The Green Stock Market Bubble

One may contend that there are some clear similarities between the market for sustainable assets now and the situation of tech stocks right before the collapse of the dot-com bubble. In fact, before 2000, the risk of the new technology was primarily idiosyncratic because of the limited scale of production and the low likelihood of a widespread adoption. Subse-quently, with an increasing probability of adoption, also, the old economy was affected by the new technology and, hence, the wealth of the representative agent. As a result, systematic risk increased, which depressed stock prices, because it pushed up the discount rates in the new and old economies. Similarly, given the high probability of a large-scale adoption of the new “green technology,” it is likely that there is not only a bubble forming in green energy stocks, but the boom is affecting the stock market as a whole. I apply a recently developed recursive testing procedure and dating algorithm that is useful in detecting multiple bubble events. Using S&P 500 stock market data, price-dividend ratios, I identify the well-known historical speculative bubbles and find an explosive movement in today’s market starting in June 2021, which can be associated with the new “green technology.” I find that an explosive movement in green stocks started roughly a year before it was migrating to the whole stock market. I argue that this is a good bubble because it will enable businesses to invest cheaply in green energy, hastening the transition away from fossil fuels and assisting in the fight against climate change.

The hype is even getting a bit irrational: Tiziana Life Sciences, a biotech firm with the ticker "TLSA," benefited from confused investors last year when they mistook it for Tesla, ticker symbol "TSLA."High valuations, aggressive interest from individual investors, and parabolic price action are all signs that there is a bubble forming in green energy stocks.
Should We Be Concerned About This Development?Well, one might argue that the market for sustainable assets now and the IT stocks immediately before the dot-com bubble burst have some clear similarities.However, as was ultimately the case with tech companies, today's inflated pricing for climate-friendly assets may reflect their longer-term potential.In the general equilibrium model of Pastor and Veronesi [24], stock price bubbles might occur for innovative firms during technological revolutions, where over time, the nature of the risks related to new technologies changes.Initially, the risk is mainly idiosyncratic, because the scale of production is small and it is unlikely that the new technology is adopted at a large scale.If a particular technology is never adopted, the risk stays idiosyncratic.However, once a technology is adopted at a larger scale, the risk becomes more and more systematic.Subsequently, with an increasing probability that the new technology is adopted, it is more likely that the old economy is affected by the new economy.Stock prices in both the new and old economies are depressed, because systematic risk is increasing, which also positively affects the discount rates.In light of the "green technological revolution," and along similar lines, given the high probability of a large-scale adoption of the "green technology" that we are currently experiencing, it is likely that there is not only a bubble forming in green energy stocks, but the boom is affecting the stock market as a whole.
How to Detect Financial Bubbles in Real Time?An effective method to identify and date financial bubbles is recursive procedures, which belong to the class of econometric detection mechanisms.Central banks and fiscal regulators conduct these procedures with realtime data as warning alerts in surveillance strategies.Phillips et al. [28,29] (PSY) develop a recursive flexible window method that is suited for practical implementation with long historical time series.For the origination and termination of multiple bubbles, the method delivers a consistent real-time date-stamping strategy.I empirically apply the PSY methodology to S&P 500 price-dividend ratios over a long historical period from January 1871 to December 2021.Next to well-known historical episodes of exuberance and collapses, the approach successfully identifies an explosive movement in today's market.Additionally, I find that an explosive movement in green stocks started roughly a year before it was migrating to the whole stock market.Hence, the current bubble can be associated with the new "green technology."I argue that this is a good bubble, because by allowing the companies to make cheap green energy investments, the bubble could speed our shift away from carbon fuels and help to combat climate change.
The rest of the paper is structured as follows: In the next sections, I briefly review the literature.I discuss the data and present the methodology in subsequent sections.Finally, the empirical results of the study are presented and the last section concludes.

