whether we have enough theoretical foundations and empirical evidence that climate change might undermine financial stability. If yes, what is the responsibility of a central bank? Is it possible and legitimate to change market incentives for a more sustainable financial system? In our analysis, we will mainly focus on physical and transition risks linked to climate change, excluding other types of risks from our analysis (such as “liability risks”, that is “the impact of lawsuits by the ones who suffered from natural catastrophes, aimed at those who are considered responsible for these events”3). We will conclude our analysis by listing the tools a central bank could use to mitigate those risks.
The insurance sector and climate change What are the “physical risks” posed by climate change to the financial system? Physical risks can be defined as “those risks that arise from the interaction of climate-related hazards (including hazardous events and trends) with the vulnerability to exposure of human and natural systems, including their ability to adapt”. (Batten et al., 2016).
There is growing evidence that climate change is likely to have an impact on meteorological events, such as storms, floods, droughts, forest fires and snow. In 2014, the IPCC estimated that “freshwater-related risks of climate change increase significantly with increasing greenhouse gas concentrations (robust evidence, high agreement) (...). The fraction [of the global population] affected by major river floods increases with the level of warming in the 21st century”. In Europe, there is “high confidence” that there will be “increased economic losses and people affected by flooding in river basins and costs, driven by (...) coastal erosion and peak river discharges”4. Concerning hurricanes, in 2012, the IPCC stated that “it is likely that there has been a poleward shift in the main Northern and Southern Hemisphere extratropical storm tracks”. Eventually, “drought coupled with extreme heat and low humidity can increase the risk of wildfire”. Hence, there is a “high confidence” of “wild-fire induced loss of ecosystem integrity, property loss, human morbidity and mortality as a result of increased drying trend and temperature trend” in North America5. Therefore, a quite substantial number of environmental indicators are expected to change. With the rising number of climate hazards, we might also expect the number of economic losses to increase.
Torsten Jeworrek, a Munich Re6 board member, explains: “We don’t discuss the question anymore of ‘Is there climate change’ (...) For us, it’s a question now for our own underwriting”7. Indeed, after hurricane Andrew hit the United States in 1992, generating $15.5 billion of insured losses, 13 insurance companies went bankrupt.
Source: NOAA National Centers for Environmental Information
What are the challenges related to an increasing number of insured losses? The number of insured losses is rising over time, with $135bn in 2017 (Munich Re), a record-breaking year. This can be notably explained by hurricanes Harvey, Irma and Maria. Torsten Jeworrek, of Munich Re board, adds: “even though individual events cannot be directly traced to climate change, our experts expect such extreme weather to occur more often in the future”.8
Source: Munich Re
If losses are insured, climate change could potentially affect the balance sheet of insurance companies.
The International Association of Insurance Supervisors (IAIS, 2018) identifies some potential effects of physical risks on underwriting activities, such as: “pricing risks arising from changing risk profiles to insured assets and property (non-life), changing mortality profiles (life and health)”, “claims risk arising from confluence of unexpected extreme events”. The IAIS highlights that: “Lloyd’s market reports to have paid out $5.8bn in major claims, most of which were climate-related. The claims burden disasters in 2017 and has have strong financial impact for non-life insurers, with industry Return on Equity dropping from 11% in 2016 to -4% in 2017”.
There are also risks concerning investment activities due to “impacts of physical climate events and trends on assets, firms and sectors, affecting profitability and cost of business, leading to impacts on financial assets and portfolios”. The capacity to pay back future claims would be diminished, whenever investments of insurers become less profitable.
Moreover, if insurance premiums are above consumers’ willingness to pay, physical risks might lead to market contraction. The US total gross insurance premium went from $0.30 million in 1985 to $2.62 million in 2015 (OECD)9. We could make the hypothesis that this trend is partly driven by the increasing number of losses. If some consumers cannot afford increasing insurance premiums, the number of customers would decrease. As Thomas Buberl, CEO of AXA, pointed out during the One Planet Summit: “more than four degrees Celsius of warming this century would make the world uninsurable”10. Some areas might thus not be covered by the insurance market. For instance, “in the US National Flood Insurance Program alone, over $600bn of property within one mile of the beach is covered, much of it will not be viable in coming decades”11. There are historical examples where coverage decreased after some climate events. For instance, after Hurricane Iniki, “the Hawaiian Insurance Group ceased trading and announced the non-renovation of existing policies, which led to a “climatic domino effect” where other insurers felt obligated to withdraw from the Pacific and Caribbean island nations” (Bank of England, 2015).
Physical risks are a major concern for the balance sheets of insurers, but for the balance sheets of banks, as well. They might lead to “disruptions to critical insurance services and systemically important financial markets, such as securities lending and funding transactions. Large-scale fire sales of assets by distressed insurers could reduce asset prices which would affect the balance sheets of other financial institutions, such as banks.” (Batten et al., 2016). The channel from insurance companies to other financial institutions can work through different mechanisms: funding transactions (if insurers hold large shares of debt securities issued by financial institutions), securities lending activities and reinsurance contracts (if a reinsurance company is in default). What does the “insurance protection gap” means for the financial system? If we also include uninsured losses, the amount of weather-related losses reaches $330bn in 2017. Therefore, beyond threats regarding the insurance sector, there is also a concern with “the insurance protection gap”.
Source: Munich Re
In 2016, ClimateWise estimated at $100bn the gap between “total economic loss and the value of assets that were covered by insurance policies”. Nonetheless, this gap has quadrupled over the last three decades (from $23bn). What’s more, Swiss Re estimated that “30 percent of catastrophe losses have been covered by insurance in the 10 years prior to 2014. Therefore, the gap is partly falling on governments’ shoulders, who are acting as the “insurer of last resort” (Climate Wise, 2016). Hence, according to the American Federal Emergency Management Assistance (FEMA), the National Flood Insurance Program has more and more trouble covering the losses induced by a greater severity of hurricanes or flooding in the past 10 years. The inability of this public insurance program to address uninsurable areas’ needs has created a large public deficit. In October 2017, Donald Trump signed a bill forgiving $16 billion, out of $25 billion debt (before servicing)12. A 2017 report from the American Congress established that: “CBO’s estimate of expected claims exceed FEMA’s estimate by $1.0 billion. Because of FEMA’s estimate is its basis for premium setting, the difference between the two estimates causes premiums to fall $1.0bn short of CBO’s estimate of expected claims”13. Therefore, physical risks might affect the credit ratings of sovereigns or municipalities. For instance, due to cyclones, “government finances would deteriorate following the cyclone, leading to an increase in government debt in the most affected sovereigns from 2016 to 2020 by between 7% in Taiwan and a maximum of 18% of GDP for the Dominican Republic, compared with the non-cyclone simulation” (Standard & Poor’s, 2015). However, physical risks will fall disproportionately on some regions, especially in the developing countries, that is, the ones who are the least insured. According to Standard & Poor’s, there is already a ratings impact of 0.5 up to 2.5 notches for some countries such as Vietnam, Bangladesh or Dominican Republic, concerning the risk of cyclones.
