The Influence of Aviation Disasters on Engine Manufacturers: An Analysis of Financial and Reputational Contagion Risks

One of the key sub-sectors in the aviation industry includes that of engine manufacturers, who have long led technological advancement and the battle to reduce airline carbon emissions. However, these same companies have been susceptible to a number of issues that have been central to international airlines due to higher costs and competition pressures. When an aviation disaster occurs, there is widespread allocation of blame and responsibility, which has left engine manufacturers exposed until the true cause is identified. This can generate many issues with regards to reputational damage and ability to generate finance. We set out to analyze such interactions over time and region. Our results indicate that engine manufacturers have had to contend with substantial income and financial leverage issues in the aftermath of a major aviation disaster, irrespective of whether they have been identified as a causation factor in the incident itself. Further, we clearly identify that there exists an average one day loss of 1.64% in the immediate aftermath of aviation incidents. Substantial corporate instability is found to persist without the company being in any way responsible for the incident. Shortly thereafter, contagion effects increase as speculation diminishes and more factual evidence arrives. The role of social media is examined as a potential contributory factor.


Introduction
In 2018, a record 4.4 billion passengers travelled by air worldwide on 46.1 million flights and the demand for air transport is apparent with almost 82% of all available seats being filled, with 22,000 city pairs connected by direct flights. However the gloss is overshadowed, as there were over 500 fatalities in 2018, which accounted for a fatal accident rate of 0.36 per million flights, or one fatal accident for every 3 million flights, predominantly because of the two 737 Max 8 related accidents, which were subsequently grounded. The previous year had a global fatality rate of 12.2 fatalities per billion passengers, representing the safest year ever on the record for aviation. 1,2 Airline safety reputation as perceived by passengers plays a substantial role in airline choices (Siomkos [2000]) and flight choices (Molin et al. [2017]). Fatal accidents in commercial aviation remain rare and the safety of commercial passenger aviation services is of major concern for the travelling public and regulatory agencies. Liu and Zeng [2007] found that demand for air travel is likely to fall as the fatality rate increases. Airline accidents trigger instantaneous activity in the financial markets because of their unanticipated and cataclysmic nature. This type of negative rhetoric has caused sharp reductions in the share price of companies throughout the world, but it significantly impacts the aviation industry when compared to other consortiums of commercial enterprises (Kaplanski and Levy [2010]). They can have significant effects on an airline's stock price and profitability (Chance and Ferris [1987]; Borenstein and Zimmerman [1988]; Bruning and Kuzma [1989]; Mitchell and Maloney [1989]; Rose [1992]; Dionne et al. [1997]; Nethercurtt and Pruitt [1997]; Bosch et al. [1998]). Akyildirim et al.
[2020] investigated a number of stylised facts relating to the effects of airline disasters on aviation stocks, where results indicate a substantially elevated levels of share price volatility in the aftermath of aviation disasters. Share price volatility appeared to be significantly influenced by the scale of the disaster in terms of the fatalities generated, with evidence suggesting the existence of significant contagion and information flow effects upon the broad aviation sector. Ho et al. [2013] examined the impact of aviation tragedies on the stock prices of the airlines that had encountered a crash together with their rival carriers and found that the afflicted airline experienced deeper negative abnormal returns as the degree of fatality increases. Investigative prognosis by Rose [1990] found that accidents can also have an impact on insurance premiums for airlines. However the financial impacts from the viewpoint of engine manufactures as a result of an accident has not been studied to date. The global engine market is sizeable as the (widebody, narrowbody, regional jets and bizjets) market was worth an estimated $70 billion in 2018 and is expected to be worth almost $1 trillion when including sales between 2014 and 2023 according to a Rolls-Royce global engine forecast. 3,4,5 Therefore, the focus of this research is on this essential linchpin in the aviation value chain in the form of airline engine manufacturers. While social media speculation lingers in the aftermath of an aviation accident, it often encapsulates to scan the wider picture by spreading to all aspects of the commercial operation including engine failure, which triggers a whole set of new interpretations as to what transpired. Engine manufacturers must assert their moral obligation of producing a safe product that is paramount to the commercial success of an aircraft in an industry that is renowned with competitive pressures through cost, efficiency and regulation (Zhang and Gimeno [2016]).