The Circular Economy
In recent years, policymakers have been addressing the integration of sustainable practices in the economy, as shown by global commitments to sustainability goals.For example, the European Commission launched in 2015 the first Circular Economy Action Plan, which is an agenda of the initiatives to implement for the transition to a circular economy, a sustainable and resilient economic system.Actions are proposed to integrate the ESG methodology-an acronym encompassing the Environmental, Social, and Governance pillars, respectively-into the traditional risk assessment process undertaken by financial institutions.Sustainable finance can be defined as the process of taking into account ESG considerations when making investment decisions in the financial sector.Sustainable financial assets are the instruments that meet these requirements and the performance of sustainable financial assets plays a crucial role in providing the financial means for the transition to a circular economy (see, e.g., Sepetis [32].As a result, recently, green bonds and stocks are performing better than their non-green benchmarks.In order to benefit the environment, promote sustainability, and draw in investors, many businesses are putting a strong emphasis on sustainability and implementing eco-friendly business models.This rise is partially attributable to a growing consensus regarding the contribution of stock markets to economic development. Krüger [18] investigates how stock markets respond to favorable and unfavorable developments relating to a company's social responsibility (CSR).Investors react substantially unfavorably to bad news and only slightly negatively to good news, especially if it stems from agency issues.Additionally, CSR news that has more robust legal and economic facts causes a more noticeable investor response.The early 1970s saw the emergence of the literature on the link between ESG factors and business financial success.In their 2015 study, Gunnar et al. compared the results of roughly 2200 distinct investigations.They demonstrate how empirically very well-founded the business case for ESG investing is.A non-negative ESG-performance association is found in almost 90% of studies, and a sizable majority of studies find positive results.Fernando et al. [10] use institutional shareholders as a lens through which to study the value effects of corporate social responsibility.They discover a stark difference between corporate strategies that reduce a company's exposure to environmental risk and those that raise its perceived environmental friendliness (or "greenness").As predicted by risk management theory, institutional investors steer clear of stocks with substantial environmental risk exposure because they trade at lower prices.According to these results, corporate environmental policies that reduce exposure to environmental risk increase shareholder value.Escrig-Olmedo (2019) compares publicly available data from the financial market's most representative ESG rating and information providers for two time periods: 2008 and 2018.The results demonstrate that in order to measure company performance more precisely and robustly in order to address new global concerns, ESG rating agencies have added additional criteria to their evaluation models.However, the business sustainability evaluation procedure used by ESG rating firms still does not completely incorporate sustainability principles.Zara et al. [36] empirically investigate a company's degree of circularity, measured by a circularity score, and show that equity investors into circular undertakings could benefit from reduced stock return volatility, as well as a greater ability to withstand exogenous negative events.

The Bubble Literature
Research in finance studies rational and irrational bubbles. 1 According to the rational expectations school Blanchard and Watson [3], Tirole (1982), and Diba and Grossman [7], bubbles are the clever creation of the invisible hand.Due to projections of cash flows [12], changes in discount rates (Pastor andStambaugh, 2006, 2009), and a combination of risk premia and learning [23,24], rational asset pricing aids in the understanding of high prices during bubble episodes.As discussed previously, the general equilibrium model put forth by Pastor and Veronesi [24] predicts that during technological revolutions, stock values of innovative companies would experience bubbles.It also follows that the new economy's stock prices decline more when the new economy discount rates grow due to a rise in the new economy's market beta.It is important to note that such a rational stock price bubble would be anticipated ex ante but visible ex post.In a similar vein, Perez [25] contends that the two most recent boom-bust cycles-the internet mania and crash around the year 2000 and the easy liquidity boom and bust around the Lehman default in 2008-are two separate elements of a single structural phenomenon.The first boom-bust cycle was based on technological innovation, while the second was based on financial innovation.These "episodes are endemic to the way in which the market economy matures and assimilates successive technological revolutions," according to her claim.When investors begin targeting capital gains instead of dividends, the real economy is becoming decoupled and bubbles begin to form.Following the ensuing collapse, there is a reconnection to the real economy and the beginning of a phase in which production capital assumes dominance over finance capital.