Source: Standard & Poor’s
Therefore, physical risks could be transmitted to the financial system through credit risks, for sovereigns, businesses or individual agents. In this case, when goods and property are used as a collateral, or if climate events are linked to decreases in revenues, this could threaten the financial situation of some financial institutions. Physical risks would both increase the probability of default, and create a higher loss given default. Indeed, it is mainly the “uninsured losses that drive the subsequent macroeconomic cost, whereas sufficiently insured events are inconsequential in terms of foregone output” (Von Peter, van Dahlen and Saxena, 2012). These authors find that “a typical median catastrophe causes a drop in growth of 0.6-0.1% on impact and results in a cumulative output loss of two to three times this magnitude (...) well insured catastrophes, by contrast, can be inconsequential or positive for growth over the medium term as insurance payouts help fund reconstruction efforts”. Klomp (2014) also shows that out of 160 countries in the period 1997-2010, meteorological and geophysical disasters increase the likelihood of bank default: “Geophysical and meteorological disasters reduce the distance-to-default the most due to their widespread damage caused”.
In addition to a rising probability of default, a natural disaster can trigger a reduction in lending activities. Lambert, Noth and Schüwer (2014) find that the Hurricane Katrina in 2005 “increased their risk-based capital ratios after the hurricane”. Indeed, “high-capitalized banks react to the lower asset quality through the hurricane by shifting investments from commercial and industrial loans to US government securities and thereby increase their risk-based capital ratios”. However, this decrease in bank lending could be associated with a fall in output, which might itself induce losses for banks. A credit contraction, at a time when it is necessary to rebuild a region would also weaken the balance sheets of households and corporates. A reduced insurance coverage would amplify this phenomenon, since collateral values would fall. Therefore, economic agents would have an even tighter access to credit.
What are the theoretical channels explaining how physical risks would result in economic losses for financial institutions? The Bank of England clarified all the ways through which a natural disaster could induce both losses for insurers and banks (see figure below). The first major transmission channel will depend on the resilience of insurers. If market agents misconceived the probability of a catastrophe to happen, we might expect a fall in asset prices. Nonetheless, uninsured losses are also a concern for the financial sector, with the reduction in creditworthiness of some economic agents.
We must stress that it does not necessarily need a catastrophe to happen in a given region, for a financial crisis to occur. Market sentiment might play a substantial role, even when a natural disaster is not happening directly. For instance, there can be a re-evaluation of risks linked to a specific sector, or region. This might generate “high volatility and trigger an abrupt decline in the value of securities held by the banking sector”. In extreme cases, “if investors reassess a bank’s balance sheet quality with respect to physical risks, they may decide to no longer refinance that bank” (Trésor - French Treasury, 2015).
Source: Bank of England
Physical risks might potentially lead to vicious circles, due to the “financial accelerator”. Indeed, Nobuhiro Kiyotaki and John Moore (Kiyotaki, Moore, 1997) explained how an exogenous shock aggravated by credit restrictions would amplify and lengthen business cycles, giving rise to large output fluctuations. The model assumes two types of identical agents: farmers (firms or borrowers) and gatherers (households or lenders). Farmers are good at producing but have a fixed quantity of land at time t, gatherers are less good at producing and possess land. Farmers consume an incompressible share of their production and sell the rest. In order to maximize profit, farmers want to produce more. Thus they want to borrow land from gatherers. A key feature of the model is that lending is collateralized in this economy, i.e. backed with an asset of the same value of the lending, that will be seized by the lender in case of default. Assuming everything else constant, if the price of land is rising, the same amount of land will be more expensive in t +1 compared to t. So the farmer can borrow a lot since his collateral is highly priced. The key point lies in the fact that land is both a collateral and a production factor. As a farmer borrows land, he increases both his production and borrowing capacity. There appears a leverage effect. The increase in production he gets through the increase in capital (namely, land) will in turn increase his net worth, and if prices of land keep increasing, his collateral values too. In t +1 he will have no problems borrowing more (as his borrowing constraint loosened). Eventually, his productive capacity will have increased a lot. This virtuous circle can also happen in the reverse way. Assuming everything else being constant, if estate prices are decreasing, a farmer will not be able to borrow and will even have to sell part of his land in order to reimburse previous debt. Production capacity decreases and so does net worth as well as the value of collateral, and the story repeats itself. We note that the price of land may not be already increasing or decreasing for such dynamics to start. This model enables us to understand how a natural catastrophe (through a lowered value of physical assets) would lead to credit contractions and output fall. This in turn, would weaken households and corporate balance sheets, leading to losses for banks.
Source: Kiyotaki, Moore (1997)
The analysis of natural catastrophes on stock prices: We look at the evolution of financial quotation from regional insurers (and other companies) surrounding the happening of hurricane Harvey in 201714. These preliminary results must be further documented, nonetheless we do see a decrease in the stock exchange values of the main insurers involved in Texas. We could think this drop in value is due to a change in investors’ expectations. Indeed, insurers’ losses were documented in the media (such as the Financial Times), as well as risks regarding airline companies or major Texan firms (such as ExxonMobil), late August this same year15. We find again evidence, suggesting a direct harm to companies we selected. Therefore, further research is needed to know if this downward trend in stock prices is driven by hurricane Harvey. If this is proven to be true, it would mean that physical risks could potentially trigger fall in asset prices, leading to losses for financial institutions or private investors.
Source: Yahoo Finance
(To construct this graph, we looked at the evolution of stocks prices for the following companies: Delta Airlines, United Continental Holdings, Jetblue Airways Corp., WestJet Airlines, Spirit Airlines Inc., Group Aeromexico SAB, ,Regions Financial Corp., BB&T Corp., Zions Bancorp., Allstate Corp., CNA Financial Corp., American International Group Inc., Chubb Ltd., ExxonMobil, Valero, The Progressive Corporation, The Hartford, HCI, Universal Insurance, Heritage Insurance.) We now look at the stock prices of three major Californian home insurance companies (All State, State Farm and Farmers Trading Company Ltd), in the aftermath of the California wildfires which happened in November 2018. We find a downward trend, even though we cannot infer any causal relationships from this result. We see a somewhat less clear pattern.