While speculation mounts about the cause of the aviation incident, it is very possible that the powerplants were not the cause, however, the engine manufacturers share price could theoretically decline due to combined reputational interlinkages. The research also considers that the engines were responsible for the cause of the accident, and that the engine manufacturer would most likely be held to account, justifying a substantially elevated negative share price response for the engine manufacturer when compared to the airline. But in two independently competitive industries, the reputational damage could be enough to trigger irreparable financial and reputational damage. Such an example was observed in the recent duel loss of MH17 and MH370 which led to the nationalisation of Malaysia Airlines ).
This research sets out to answer a number of key questions. First, a thorough analysis of all aviation disasters was conducted through an international database to identify the key statistics associated. The study then focused on the financial stock market interactions between the aircraft companies and their selected engine manufacturer that considered the cumulative abnormal returns after an accident when compared to the market index over a specific time period. Such interactions are then analysed in detail to understand as to whether there exists variation by type of engine and by type of airframe manufacturer, while the volatility of the stock returns is also measured. Finally, a number of robust tests are applied to validate the results.
The results provide a variety of outcomes based on the varying behaviour and interactions between aviation companies in periods of crises. First, when analysing the effects of airline disasters on engine manufacturer's profitability and financing structure, the associated findings reveal that the net income falls sharply in the months following a major aviation disaster that involves an aircraft that utilises an engine created by one of the engine manufacturers in our sample. Further, there is evidence of a dampening effect on financial leverage of the company. These results are found to be robust across a variety of methodological structures, indicating that engine manufacturers have to contend with substantial income and financial leverage issues in the aftermath of a major aviation disaster, irrespective of whether they have been identified as a causation factor in the incident itself. The second stage of the research focuses on the calculation of cumulative abnormal returns that exist for engine manufacturers in the period after an aviation disaster, where we identify that there exists a sharp one day loss of 1.64% in the aftermath of aviation incidents, and is found to gradually extend such poor performance in each of the following analysed periods thereafter. In the six months thereafter, the calculated CARs of the engine manufacturers of these airlines are found to have under-performed the market index by over 6.5%. In the third stage of the analysis, the changes in stock market volatility are analysed as a result of these disasters. Companies such as General Electric, Rolls Royce, Textron and United Technology Corporation are found to exhibit quite a substantial increase in share price volatility during the period immediately following an aircraft incident that included aircraft using their components. The authors conclude the research by focusing on the interaction between the returns of the engine manufacturers and the aviation sector in the aftermath of the aviation disasters. When focusing on the inherent contagion effects of these incidents, there is evidence to suggest that the largest events relating to companies such as General Electric, Rolls Royce and United Technology Corporation, all of which possess significant links with Airbus and Boeing, are collectively found to possess gradually elevated dynamic correlations with the aviation sector in the sixty day period after such aviation disaster. This suggests that the immediate price volatility particularly targets the designated engine manufacturer as news and speculation develops and spreads about the related tragedy. However, over a brief period of time, there is evidence to suggest that contagion effects take hold as speculation diminishes and more factual evidence arrives.
The rest of this paper is as follows. Section 2 presents a thorough review of the literature relating to the interlinkage between aviation disasters and airline engine manufacturers. Section 3 presents a concise overview of the data used in this research along with the various methodologies employed to capture firm-level volatility, both intra-sectoral and geographic volatility transmission, and indeed contagion effects by type of aviation incident. Section 4 presents a concise overview of the results presented, while Section 5 concludes.

