付迎杰

In recent years regulators have begun warning about the threat that climate change poses to the stability of the financial system. Following its strategy review in July， the European Central Bank （ECB） will assemble a “climate change action plan”. Mark Carney， the former governor of the Bank of England， warned of financial risks from climate change as long ago as 2015. In America the Commodity Futures Trading Commission last year published a 200-page report beginning “Climate change poses a major risk to the stability of the US financial system.” But progressive Democratic politicians are calling on President Joe Biden not to reappoint Jerome Powell as the chairman of the Federal Reserve， partly because they think he has done too little to eliminate climate risk.

Just how damaging does climate risk stand to be， though？ Early stress tests by central banks and disclosures of companies are starting to shed light on the question. For the most part， the evidence that it could bring down the financial system is underwhelming. But a lot hangs on whether governments set out a clear path for reducing emissions， such as through carbon taxes and energy-efficiency standards， giving banks enough time to prepare.

Climate change can affect the financial system in three ways. The first is through what regulators describe as “transition risks”. These are most likely to arise if governments pursue tougher climate policies. If they do， the economy restructures： capital moves away from dirty sectors and towards cleaner ones. Companies in polluting industries may default on loans or bonds; their share prices may collapse.

The second channel is financial firms’ exposure to the hazards of rising temperatures. Attributing individual natural disasters to climate change is tricky， but the Financial Stability Board， a group of regulators， estimates that global economic losses resulting from weather-related catastrophes went from $214bn in the 1980s， in 2019 prices， to $1.62trn in the 2010s. These losses are often borne by insurers （though over time the costs should be passed on to customers through higher premiums）.

The financial system could also be exposed to any wider economic damage caused by climate change， say if it triggered swings in asset prices. This third channel is harder to quantify. Academic estimates of the effect of 3°C of warming （relative to pre-industrial temperatures） veer from financial losses of around 2% to 25% of world GDP， according to the Network for Greening the Financial System， a group of supervisors. Even the gloomiest estimate might prove too rosy if climate change triggers conflicts or mass migrations.

Perhaps the worst-case scenario for the financial system is where transition risks crystallize very suddenly and cause wider economic damage. In 2015 Mr. Carney described a possible “Minsky moment”， named after Hyman Minsky， an economist， in which investors’ expectations about future climate pol-icies adjust sharply， causing fire sales of assets and a wide-spread repricing of risk. That could spill over into higher borrowing costs.

The value of financial assets exposed to transition risk is potentially very large. According to Carbon Tracker， a climate think-tank， around $18trn of global equities， $8trn of bonds and perhaps $30trn of unlisted debt are linked to high-emitting sectors of the economy. That compares with the $1trn market for collateralized debt obligations （CDOs） in 2007， which were at the heart of the global financial crisis. The impact of losses， however， would depend on who owns the assets. Regulators might be especially concerned about the exposures of large， “systemically important” banks and insurers， for instance.

Preliminary stress tests conducted by central banks suggest that the impact of climate change on these sorts of institutions might be manageable. In April the Banque de France （BDF） released the results from such an exercise. It found that French banks’ exposures to transition risks were low. Claims on insurers， though， did rise as a result of worse droughts and flooding， by more than five times in some regions.

In a recent paper the ECB and the European Systemic Risk Board found similar results. The exposures of euro-area banks and insurers to the highest-emitting sectors were “l(fā)imited”， although losses in a “hot-house world” scenario where temperatures rise by 3.5°C compared with pre-industrial times were more severe. Still， in both cases， banks’ losses on their corporate loan books were only around half the level of those in the regular stress tests1 of euro-area lenders， which they were deemed to be well-capitalized enough to pass.

Those findings are consistent with an exercise by the Dutch central bank （DNB） in 2018， which found that the impact on Dutch financial firms from transition risks was “manageable”. In its most severe scenario， there was a sudden change in climate policy alongside rapid progress in renewable energy development， causing a “double shock” for companies and a severe recession. Even then， banks’ capital ratios fell by about four percentage points. That is sizeable， but still less than what the banks experienced in this year’s regular stress tests by the European Banking Authority， which they were deemed to pass.

To what extent are these stress tests realistic？ Mark Campanale of Carbon Tracker is skeptical， pointing out that most firms are using out-of-date models. If auditors were ever to stress companies’ assets against a much lower oil price， the associated write-downs could trigger a collapse in investor sentiment of the sort regulators fear， he claims. Nor do the stress tests include a full-blown Minsky crisis.

Yet in other respects they are conservative. Most of the tests used an accelerated time frame—five years in the DNB and BDF cases—in effect assuming that firms are stuck with the balance-sheets they have today. But it seems reasonable to think that banks and insurers will change their business models as the climate transition progresses， curbing the impact on the financial system. The BDF ran a second exercise where firms were allowed to make realistic changes to their business models over 30 years. Unsurprisingly， that allowed banks to sharply reduce lending to fossil-fuel sectors， and insurers to raise premiums.