Source: Yahoo Finance
As said earlier, risks will not be distributed in the same way all around the world. The French Directorate General of the Treasury together with the Prudential Supervisory and Resolution Authority identified that “sectors potentially exposed to physical risks represent 56% of total ‘corporate exposure’. The highest exposure appears in the real estate and construction sectors (22.3%), followed by transport (10%) and wholesale trade (9.7%)”. For the French banking sector, the risks seem to be relatively low (Trésor – French Treasury, 2015). The rise of “catastrophe bonds”: A worth noting market reaction to physical risks is the replacement of insurance companies by catastrophe bonds. They are an “example of insurance securitization to create risk-linked securities which transfer a specific set of risks from an issuer to investors. In this way, investors take on the risks of a specified catastrophe or events occurring in return for attractive rates of investment. Should a catastrophe or event occur, the investor will lose the principal they invested and the issuer will receive that money to cover their losses”16. When a catastrophe occurs, the number of claims sharply increases in a short period of time. Insurance companies usually use third parties, namely reinsurance companies, to spread their risks. However, in that case, bonds replace insurance contracts. Their attractiveness is explained by a high rate of return and their relative disconnection from other macroeconomic variables. Nonetheless, this comes at a price, investors cannot be sure to preserve their principal. On the other side, this reduces the cost of raising capital, in the event of a huge financial loss. Catastrophe bonds could reduce the rise of premiums or the limitation of coverage. Nonetheless, the increasing number of such risky financial assets also show that traditional instruments to hedge against risks might become less relevant in +2°C world.
“Transition risks”, how public announcements related to climate change influence financial markets What do we actually mean by “transition risks” and how could some assets become “stranded”? Transition risks can be defined “as the risks of economic dislocation and financial losses associated with the transition to a lower-carbon economy” (Batten et al., 2016). There is a possibility of a quite substantial loss in the value of assets related to carbon-intensive environments. The hypothesis of a carbon bubble states that some financial assets might be overvalued. Indeed, the Carbon Tracker Initiative estimates that to “reduce the chance of exceeding 2°C warming to 20%”, the global carbon budget for 2000-2050 is around 880 GtCO2 (Carbon Tracker Initiative, 2011). Nonetheless, the total carbon potential of earth’s known fossil fuel reserves is estimated at 2795 GtCO2, that is five times the budget planned for the next 40 years. According to the IEA, the carbon budget would be reached in 16 years’ time, if consumption does not slow down.
Source: Carbon Tracker Initiative
Moreover, the top 100 oil and gas companies can potentially emit 745 GtCO2. “his exceeds the remaining carbon budget (...) by 180 GtCO2”. Hence, “using just the listed proportion of reserves is enough to take us beyond 2°C of global warming”. If we want to still stick to that limit, considering reserves held by governments, “only 149 of the 745GtCO2” can be used. This leads to the conclusion that there is a possibility of some assets of these companies to become “stranded”, if carbon becomes ‘unburnable’. The report from the Carbon Tracker Initiative further explains: “up to 80% of declared reserves owned by the world’s largest listed coal, oil and gas companies and their investors would be subject to impairment”. A carbon budget depends on analytical assumptions. The future amount of stranded assets is not fully certain.
Source: Carbon Tracker Initiative
However, even with full investment in CCS (carbon capture and storage technology), a carbon budget would not increase by far more than 12-14%, since it is a new technology which has yet to be developed. In any case, only 20 to 40% of current proven reserves could be used. What does that mean for the financial sector? The Carbon Tracker Initiative quotes a report from McKinsey and Carbon Trust demonstrating that 50% of the value of oil and gas companies “resides in the value of cash flows to be generated in year 11 onwards”. Market agents seem to react to a company’s future reserves. For instance, when in 2004, Shell announced that its level of reserves would be 20% lower, its share price decreased by 10% in a week (resulting in a loss of £3 billion).
Source: Carbon Tracker Initiative
The energy transition impact on the expected loss of carbon-intensive companies will depend on the policy scenario chosen by governments. A study from Kepler-Cheuvreux (Lewis, 2014) takes the IEA scenario (“450 Scenario”) consistent with a + two degree world, and estimates that “fossil-fuel industry stand to lose USD28trn (in constant 2012 US dollars) of gross revenues over the next two decades, compared with business as usual (...) the revenues most at risk would be concentrated in the high-cost, high-carbon sources of production [namely] deepwater, oil-sands and shale-oil plays”. While the Current Policies Scenario only considers “policy firmly enshrined in legislation”, the 450-Scenario is more ambitious. The 450-S also implies lower volumes in order to reach carbon neutrality by 205017.
Source : Kepler Cheuvreux
Compared to this baseline scenario, fossil fuel prices are projected to be lower, if countries implement more stringent regulations in order to diminish their carbon emissions. This will reduce the profitability of heavy projects with a high cash cost per barrel.
Source : Kepler Cheuvreux
Lower demand accompanied by lower price explains why fossil-fuel companies could face substantial financial losses. The global market capitalisation of fossil-fuel companies was a $5trn in 2014. HSBC (Robins, 2012) looked at the effect of US$50 price for oil and gas barrel, and estimated that the impact on equity valuations is “in the range of 40-60% for most companies”. However, not only fossil fuel industries would be impacted by stranded assets, but also other economic agents using fossil fuels such as car producers. A sudden shift to more pro-environmental policies would trigger a stronger repricing of carbon-intensive assets such as fossil fuel reserves, but also of all assets related to the use of fossil fuel. (European Systemic Risk Board, 2015). A late transition would not be without risks for the financial system. According to the European Systemic Risk Board, “fossil-fuel firm and electricity utilities are substantially debt financed, exacerbating the potential financial stability impact of a sudden revaluation of stranded assets”.
Source: Batten et al. (2016)
In 2015, fossil-fuel and carbon-intensive companies represented one third of the $2.6trn leveraged loan market. In the European Union, total exposure of the financial system (banks, hedge funds, insurers) exceeds €1trn, and losses would be ranging from €350 to €400bn under “an orderly transition scenario” (Weyzig, 2014). The first-order impact seems relatively small, relative to the overall size of the financial sector, but feedback loops are a source of concern. At a global level, losses could be more or less substantial, depending on substantial technology diffusion (with higher energy efficiency and lower fossil fuels demand) or on a steady continuation of investment in fossil fuels. Under the first assumption, a two-degree scenario means that the discounted cumulative fossil fuel value loss to 2035 could reach US$4 trillion (with a 10% discount rate). If instead, there is some technology diffusion, the potential loss decreases to US$3 trillion. We can identify two channels which could threaten financial stability: “wealth losses and value of fossil fuels companies” and “macroeconomic change (GDP and structural change)”. OPEC countries, Russia and the US would be the most heavily impacted by the economic losses due to stranded assets (Mercure et al., 2018).
Source: Mercure et al. (2018)
The possibility of a late transition: One can explain the transition risks using game theory. Carbon-intensive companies will invest in CCS or other technologies reducing carbon emissions, if and only if they believe that at time t +1, government will “shut down unabated electricity production”. Then, their investment will be worthwhile. However, if the government does not decide to invest in the ecological transition at time t, there might be a “high carbon equilibrium” where, it will be anyway too late to invest in carbon-reducing technologies, since the costs would be too high.