Previous Literature
Fuel is an airlines highest cost item representing 23.5% of the total operating costs, whose expenditure amounted to $180 billion for the industry in 2018. Although substantial, these numbers are in fact already much lower than the historical highs in 2013 (33% and $231 billion). The wafer thin margins of the airline industry were evident throughout 2018 as it generated just $6.12 profit per passenger, indicating the importance of efficiency and cost containment of each element of operating an aircraft and particularly that of engine fuel burn as reducing fuel consumption is a unifying goal across the aviation industry. 6 Engine technology is continuously improving as Hu et al. [2018] found that the average fuel burn of new aircraft fell approximately 45%, or a compound annual reduction rate of 1.3% between 1986 and 2014. The linkages between fuel (or, more generally, operating costs) and an airline profitability have been touched upon through a variety of papers, including Kang and Hansen [2018]; Heshmati and Kim [2016b]; Heshmati and Kim [2016a]; Kwan and Rutherford [2015];De Poret et al. [2015]; Zou et al. [2014]; Morrell [2011] and Swan and Adler [2006]. The Industry has two key manufacturers of large engines, General Electric and Rolls Royce, while three key manufacturers reside in the narrowbody or regional jet market, notably General Electric, Safran and Pratt & Whitney. These power-plants equip the seven original equipment manufacturers in these segments (Airbus, Boeing, Bombardier, Embraer, Mitsubishi, COMAC and Sukhoi). The developmental cost of a new Jet engine is sizeable, costing between $1 and $2 billion, while a similar amount is required for plant capital and sales concessions, creating high barriers to industry. Product life cycles tend to be long, at 25 to 35 years in duration, with a stream of services or spares revenues attached to each new engine worth 3 to 5 times of its value. These initial power-plants are typically sold at less than unit manufacturing costs while cash flow remains negative for 10 to 15 years (Epstein [2014]).
Much of the literature encapsulating the economics of airline operations bundles the following entities as one unit of aircraft performance: the airframe type including its aerodynamic envelope; maximum takeoff weight; engine performance; as the overall entity that determines fuel burn, which epitomises its technical cost efficiency. Studies have traditionally sought to establish how technical efficiency and productivity have evolved over time (Good et al. [1995]; Oum and Yu [1995]). Other studies investigate airline cost efficiency (Oum and Zhang [1991]; Oum and Yu [1998]) or both productivity and cost competitiveness (Oum and Yu [2012]; Windle and Dresner [1992]; Windle [1991]).
The literature dedicated to the analysis of airline financial performance focuses on efficiency and airline productivity (Schefczyk [1993]; Gittell et al. [2004]; Tsikriktsis [2007]; Barros and Peypoch [2009]). The literature is replete with references to airline safety. Airline accidents represent a perilous impass for the involved party, as they have a significant impact on the demand for air travel, which ultimately affects the finances of an airline, compounding its share price volatility.
Borenstein and Zimmerman [1988] estimated that the loss in enplanement as a result of a crash accrued to about 10-15% of one months worth of traffic. A majority of extant studies on the effect of profitability on safety takes place in the transportation industries. For example, Golbe [1983] examined the connection between profitability and safety in the US railroad industry and found a positive association between contemporaneous profitability and safety -such that railroads are more profitable when they are involved in fewer accidents. However, Golbe [1986] subsequent analysis of the pre-deregulation US airline industry finds no significant relationship between airline profitability and contemporaneous safety. Rose [1990] revisits the profit-safety link for US airlines, finding a marginally significant positive relationship between airline profitability and safety for all airlines and a stronger positive relationship for smaller airlines. Rose improves on the methods employed by Golbe [1983] and Golbe [1986] by utilising lagged profitability measures, eliminating the possibility of reverse causality whereby the observed relationship between profitability and safety is driven by the effect of an accident on its profitability rather than the effects of profitability on safety processes. Raghavan and Rhoades [2005] concurs with the findings of Rose. Both Borenstein and Zimmerman [1988] together with Mitchell and Maloney [1989] analysed changes in equity value following accidents and found evidence that airlines that experienced fatal accidents were subsequently penalised by modest profitability declines. Subsequently, the literature on airline safety is highly influenced by event studies assessing the impact of crashes on stock market prices (see for example Bosch et al. [1998], Borenstein andZimmerman [1988], Chalk [1986]; Chalk [1987]; Mitchell and Maloney [1989]). Airline's parent company was mistakenly identified as a new bankruptcy filing, causing a 76% fall in the company's share price, but after the case was identified as an error, the stock remained over 11% below opening prices, as the authors identify that contagion effects would dominate competitive effects. Luo [2007] used longitudinal real-world data set that matches consumer negative voice (complaint records) in the airline industry with firm stock prices, this article finds that higher levels of current consumer negative voice harm firms' future idiosyncratic stock returns. Guzhva et al.