Nonetheless， the stress tests reveal the importance of giving firms time to adapt. And that makes a predictable path for government policy important. The BDF found that credit losses were highest when policy was delayed and there was a sudden transition. Perhaps the most plausible scenario in which climate change affects financial stability is one in which governments dawdle， and then have no choice but to take drastic action in the future.

最近幾年，監(jiān)管機(jī)構(gòu)已經(jīng)開始警告人們氣候變化將威脅到金融系統(tǒng)的穩(wěn)定。在2020年7月份的戰(zhàn)略審查之后，歐洲中央銀行將制訂一項(xiàng) “氣候變化行動(dòng)計(jì)劃”。英國央行前行長馬克·卡尼早在2015年就警告過氣候變化會(huì)帶來金融風(fēng)險(xiǎn)。在美國，商品期貨交易委員會(huì)去年發(fā)表了一份長達(dá)200頁的報(bào)告，開頭寫道：“氣候變化對美國金融系統(tǒng)的穩(wěn)定性構(gòu)成重大風(fēng)險(xiǎn)”。但民主黨進(jìn)步派政治家呼吁總統(tǒng)喬·拜登不要重新任命杰羅姆·鮑威爾為美聯(lián)儲(chǔ)主席，部分原因是他們認(rèn)為鮑威爾在消除氣候風(fēng)險(xiǎn)方面做得太少。

不過，氣候風(fēng)險(xiǎn)的破壞性到底有多大？各國中央銀行的早期壓力測試和各公司披露的消息讓這個(gè)問題初露端倪。在大多數(shù)情況下，氣候風(fēng)險(xiǎn)可能導(dǎo)致金融系統(tǒng)崩潰的證據(jù)是不充分的。但這在很大程度上取決于政府是否制定了明確的減排路徑（比如通過征收碳稅和推行能效標(biāo)準(zhǔn)），給銀行足夠的準(zhǔn)備時(shí)間。

氣候變化可以通過三個(gè)方面影響金融系統(tǒng)。第一個(gè)方面是通過監(jiān)管機(jī)構(gòu)所描述的“轉(zhuǎn)型風(fēng)險(xiǎn)”。如果政府推行更嚴(yán)厲的氣候政策，這些風(fēng)險(xiǎn)最有可能出現(xiàn)。如果他們這樣做，經(jīng)濟(jì)就會(huì)重組：資本從重污染行業(yè)轉(zhuǎn)移到更清潔的行業(yè)。高污染行業(yè)的公司可能會(huì)拖欠貸款或出現(xiàn)債券違約;他們的股票價(jià)格也可能會(huì)暴跌。

第二個(gè)方面是金融公司對氣候變暖存在風(fēng)險(xiǎn)敞口。將個(gè)別自然災(zāi)害歸咎于氣候變化有欺騙性，但金融穩(wěn)定委員會(huì)（由各國金融管理機(jī)構(gòu)組成）估計(jì)，與天氣有關(guān)的災(zāi)難造成的全球經(jīng)濟(jì)損失從20世紀(jì)80年代的2140億美元（按2019年價(jià)格水平計(jì)算）上升到2010年代的1.62萬億美元。這些損失通常由保險(xiǎn)公司承擔(dān)（盡管最終這些成本將通過更高的保險(xiǎn)費(fèi)轉(zhuǎn)嫁給客戶）。

如果氣候變化引發(fā)了資產(chǎn)價(jià)格的波動(dòng)，金融系統(tǒng)也可能遭受氣候變化造成的更廣泛的經(jīng)濟(jì)損失。第三種風(fēng)險(xiǎn)更難量化。央行與監(jiān)管機(jī)構(gòu)綠色金融網(wǎng)絡(luò)（由各國金融監(jiān)管機(jī)構(gòu)組成）稱，學(xué)術(shù)界對氣候變暖3攝氏度（相對于工業(yè)化前的溫度）的影響做出估計(jì)，相關(guān)金融損失占全球GDP的比例將從 2%左右增至25%。如果氣候變化引發(fā)沖突或大規(guī)模移民，后果將不堪設(shè)想。

對金融系統(tǒng)來說，也許最糟糕的情況是轉(zhuǎn)型風(fēng)險(xiǎn)突然具體化，并造成更大范圍的經(jīng)濟(jì)損失。2015年，卡尼先生描述了一個(gè)可能會(huì)出現(xiàn)的“明斯基時(shí)刻”（以經(jīng)濟(jì)學(xué)家海曼·明斯基的名字命名），即投資者對未來氣候政策的預(yù)期急劇調(diào)整，導(dǎo)致資產(chǎn)的低價(jià)拋售和廣泛的風(fēng)險(xiǎn)重新定價(jià)。這可能會(huì)導(dǎo)致更高的借貸成本。