More recently, Eugene Fama's assertion that stock prices do not experience price bubbles was analyzed by Greenwood et al. (2018).According to their findings, it is generally impossible to forecast that significant price rises will be followed by extremely low returns at the industry level.Nevertheless, keeping track of those industry portfolio boom events greatly raises the likelihood of a subsequent crash.By depending on some of the traits of the price run-up, investors can predict the bubble.In a similar vein, Lehnert [19] assesses Alan Greenspan's assertion that stock price bubbles develop during times of euphoria and typically burst because of rising fear.Based on US industry returns from 1959 to 2014, he discovers that, on average, euphoria and positive market sentiment change into fear around a year before an industry crash.For price runs in industries that do not ultimately crash, there is no specific euphoria-fear pattern.

Bubble Detection in Real Time
I take into account a lengthy historical time series with numerous recognized crisis events.The real S&P 500 stock price index and the real S&P 500 stock price index dividend are included in the statistics, which were both taken from Robert Shiller's website.From January 1871 to December 2021, the data were sampled monthly, totaling 1812 observations.The price-dividend ratio is the ideal variable to analyze, because it reflects asset prices in relation to fundamentals, according to the following pricing equation: where P t is the after-dividend price of the asset, D t is the payoff received from the asset (i.e., dividend), r f is the risk-free interest rate, U t represents the unobservable fundamentals, and B t is the bubble component.
In order to identify bubbles, one method is to look for empirical evidence of explosive behavior in house price ratios (house prices adjusted for rent).One of the earlier techniques is the variance bound test put forth by Shiller (see [34], which states that in the event of a rational bubble, the variation of the observed asset price should exceed the bound imposed by the variance of the underlying value.With the restriction that focus is placed on the volatility, a factor which can also be influenced by changes in expected returns, it can be used for this purpose even though it was not initially intended to test for the presence of bubbles.By drawing on the bubble literature, Campbell and Shiller [4] present a new indirect technique for bubble detection based on unit root testing.They base their strategy on the notion that if there is a discrepancy between an asset's price and its intrinsic value, it will behave explosively as a bubble forms.The two cases they identify are (i) when the asset price is non-stationary in level but the fundamental value is, and (ii) when both the asset price and the fundamental value are non-stationary.However, in that second scenario, a co-integration test is required,if the asset price and its fundamental value are co-integrated, and as a result, have co-movement over time, their non-stationary behavior is not an indication of the presence of a bubble.The hypothesis put forth by Campbell and Shiller [4] is shown to be sufficient to establish the presence of a bubble (see Diba and Grossman [7].Left-tailed unit root tests and co-integration tests have been the standard method in bubble identification notwithstanding their drawbacks.Due to their inability to distinguish between a periodically collapsing bubble trend and a stationary process, these tests are unable to detect explosive bubbles when there are periodically collapsing bubbles in the time series, meaning that the collapsing bubbles "break" the non-stationary characteristics of the sample.The typical left-tailed unit root test can therefore be used to mistakenly conclude that a time series with many bubbles is stationary and draw the incorrect conclusion that the data is bubble-free.A recent direct bubble testing method created by Phillips et al. has taken into account the drawbacks of the left-tailed unit root tests (2011).Instead of the usual lefttailed test, the authors adopt the right-tailed ADF test.While Philips et al. (2011) utilized both the left-tailed and right-tailed ADF tests to examine the unit root behavior, their alternative hypotheses-"stationary behavior" for the former and "mildly explosive" for the latter-differ.Phillips et al. [26] overcame the problem of periodically collapsing bubble detection by directly looking for evidence of non-linear explosive behavior in the data, avoiding the possibility of misinterpreting a rejection of the null hypothesis owing to stationary behavior.We use Phillips et al.'s recently created multiple bubble detecting method (2015a, 2015b).The objective of the PSY technique is to identify the regional explosive dynamics of speculative bubbles.The testing methodology is based on a right-tailed unit root test with an explosive (bubble) alternative and a unit root null (market fundamental).The PSY test requires conducting subsample regressions.The purpose is to take care of any potential structural break or regime switching within the sample period.Let r 1 and r 2 be the fractional starting and ending points of a subsample regression, with the corresponding ADF statistic denoted by ADF r 1 r 2 . The algorithm calculates the ADF statistic repeatedly on a sample sequence.Suppose r is the observation of interest.The ending points of the samples r 2 vary from r 0 to 1 and the starting points of the samples r 1 vary from the first observation to r 2 -r 0 , where r 0 is the minimum window size required to initiate a regression (see Phillips et al. [28,29] for details).In the empirical application, I set the window size to 95 observations of the sample, based on the rule r 0 = 0.01 + 1.8/√1809, to minimize the probability of size distortion.Inference of explosiveness for observation r is based on the sup value of the ADF sequence.Phillips et al. [29] show that this strategy can consistently estimate the origination and termination dates of multiple bubbles.Critical values for the test can be obtained using a wild bootstrapping procedure (see Milunovich et al. (2016)). 2  Phillips and Shi [30] make use of the method to analyze S&P 500 stock prices and sovereign risk in European Union countries over 2001-2016 using government bond yields and credit default swap (CDS) premia.A pseudo-real-time empirical analysis of these data shows the effectiveness of the monitoring strategy in capturing key events and turning points in market risk assessment.