Source: Batten et al. (2016)
Nonetheless, one cannot rule out a sudden change in governments’ policies. If a government was to force a sudden reduction in the amount of carbon emissions, then the value of a firm could decrease quite dramatically. In that case, one ends up in a bad equilibrium where “some fossil fuels and unabated power plants become stranded”.
Source: Batten et al. (2016)
This late transition would in turn trigger financial instability with repricing of financial assets. As said earlier, major carbon-producing companies rely on debt to finance themselves. The analysis of COP 21 and the “One Planet Summit” on stock prices
In order to construct the following graph, we took the average of all stocks from the below-mentioned companies (which gather car-producing companies, as well as some of the top 100 most polluting firms listed in the Carbon report18). For non-US-based companies, we converted stock prices by using the yearly exchange rate. Then, we plotted values for each day, and observed that indeed, the value of stock prices for these major companies is, on average, decreasing19.
Source: Yahoo Finance
(In order to construct this graph, we looked at the market valuation of the following companies: Total, Shell, BP, PetroChina, Chevron, Eni, Gazprom, Valero, Exxon Mobil, Brasileiro, Bhp Limited, Lukoil, Rio Tinto, Anglo American, Surgutneftegas, Statoil, Conoco Philips, Glencore, Rosnet OAO, Anadarko, Repsol, Sasol, RWE). If we compare our first estimation with the evolution of prices with three major companies involved in the solar energy (7C Solar Paken, Atlantica Yield, Canadian Solar), we observe, on the other hand, that stock prices increased quite substantially. These first striking results demonstrate for us the need to further investigate these findings, with an expanded database, and controls for sample heterogeneity.
Source: Yahoo Finance
The One Planet Summit  was co-organized by the French government, as well as the World Bank and the United Nations. It has been officially supported by 89 top French companies, which released public statements of their strategic action to tackle climate change. Our hypothesis is that these announcements worked as a temporary “signalling effect” during that time. Indeed, if we take the average of stock prices at that time, we can see a local pinnacle on December 12th 2017. Nonetheless, the overall impact, if ever there is one, seems relatively limited.
Source: Yahoo Finance
Are investors already taking into account transition risks? Vikash Ramiah, Belinda Martin, and Imad Moosa (Ramiah, Martin, Moosa, 2013) looked at the impact of 19 announcements of environmental regulation on the Australian Stock Exchange equities over the period 2005-2011. They see an “abnormal return of -31%” in the energy sector after Australia submitted its target range to the Copenhagen Accord. They point out that “60% of the Australian stock market was influenced by these policies”. It is primarily the oil and gas, real estate and the general industrial sectors which experienced abnormal negative returns, especially the electricity companies which saw a -2.84% abnormal return. The authors understand “abnormal return”, using a capital asset pricing model. Nonetheless, this enters in contradiction with the analysis of COP 21, conducted by the Bank of England (Batten et al., 2016). The authors computed “cumulative abnormal returns experienced by a petroleum refining company (CVR Energy) and a wind turbine manufacturer (Nordex)” in the aftermath of the Paris Agreement. They find negative, but insignificant results, concerning CVR Energy. However, the return on this specific renewable energy company is both positive and significant.
Source: Batten et al. (2016)
Corroborating this finding, the publication of a paper21 in the Nature journal of science highlighting that only part of oil, coal and gas reserves could be emitted not to go beyond the 2°C threshold, triggered little market reactions. This took place, despite the fact that this article is considered as “one of the most cited environmental science studies in recent years” (top 0.1% of science papers in 2009). Researchers find an “average price drop of 1.5% to 2%” in their sample of the 63 largest US oil and gas firms in the two weeks after publication, that is an aggregate loss of $16.5 billion in price drop. Nonetheless, they find no impact on stock prices (in 2012-2013), due to news related to the carbon bubble (Griffin et al., 2015). Why don’t we see a sudden movement in stock prices? It might be because investors are expecting some public compensations for their stranded assets. Another study, based on a German climate policy case, suggests that investors do care about stranded asset risk (Sen, von Schickfus, 2017). This policy intended to target lignite assets, and the authors find that “investors did not react to the announcements of the initial (...) extra fee on carbon emissions”. “Only announcements that the compensation mechanism may not go through due to violating state aid rules resulted in a significant and negative reaction”. Indeed, investors did not react to the first proposal of taxation, but only to the news that there will be no compensation mechanism from the German state : “only on the publication dates of the media reports of the assessment do we observe a significant reaction” on the three main German utility companies (RWE, E.ON, EnBW). The authors conclude that investors “do price in the stranded asset risk, but with an expectation of compensation”. Therefore, it implies that a clear strategy to fight climate change, with a credible announcement that carbon-intensive companies will not be compensated, would incentivize investors to consider transition risks in their valuation procedures. This limited reaction might also be explained by the fact that investors actually integrated CCS. It has been shown that there has been “significant positive reactions to CCS breakthroughs” between 2011 and 2015 (Byrd et al., 2016). Other explanations would state the fact investors are skeptical about the demand for oil and gas to decrease in the following years. (Griffin et al., 2015). Moreover, we cannot rule out the fact that investors lack the appropriate information to value some firms. Indeed, Griffin and his co-authors find “no mention of unburnable carbon or an equivalent phrase” in the 10-K filings of top US carbon-emitting companies, which are required by the Security and Exchange Commission. Nonetheless, there is also growing evidence that “green” assets are more highly valued (Baker et al., 2018). The authors find that “green bonds will sell for a premium”. Hence, “after controlling for numerous fixed and time-varying factors, [they] find that green bonds indeed are issued at a premium, with yields lower by several basis points”. Indeed, “overall average after-tax yields are somewhat lower for green bonds than ordinary bonds, at 2.28% versus 2.50%”. Green bond are priced as if they were ‘half a notch’ more highly rated”. Concerning the CBI certified bonds, authors demonstrate that they have “yields 26 basis points lower than ordinary bonds with similar characteristics and timing”. Even with the addition of several controls and interactions, “the average difference between the after-tax issue yield on CBI certified green bonds and ordinary bonds amounts to 15 basis points per year”. This is why, even if there is a cost to pay for a bond to be certified, the premium that an issuer can obtain could be a substantial incentive for divestment. They also find that “green bond ownership is more concentrated, with a subset of investors holding them at higher weights, particularly when the par value is small or the bond is especially low risk”. The fact that both pricing and ownership effects are stronger for certified green bonds, emphasized the need for a reliable taxonomy (see section IV).