面臨轉(zhuǎn)型風(fēng)險(xiǎn)的金融資產(chǎn)可能價(jià)值巨大。根據(jù)氣候智庫“碳追蹤”的數(shù)據(jù)，全球約有18萬億美元的股票、8萬億美元的債券和約30萬億美元的未上市債務(wù)與經(jīng)濟(jì)中的高排放行業(yè)有關(guān)。與之相比，2007年的擔(dān)保債務(wù)憑證市場規(guī)模為1萬億美元，是全球金融危機(jī)的核心所在。然而，損失的影響將取決于誰擁有這些資產(chǎn)。例如，監(jiān)管機(jī)構(gòu)可能會(huì)特別關(guān)注“具有系統(tǒng)重要性”的大型銀行和保險(xiǎn)公司的風(fēng)險(xiǎn)敞口。

各國中央銀行進(jìn)行的初步壓力測試表明，氣候變化對這些機(jī)構(gòu)的影響可能是可控的。2020年4月，法國央行發(fā)布了這樣的測試結(jié)果。它發(fā)現(xiàn)法國銀行轉(zhuǎn)型風(fēng)險(xiǎn)敞口很小。但是，由于干旱和洪澇災(zāi)害加劇，對保險(xiǎn)公司的索賠確實(shí)增加了，在一些地區(qū)增加了五倍以上。

在最近的一份文件中，歐洲央行和歐洲系統(tǒng)性風(fēng)險(xiǎn)委員會(huì)也得出了類似的結(jié)果。歐元區(qū)銀行和保險(xiǎn)公司對碳排放量最高的行業(yè)的風(fēng)險(xiǎn)敞口“有限”，盡管在“溫室世界”（溫度比工業(yè)化前上升3.5攝氏度）的情形下?lián)p失更為嚴(yán)重。不過，在這兩種情況下，銀行在企業(yè)貸款賬面上的損失只有歐元區(qū)銀行定期壓力測試中的一半左右，它們被認(rèn)為資本充足，足以通過壓力測試。

這些發(fā)現(xiàn)與荷蘭央行2018年的一項(xiàng)測試一致，該測試發(fā)現(xiàn)轉(zhuǎn)型風(fēng)險(xiǎn)對荷蘭金融企業(yè)的影響是“可控的”。在形勢最嚴(yán)峻的情景中，隨著可再生能源開發(fā)取得快速進(jìn)展，氣候政策也發(fā)生了突然變化，對企業(yè)造成了“雙重沖擊”和嚴(yán)重的經(jīng)濟(jì)衰退。即使如此，銀行的資本充足率也下降了大約四個(gè)百分點(diǎn)。這一降幅十分顯著，但仍然低于今年歐洲銀行管理局對銀行進(jìn)行的定期壓力測試中的降幅（銀行已經(jīng)通過測試）。

這些壓力測試在多大程度上有現(xiàn)實(shí)意義？智庫“碳追蹤”的馬克·坎帕納萊對此持懷疑態(tài)度，他指出，大多數(shù)公司都在使用過時(shí)的模型。他說，如果審計(jì)機(jī)構(gòu)在對公司資產(chǎn)進(jìn)行壓力測試時(shí)設(shè)置了低得多的石油價(jià)格參數(shù)，相關(guān)的資產(chǎn)減值可能會(huì)引發(fā)投資者情緒崩潰這種令監(jiān)管機(jī)構(gòu)擔(dān)憂的情況。這些壓力測試也不包括全面爆發(fā)的明斯基危機(jī)。

然而，在其他方面，壓力測試是保守的。大多數(shù)測試使用了一個(gè)加速的時(shí)間框架——在荷蘭央行和法國央行是5年——實(shí)際上是假設(shè)公司無法擺脫它們今天的資產(chǎn)負(fù)債困境。但似乎有理由認(rèn)為，隨著氣候轉(zhuǎn)型的進(jìn)展，銀行和保險(xiǎn)公司將改變它們的商業(yè)模式，用以遏制對金融系統(tǒng)的影響。法國央行進(jìn)行了第二次測試，允許企業(yè)在30年內(nèi)對其商業(yè)模式進(jìn)行務(wù)實(shí)的改變。不出所料，這使得銀行大幅減少對化石燃料部門的貸款，使保險(xiǎn)公司提高保費(fèi)。

盡管如此，壓力測試揭示了給公司預(yù)留足夠時(shí)間去適應(yīng)的重要性。而這使得政府政策的可預(yù)測路徑變得重要。法國央行發(fā)現(xiàn)，當(dāng)政策被推遲并突然轉(zhuǎn)型時(shí)，信貸損失最高。也許氣候變化影響金融穩(wěn)定最有可能出現(xiàn)的情況是，政府拖拖拉拉，然后別無選擇，只能在未來采取激進(jìn)的措施來應(yīng)對。