Given that the GSADF statistics for the entire sample exceeds its 1% right-tailed critical threshold, I find strong evidence that overall, the S&P 500 price-dividend ratio experienced explosive subperiods.This is consistent with Phillips et al.'s [28] findings.Then, utilizing the PSY dating techniques, I conduct a (pseudo) real-time bubble monitoring experiment for the S&P 500 stock market.I track the price-dividend ratio time series behavior for the market from June 1878 through the conclusion of the sample period using a training set of 95 observations.The empirical results from the PSY procedure are shown in Fig. 1, which plots the S&P 500 price-dividend ratio (in blue, right axis) and the Fig. 1 The figure presents the empirical results for the S&P 500 price-dividend ratio from the PSY procedure 2 The PSY algorithm faces the problem of multiplicity in testing that taints conventional significance values, much like all other recursive testing methods.To solve this problem, Shi and Phillips [33] suggest a multiple testing approach that uses simulations to demonstrate adequate performance in finite samples while identifying the proper test critical values.When the multiplicity problem is not taken into account, they find that the identified origination (or termination, as the case may be) date is slightly earlier (later).However, in our analysis, this problem is only marginal and does not affect our conclusions.
backward SADF statistic sequence (in green, left axis) against the 95% ADF critical value sequence (in red, left axis).As can be seen and in line with Phillips et al. [28], the identified periods of exuberance in the market include the so-called post long-depression period (1879M10-1880M04), the great crash episode (1928M11-1929M10), the postwar boom (1955M01-1956M04), the BlackMonday in October 1987 (1986M06-1987M09), and the dot-com bubble (1995M11-2001M08).The PSY strategy also identifies two episodes related to market collapses instead of bubble expansion, namely the 1917 stock market crash (1917M08-1918M04) and the subprime mortgage crisis (2009M02-M04).Interestingly, the PSY strategy also identifies an explosive movement in today's market starting in June 2021.

Bubble Migration
I hypothesize that the boom in the overall stock market is triggered by a bubble forming in green energy stocks.Hence, in the following, I apply the PSY testing procedure to the NASDAQ Clean Edge Green Energy Total Return Index (CEXX), which I use as a proxy for green energy stocks.While the price-dividend ratio is the ideal variable to analyze, because it reflects asset prices in relation to fundamentals, Philips et al. (2015) also apply the PSY procedure to the logarithm of the S&P 500 price index.They find that these adjustments produce only minor discrepancies, and the empirical results are qualitatively unchanged.Hence, I use the logarithm of the NASDAQ Clean Edge Green Energy Total Return Index as the main variable.The available data are sampled monthly over the period from September 2011 to December 2021, constituting 124 observations; hence, the window size is set to 21 observations of the sample.