The cost of inaction: The cost of inaction can be identified through the “climate value-at-risk” (Dietz et al., 2016). A value at risk can be defined as: “the size of loss on a portfolio of assets over a given time horizon, at given probability”. This estimate can also be seen as “a measure of the potential for asset-price corrections due to climate change”. The authors use an aggregated Integrated Assessment Models, in order to get the “probability distribution of the present market value of losses on global financial assets due to climate change”. For this, they consider the present value of future interest payments (which is a discounted cash flow). The climate value at risk is then estimated, taking into account GDP projections depending on climate change (since corporate earnings represent a given share of GDP and since part of these earnings ultimately benefit to the ones who own financial liabilities), the stock of global financial assets, and a discount rate. They use the Dynamic Integrated model of Climate and the Economy (DICE), developed by Nordhaus. It is a neoclassical growth model, where individuals trade off their consumption today with their consumption in the future. The DICE model includes natural capital as an additional kind of capital stock. This model might include a constraint over a certain limit in temperature rise22. The authors explain that their Integrated Assessment Model leads them to estimate a climate value at risk of 1.8%, representing an amount of $2.5 trillion. This result comes from a scenario where the increase in global mean temperature in 2100 is 2.5°C above pre-industrial levels. They also see that risks are heavily concentrated: “the 95th percentile is 4.8% and the 99th percentile climate VaR is 16.9%”. The most interesting element in this analysis is that results differ depending on the scenario chosen. If we want to limit warming to no more than 2°C with a probability of ? , then “the expected VaR is [only] 1.2%, the 95th percentile is 2.9% and the 99th percentile is 9.2%”. The estimated present value of a loss to 2100 (using the current sock of global financial assets) is about $2.5 trillion (but $24.2 trillion at the 99th percentile). It can be reduced to $1.7 trillion ($13.2 trillion at the 99th percentile), under a two-degree scenario. However, according to these authors, most of the costs will arise in the second half of the century. Depending on the discount rate that is chosen, losses can vary quite substantially. In the event of a 5°C warming, the present value of potential losses would reach US$7trn to 2100 (and $13.8trn with 6°C warming), that is “more than the total market capitalization of the London Stock Exchange”23. However, if the authors choose the same discount rates as the Stern Review, the present value of losses jumps to US$43trn, with a 6°C degrees warming24.
Therefore, under a pure cost-benefit analysis, it makes sense to mitigate the rise in global temperatures, since the latest this transition happens, the higher the transition risks (and so the costs for the economy) will be.
Fat-tail probability distributions, radical uncertainty and the hypothesis of a climate-related systemic risk The essential problem with climate change is the radical uncertainty of its consequences. Weitzman argues that we have to take into account the possibility of extreme values, due to probability distributions of climate events. The probability that temperatures will rise dramatically does not follow a normal distribution, but rather a fat-tail distribution, with still a substantial likelihood that one observes a radical increase in temperatures. “What is especially striking to me is the reactiveness of high-temperature probability to the level of GHGs.” He further emphasizes: “Deep structural uncertainty about the unknown unknowns of what might go wrong is coupled with essential unlimited downside liability on possible planetary damages”.
Source: Weitzmann (2011)
Weitzmann identifies a long chain of structural uncertainties, among which: “unknown base-case GHG emissions”, “how available policies and policy levers will affect actual GHG emissions”, “how and when GHG stock concentrations translate into global average temperature”, “how global average temperature decompose into specific changes in regional weather patterns”, “what discount rate should be used” and which utility function to apply (Weitzmann, 2011). Indeed, beyond the uncertainty about the political cycle, there is also a question of economic modelling which remains quite difficult to solve. Nordhaus explained: “there has been virtually no work applying Weitzman’s insights in empirical IAMs (...) the question is particularly demanding because it requires estimating the shape of the tails of distributions for events, such as damages to future consumption, where there is very sparse experience on which to estimate the distribution” (Nordhaus, 2013). The fat-tail distribution implies nonetheless that market agents cannot rely on historical data to measure their future risks. Therefore, the question of modelling dynamic interactions becomes particularly important. There is a need to develop complex and adaptive systems, as climate is itself a complex system. In mathematics, this is defined as a system with multiple interactions, and characterized by some features such as: nonlinearity, spontaneous order (with no leader but only local interactions), or feedback loops. In order to forecast the future impact of climate-related risks on the financial system, it becomes necessary to develop models with “network effects that stem from the interactions between agents” and where “disequilibrium phenomena play a key role”. For now, models used in central banks or private entities, such as dynamic stochastic general equilibrium (DSGE) do not take integrate climate change or environmental policies (Campiglio, 2018). This is why Michel Aglietta and Etienne Espagne (Aglietta, Espagne, 2016) argue that the “systemic risk approach arises in concrete economies with incomplete and imperfect markets”. Indeed, “the combination of increasing uncertainty on critical climate parameters with the complex political economy of climate policy action cannot be apprehended in a satisfying manner through this traditional first/second-best policy nexus”. Therefore, what a climate-related systemic risk means for financial markets is that it could result in “a potential financial turmoil and this in turn can increase the tension around the provision of the ultimate liquidity”. They also identify another source of concern, which is that “the power of ultimate liquidity to restore confidence into the payment system can potentially be put into question”, since the risks are not confined to the macroeconomic environment, but also affect the “fundamental support of life”.
Source: Aglietta, Espagne (2011)
Moreover, when we depart from the hypothesis of static Arrow-Debreu economy with complete financial markets, uncertainty on the value of assets plays a substantial role. For instance, we might not rule out the hypothesis of “sunspots” on financial markets, where the variation of stock prices is disconnected from fundamentals, and only rely upon market sentiment (Cass, Shell, 1983). A report from the Inter-American Development Bank (Caldecott et al., 2016) shows that the financial system did not take well into account climate risks in their strategy. “There is a lack of knowledge of climate issues, particularly in the mainstream investment industry, and a need for easier-to-digest information to assist in the decision-making process.” There is also an heterogeneity of awareness: “in Latin American countries, it appears to be a clear gap (...) there is a significant omission, given the region’s exposure to environment-related risk factors, [and] the presence of extensive fuel resources that may become “unburnable” given carbon budget constraints”.
Which financial regulation for climate change? The debated proposition of a “green quantitative easing” The supporters of a European finance pact advocate for a monetary policy which would encompass sustainability goals. Recent research highlights that there is indeed a climate impact of quantitative easing. Theory would predict a priori that the effect of QE is neutral and would stimulate the economy as a whole. The law of supply and demand implies that investors would sell assets in high demand to buy other ones, and so asset prices would increase overall. In fact, transmission channels are not perfect and the ECB asset purchasing program apparently is not neutral. To be eligible to the corporate bond purchase program (CSPP) of the ECB, a corporate bond must be denominated in euros, rated “investment-grade” and with a maturity of between 6 months and 30 years. Hence “renewable energy companies, already at a small portion of the bond market to begin with, are not represented at all in ECB purchases, while oil and gas companies make up an estimated 8.4 per cent of its portfolio”. Moreover, the CSPP market does not represent the real economy. Hence, “the sectoral contribution of purchases appears to be inconsistent with the sectoral distribution of the euro-area economy in terms of contribution to gross value added, and skewed towards sectors characterised by high greenhouse gas emissions” (Matikainen, 2017).
Source: Matikainen et al. (2013)
What does that mean? The ECB could ease financial conditions more favourably for carbon-intensive companies. For instance, Henkel and Sanofi demonstrated negative yielding bonds, while their bonds were purchased by the ECB25. There has also been an increase in debt issuance26. It seems that CSPP-eligible bonds were relatively more favoured than other types of bonds. Therefore, when a central bank buys some assets in large quantities, this might send a signal to market participants that this one is relatively less risky, or more liquid. We might consequently fear some long-lasting consequences of quantitative easing, which could favour incumbent industries.
Source : Financial Times
One could imagine the implementation of a “green quantitative easing” by the ECB, expanding purchases in green bonds. This proposal remains highly controversial. This would mean a change in the mandate of the central bank, which remains price stability in the eurozone. Moreover, low-carbon assets tend to be riskier ones, and this could reduce the quality of a central bank’s portfolio. In addition, the number of purchasable assets would decrease, if the ECB had to take into account environmental standards. Jens Weidmann, President of the Deutsche Bundesbank advocated for instance that: “some observers are actually calling for monetary policy to take climate risks into account. However, neutrality is an important principle of the Eurosystem’s operational framework. (...) to avoid opening Pandora’s box, we should not award preferential treatment to green bonds (...) monetary policy should not be overburdened by other policy objectives”27.
What remains today is indirect forms of green quantitative easing, as the ECB dedicates 10% of its public sector purchase program to European institutions such as the European Investment Bank, which itself gives 25% of its lending facility to environmental projects.
Should we aim at a financial regulation taking into account environmental constraints? Even if they did not implement a green “quantitative easing”, some central banks have chosen to go further. Emerging markets are the ones which went the furthest in climate-related financial regulation. For instance, the Reserve Bank of India “requires that commercial banks allocate a certain proportion of lending to a list of “priority sectors”, which now includes renewable energy” (Campiglio et al., 2018). Banque du Liban also already incentivized commercial banks to “increase the share of green lending projects of their loan portfolio” through the allowance of lower reserves”28. The People’s Bank of China is about to implement such measures in its Macro-Prudential Assessment framework (Campiglio et al., 2018). The Bangladesh Bank (Barkawi, Monnin, 2015) “provides a BDT 2 billion refinancing line to promote green finance. (...) Bank loans for projects in [solar energy, biogas and effluent treatment projects] can be refinanced by Bangladesh bank at 5% provided that the interest charged to bank customers does not exceed 9%”. In addition, the Bank provided US$200 million in a longer-term refinancing window to “green initiatives including water and energy efficiency measures in the textiles industry”. There is also a credit quota of 5% for green loans. The Bank of Japan (Bank of Japan, 2010) has adopted a relatively similar financing scheme, offering preferential rates to financial institutions which finance “green” businesses.
In Europe, the EU High-Level Expert Group on Sustainable Finance published a proposal to create different capital requirements, depending on the direction of loans towards a greener economy or to carbon-intensive sectors: “a well-identified “green” and, potentially “brown” asset class is needed to which differential capital requirements could be applied” (HLEG, 2018). However, this requires creating an appropriate taxonomy of what should be denominated as a “green” or “brown” asset. Therefore, the HLEG recommends developing “taxonomies with definition, together with screening criteria, thresholds and metrics covering the complete scope of the sustainability taxonomy framework”. The European Commission advised the European Supervisory Authorities to “take into account the environmental, social and governance (ESG) factors arising within the framework of their mandate”. They suggest to “provide guidance on how sustainability considerations can be effectively embodied in relevant EU financial legislation”29. Banque de France is already incorporating the ESG criteria in its investment decisions30. If it does not finance environmental projects, a central bank could still stop purchasing assets with high-risk profiles (among which climate-related risks).
Some of these measures are again subject to controversy (such as the creation of different capital ratios for “green” and “brown assets”). First, restraining credit access to carbon-intensive companies might increase their cost of finance, and therefore delay their investments in emissions-reducing technologies. Moreover, easing credit requirements might trigger a “green bubble”, and have a negative effect on the soundness of banks. There is also the question of the screening of projects, which comes at a cost. Indeed, banks may not be accustomed to judge if an investment should fall under a green label or not. Eventually, nothing would prevent major carbon-emitting companies to finance themselves on financial markets where this type of prudential rules isn’t applied. There would be thus a need for an international coordination. Climate stress-tests and disclosure of risks: There is still a consensus over mitigating climate-related risks. In order to assess those risks, it is necessary to develop “climate stress-tests”. Thanks to the Orbis database, some researchers pointed out that, although the exposure of banks regarding fossil fuels is quite low (3-12%), 40% of equities would be considered at-risk, if we integrate all sectors (Battiston et al., 2017). In order to do this, they use the “top sectors by direct GHG emissions according to Eurostat”, as well as the “carbon leakage risk classification”, defined by a 2015 European Commission Directive, as the list of activities which would be impacted by a carbon price. This way, they are able to run a 100% shock in the market capitalization of climate-sensitive sectors for the top 50 EU banks. They also estimate a second round amplifying shock, taking into account the interconnection between financial institutions.
Table : The impact on the top 50 listed EU banks of a 100 shock in the market capitalization of the climate-sensitive in different, progressive aggregations
Source: Battiston et al., (2017)
The researchers also look at the types of investors which might be the most heavily impacted by a shock (see figure below). On the one hand, the fraction of market capitalization helps to understand the “size of the equity exposure in climate-sensitive sectors” (investment funds hold a greater fraction of the market capitalization of carbon-emitting companies than banks). On the other hand, the fraction of equity portfolio (of the shareholders) quantifies “which actors are potentially more exposed to the climate sensitive sectors” (the authors find that this mainly concerns industrial companies, governments and some credit institutions). Such methodologies could be more and more developed by central banks.
Relative equity exposures of major banks to fossil fuel, energy-intensive and housing companies. Bubble size proportional to total equity holdings in EU and US companies.