The empirical results for the NASDAQ Clean Edge Green Energy Total Return Index from the PSY procedure are shown in Fig. 2. I plot the NASDAQ Clean Edge Green Fig. 2 The figure presents the empirical results for the NASDAQ Clean Edge Green Energy Total Return Index from the PSY procedure Energy Total Return Index (in blue, right axis) and the backward SADF statistic sequence (in green, left axis) against the 95% ADF critical value sequence (in red, left axis).As can be seen, besides some minor historical explosive movements, more recently, the PSY strategy identifies a period of exuberance in today's market starting in July 2020.Hence, the explosive movement in green stocks started roughly a year before it was migrating to the whole stock market, as it was the case for tech stocks around the year 2000.I implement the Greenaway-McGrevy and Phillips [13] methodology to statistically detect the presence of migration between the two types of bubbles.Using the same methodology, Deng et al. [6] detect significant migration from the stock to the housing market bubble in China in 2009 and a temporary spillover in 2007.I am interested in the bubble spillover effect from the markets for green energy stocks to the overall stock market.The previously obtained autoregressive coefficients form the basis of the PSY test and hence, to some extent, capture the degree of explosiveness.The contagion regression is based on the two autoregressive coefficient sequences for both markets.I am interested in the relationship between the autoregressive coefficient of the market for green energy stocks and the multiple-periodahead autoregressive coefficient of the overall stock market.Results confirm my previous conjecture and suggest that long lags (approximately eleven months) are at work in the migration of the bubble from green energy stocks to the overall stock market.My results are in line with the experimental evidence presented in Draganac and Lu [8], who find that non-pecuniary factors influence financial markets, which can lead to stock price bubbles.They present a lab experiment in which subjects simultaneously trade two assets in the market.Depending on the assigned treatments, one asset is associated with positive environmental externality or image considerations.When prosocial preference is accompanied by image consideration, individuals with strong non-pecuniary considerations seem to push their portfolios into green assets relative to the regular asset, leading to a growing risk of a price bubble in environmentally friendly focused asset markets.Their results also indicate that the non-green assets might further deviate from their fundamental value simply because of the existence of non-pecuniary considerations in the asset market.
One might want to refer to the current scenario as a bubble with beneficial side effects in contrast to other earlier speculative bubbles.It would be ideal to see an equity market bubble that rewarded businesses that concentrate on renewable energy, as cheap money would enable successful long-term green investments.A booming IPO market implies easy access to money for new green entrants.In this sense, the bubble might hasten our transition away from fossil fuels and contribute to the fight against global warming.Like in 2000, there is no question that the stock market is correct once more: green technology businesses will prosper in the future; however, the value is currently going a little high, so it will decline again, but the green technological development will continue.

Conclusions
Green stocks are currently in high demand.Stock prices of green firms like solar companies or electric vehicles makers have recently risen up to nine-fold.High valuations, aggressive interest from individual investors, and parabolic price action are all signs that there is a bubble forming in green energy stocks.One might argue that there are some obvious parallels between today's market for sustainable assets, and tech stocks just before the dot-com bubble burst.Indeed, before 2000, due to the small scale of production and a low probability of a large-scale adoption, the risk of the new technology was initially mostly idiosyncratic.As the probability of adoption increased, the new technology affected the old economy and with it the representative agent's wealth.The resulting increase in systematic risk depressed stock prices in both the new and old economies, because it pushed up the discount rates.Similarly, given the high probability of a large-scale adoption of the "green technology," it is likely that there is not only a bubble forming in green energy stocks, but the boom is affecting the stock market as a whole.Using a recently developed recursive testing procedure and dating algorithm and S&P 500 stock market data, I identify wellknown historical speculative bubbles and find an explosive movement in today's market starting in June 2021.I find that an explosive movement in green stocks started roughly a year before it was migrating to the whole stock market.Hence, the current bubble can be associated with the new "green technology." Understanding the financial bubble phenomena is important in light of the significant economic implications.Policymakers have emphasized the importance of research aimed at developing measures to address financial bubbles in order to avoid further crises or to ameliorate their effects.Indeed, the exponential growth of the green assets has already drawn the attention of key market participants.While ESG enthusiasts expect green investments to shoot even higher, some executives and analysts warned of "crazy" valuations.The Bank for International Settlement has already warned of the growing risk of a price bubble in environmentally friendly focused asset markets.However, I argue that this is a good bubble because it will enable businesses to invest cheaply in green energy, hastening the transition away from fossil fuels and assisting in the fight against climate change.