Source: Battiston et al., (2017)
Beyond climate stress-tests, disclosure of risks in various domains is of a critical importance. The Financial Stability Board created a Task Force for Climate-related Financial Disclosures. It identified several areas of disclosure such as: “governance” (“the board’s oversight of climate-related risks”), “strategy” (risks related to the “organization’s business, strategy and financial planning”), “risk management” (processes to assess risks) and “metrics and targets” (methodologies of risks metrics)31. With continuing research on how to assess risks, this should lead to a more standardized methodology. Investors must not only have access to information, but also take advantage of metrics, which are relatively easy to compare across businesses. This is in line with the recent steps taken by the Bank of England, which told banks and insurers “to identify a senior executive to take charge of managing climate-change risks and report to the board”. The Bank of England, with the ones of Germany, France, Japan and China, considers the risk of climate change as “system-wide and potentially irreversible if not addressed”32. According to the Bank of England, only “10 percent of banks were taking a long enough view of climate-related risks and on average the banks questioned were found to have a four-year planning horizon”. In France, Article 173 of France’s Energy Transition Law “requires all major institutions (listed companies but also banks and institutional investors) to evaluate, report and address their exposure to long-term climate-related financial risk” (HLEG, 2018). Haldane, Bank of England’s Chief Economist demonstrated that finance is more and more prone to short-termism with the shortening of performance assessment intervals and an increasing frequency of corporate reporting (Haldane, 2010). “In 1940, the mean duration of US equity holdings by investors was around 7 years (...) by 2007, it was around 7 months”. This is why the HLEG suggests a comprehensive reform of the financial system from accounting standards (non-financial information, presentation of a more “long term value” rather than a “mark-to-market valuation”33) to investors duties (investors need to understand ESG criteria and provide relevant information to their clients) and governance (“supervisory manuals should be updated to include sustainability as a risk against which the skills and competences of the members of the governing bodies of companies should be addressed”). Climate change is likely to have an impact on monetary policy, not only because changes in commodity prices would affect inflation expectations, but also because it might heighten financial risks. By including more and more physical and transition risks into their analysis, central banks would remain faithful to their mandate. Further economic research will be needed more than ever, with the development of climate stress-tests, asset-level databases or appropriate frameworks for risks disclosure, in order to design the right financial regulation to adopt, which might include macro-prudential measures in the future? To break “the Tragedy of the Horizon”, it becomes imperative to influence assets pricing and to support an accurate understanding of long-term risks by investors. In the end, to be in a world where “every company, investor, and bank that screens new and existing investments for climate risk is simply being pragmatic”34, information and credible policy commitments will be key.
1 Coeuré, B. (2018). “Monetary policy and climate change”, European Central Bank. [online] Available at : https://www.ecb.europa.eu/press/key/date/2018/html/ecb.sp181108.en.html [Accessed 12/01/2018] ;
2 Carney, M. (2015). “Breaking the tragedy of the horizon - climate change and financial stability”, Bank of International Settlements. [online] Available at : https://www.bis.org/review/r151009a.pdf [Accessed 12/01/2018].
3 Aglietta, M., Espagne, E. (2016). “Climate and Finance Systemic Risks, more than an Analogy? The Climate Fragility Hypothesis”, Working Paper CEPII.
4 IPCC. (2014). Climate Change 2014. Impacts, Adaptation and Vulnerability., 34p.
5 IPCC. (2012). Managing The Risks of Extreme Events and Disasters to Advance Climate Change Adaptation, 20p.
6 Munich Re is a prominent german reinsurance firm.
7 Hope B., Friedman N. (2018), “Climate Change Is Forcing the Insurance Industry to Recalculate”, Wall Street Journal.
8McGee S. (2018), “2017 global insured losses reach $135bn - Munich Re”, Insurance Times.
9 OECD. (2018)“Insurance - Gross insurance premiums”. [online] Available at: https://data.oecd.org/insurance/gross-insurance-premiums.htm [Accessed 12/01/2018].
10 Remarks by AXA CEO Thomas Buberl at the One Planet Summit in Paris, 12 December 2017.
11International Association of Insurance Supervisors. (2018). Compiled Comments on Issues Paper on Climate Change Risks to the Insurance Sector, 72p. (p.27).
12 Gonzalez G. (2017) “Trump signs bill forgiving $16 billion NFIP debt”, Business Insurance.
13 Congress of the United States. (2017) “The National Flood Insurance Program: Financial Soundness and Affordability”, CBO, 44p.
14 Hurricane Harvey began on August 17th, 2017 and was followed by hurricane Irma, which ended on September 13th, 2017.
15Meyer, G., Gray, A., Fleming, S. (2017). “Insurers and energy companies count cost of tropical storm Harvey”, Financial Times.
Zhang, B. (2017). “ Hurricane Harvey could cost United Airlines more than $265 million”, Business Insider.
16“What is a catastrophe bond?”, Artemis [online] Available at: http://www.artemis.bm/library/what-is-a-catastrophe-bond.html [Accessed 01/12/2018].
17 International Energy Agency. (2018). “WEO Model” [online] Available at : https://www.iea.org/weo/weomodel/ [Accessed 12/01/2018].
18 Paul Griffin (July 2017). The Carbon Majors Database. CDP Carbon Majors Report 2017. Climate accountability Institute.
19 The Cop 21 took place between November 30th, 2015 and December 12th, 2015.
20 The One Planet Summit took place on December 12th, 2017.
21 Meinshausen, M., Meinshausen, N., Hare, W., Raper, S., Frieler, K., Knutti, R., Frame, D., Allen, M., 2009. Greenhouse-gas emission targets for limiting global warming to 2 degrees C. Nature 458 (7242), 1158–1163.
22 Nordhaus, W. (2017). “Integrated Assessment Model of Climate Change”, NBER [online] Available at: https://www.nber.org/reporter/2017number3/nordhaus.html [Accessed 01/12/2018].
23 The Economist Intelligence Unit. (2015). The cost of inaction: Recognising the value at risk from climate change, 64p.
25 Jackson, G. (2016) “Henkel and Sanofi set new milestone with negative yielding bonds”, Financial Times.
26 Lester, K. (2016) “Corporate Europe Embraces Bond as Yields Plunge to Record Low”, Bloomberg [online] Available at: https://www.bloomberg.com/news/articles/2016-09-07/corporate-bond-sales-swell-in-europe-as-yields-slump-to-record Accessed [01/12/2018].
27Welcome and Opening Speech by Jens Weidmann, Global Public Investor Symposium on “Green bond issuance and other forms of low-carbon finance”, 13 July 2017.
28 Dikau, S., Ryan-Collins, J. (2017). Green Central Banking in Emerging Market and Developing Countries, 23p. New Economics Foundation.
29European Commission. (2017) “Reinforcing integrated supervision to strengthen Capital Markets Union and financial integration in a changing environment”, [online] Available at : http://ec.europa.eu/finance/docs/law/170920-communication-esas_en.pdf [Accessed 01/12/2018].
30 Banque de France. (2018). Responsible Investment Charter of the Banque de France, 4p.
31 Task Force on Climate-Related Financial Disclosure. (2017). Recommendations of the Task-Force on Climate-related Financial Disclosures, 66p.
32 Hook, L., Binham, C. (2018) “Bank of England tells institutions to prepare for climate change”, Financial Times.
33 Mark-to-market “is an accounting practice that involves recording the value of an asset to reflect it current market level”. See Investopedia. “Mark to Market” [online] Available at: https://www.investopedia.com/terms/m/marktomarket.asp [Accessed 12/01/2018].
34 World Bank Group President Jim Yong Kim Remarks at the 2014 Davos Press Conference.
Bibliography AGLIETTA, M. & E. Espagne. “Climate and Finance Systemic Risks, more than an Analogy? The Climate Fragility Hypothesis”, Working Paper CEPII, 2016. Bank of England. The Impact of Climate Change on the UK Insurance Sector, Bank of England Prudential Regulatory Authority, London : 2015. Bank of Japan. Principal Terms and Conditions for the Fund-Provisioning Measure to Support Strengthening the Foundations for Economic Growth Conducted Through the Loan Support Program, Tokyo : 2010. BARKAWI, A., & P. Monin. Monetary Policy and Sustainability - The Case of Bangladesh, Unep Inquiry into the Design of a Sustainable Financial System, Nairobi : 2015. BATTEN, S., R. Sowerbutts, and M. Tanaka. “Let's talk about the weather: The impact of climate change on central banks”, Bank of England Staff Working Paper, London : 2016. BATTISTON, S., A. Mandel, I. Monasterolo, F. Schuetze & G. Visentin. “A Climate Stress-Test of the Financial System”, Nature Climate Change, Vol. 7, pages 283-298, London : 2017. BYRD, J. & E. Cooperman, “Investors and Stranded Asset Risk: Evidence from Shareholder Responses to Carbon Capture and Sequestration (CCS) Events”, Journal of Sustainable Finance & Investment, 2018. CALDECOTT, B., Harnett, E., Cojianu,T., Kok, I., Pfeiffer, A. (2016) Stranded Assets: A Climate Risk Challenge, Washington, D.C.: Inter-American Development Bank (IDB) CAMPIGLIO, E., Dafermos, Y., Monnin, P., Ryan-Collins, J., Schotten, G., Tanaka, M. (2018). “Climate change challenges for central banks and financial regulators”, Nature Climate Change, Vol.8, pages 462-468 Carbon Tracker Initiative. (2011). Unburnable Carbon: Are the world’s financial markets carrying a carbon bubble?, 36p. CASS, D. and Shell, K. (1983). “Do Sunspots Matter?”, Journal of Political Economy , 91(2), pp.193-227. Climate Wise. (2016). Closing the Protection Gap, University of Cambridge Institute for Sustainability Leadership, 40p. DIETZ, S., Bowen, A., Dixon, C., Gradwell, P., “Climate value at risk of global financial assets”, Nature Climate Change, Vol. 6, pages 676-679 Directorate General of the Treasury, Banque de France, Prudential Supervisory and Resolution Authority. (2015). Evaluating Climate Change Risks in the Banking Sector, 107p. European Systemic Risk Board. (2016). Too late, too sudden: Transition to a low-carbon economy and systemic risk, 22p. GRIFFIN, P. A., Jae, A. M., Lont, D. H., and Dominguez-Faus, R. (2015). “Science and the stock market: Investors' recognition of unburnable carbon”. Energy Economics, 52, 112. HALDANE, A. (2010). “Patience and Finance”, Bank of International Settlements. High-Level Expert Group on Sustainable Finance (2018). Financing a Sustainable European Economy, 99p. International Association of Insurance Supervisors. (2018). Issues Paper on Climate Change Risks to the Insurance Sector, 81p. KIYOTAKI, N. and Moore, J. (1997). Credit Cycles. Journal of Political Economy , 105(2), pp.211-248. KLOMP, J. (2014). “Financial fragility and natural disasters: An empirical analysis”, Journal of Financial Stability, Elsevier, vol.13(C), pages 180-193. KRAEMER, M., Mrsnik, M., Petrov, A., Glass, B. (2015). Storm Alert: Natural Disasters Can Damage Sovereign Creditworthiness, Standard & Poor’s, 32p. LAMBERT, C., Noth, F., Schüwer, U. (2014). “How Do Insured Deposits Affect Bank Risk? Evidence from the 2008 Emergency Economic Stabilization Act”, Journal of Financial Intermediation, SAFE Working Paper No. 94. LEWIS, M., (2014). Energy transition & climate change, Kepler Cheuvreux, 35p. RAMIAH V., Martin B., Moosa I. (2013). “How does the stock market react to the announcement of green policies?”, Journal of Banking & Finance, vol.37, Issue 5, pages 1747-1758 MATIKAINEN, S., Campiglio E., Zenghelis, D. (2017). The climate impact of quantitative easing, Grantham Research Institute on Climate Change and the Environment MALCOLM, B., Bergstresser, D., Serafeim G., Wurgler, J. (2018). "Financing the Response to Climate Change: The Pricing and Ownership of U.S. Green Bonds", NBER Working Papers 25194, National Bureau of Economic Research, Inc. MERCURE, J., Pollitt, H., Vinuales, J., Edwards, N., Holden, P., Chewpreecha, U., Salas, P., Sognnaes, I., Lam, A., Knobloch, F. (2018) “Macroeconomic impact of stranded fossil fuel assets”, Nature Climate Change, Vol.8, pages 588-593 NORDHAUS, W. (2013). "Integrated Economic and Climate Modeling," Handbook of Computable General Equilibrium Modeling, Elsevier. RAMIAH, V., Martin, B., Moosa, I. (2013). “How does the stock market react to the announcement of green policies?”, Journal of Banking & Finance, Vol. 37, pages 1747-1758 ROBINS, N., Keen, A., Knight, Z. (2012). Coal and carbon stranded assets: assessing the risk, HSBC, 20p. SEN, S., Von Schickfus, M. (2017). “Will Assets be Stranded or Bailed Out? Expectations of Investors in the Face of Climate Policy”, ifo Working Paper Series 238, ifo Institute - Leibniz Institute for Economic Research at the University of Munich. VON PETER, G., Von Dahlen, S., Saxena, S. (2012). “Unmitigated disasters? New evidence on the macroeconomic cost of natural catastrophes”, BIS Working Papers No 394. WEITZMAN, M. (2011). “Fat-tailed Uncertainty in the Economics of Catastrophic Climate Change”, Review of Environmental Economics and Policy, Vol.5, Issue 2, pages 275-292 WEYZIG, F., Kuepper, B., van Gelder, J.W., and van Tilburg, R. (2014). “The Price of Doing Too Little Too Late; the Impact of the Carbon Bubble on the European Financial System”, Green New Deal Series, Vol. 11, 69p.