[2010] assess the market valuation of airline convertible preferred stocks to suggest that airlines undervalue by approximately 10% when they raise capital by issuing convertible securities. Hung and Liu [2005] use the beta value, an indicator of systematic risk, to estimate the costs of equity and the evaluation of a stock's reasonable price, to find that airline betas are volatile over time and that crashes and stock market trends may also impact them.
A number of previous pieces of research focused on the measurement of share price volatility effects outside of the aviation industry. Such research develops on the work of French et al. [1987] who examined the relation between stock returns and stock market volatility to find evidence that the expected market risk premium (the expected return on a stock portfolio minus the Treasury bill yield) is positively related to the predictable volatility of stock returns. While Schwert [1989] found stock market volatility changed over time, while Campbell and Hentschel [1992] developed a formal model of this volatility feedback effect using a simple model of changing variance to find that volatility feedback normally has little effect on returns, but it can be important during periods of high volatility. Marais and Bates [2006] and Sander and Kleimeier [2003] focused on the contagion effects of the Asian financial crisis of the late-1990s, while Arghyrou and Kontonikas [2012], Kenourgios et al. [2011], Samarakoon [2011], Samitas and Tsakalos [2013] and Philippas and Siriopoulos [2013] focused on the contagion effects from both the US and European elements of the subprime and European financial crises. Koutmos and Booth [1995] and Ramchand and Susmel [1998] identified that volatility spillovers in a given market are much more pronounced when the news arriving from the last market to trade is bad, while the correlations between the US and other world markets are on average 2 to 3.5 times higher when the US market is in a high variance state as compared to a low variance regime. Veronesi [1999] found that in equilibrium, investors' willingness to hedge against changes in their own 'uncertainty' on the true state makes stock prices overreact to bad news in good times and under-react to good news in bad times. Bekaert and Wu Focusing on the use of methodologies to specifically analyse the contagion effects of such airline disasters and the interactions between the airlines and the engine manufacturers, the authors of this paper developed a preferred analysis while considering a number of previous works. Forbes and Rigobon [2002] were amongst the first to tests for contagion based on correlation coefficients which they defined as a significant increase in market comovement after a shock to one country.
Such correlation coefficients were found to be conditional on market volatility, where the authors identify a high level of market comovement during periods of crisis. Bae et al. [2003] based their measure of contagion by capturing the coincidence of extreme return shocks across countries within a region and across regions find that contagion is predictable and depends on regional interest rates, exchange rate changes, and conditional stock return volatility in emerging markets during the 1990s.
Such contagion was found not just to be identified in stock markets. Longstaff [2010] identified contagion effects in markets for subprime asset-backed collateralised debt obligations (CDOs) and their contagion effects on other markets, while Sadorsky [2012] found evidence of correlations and volatility spillovers between oil prices and the stock prices of clean energy companies. Swidan and Merkert [2019] investigated the relative effect of operational hedging on airline operating costs, presenting evidence of reduced operating costs, but also providing the caveat that such decision-making does not, in isolation, oprovide an optimal strategy to manage jet fuel risk exposure. Berghöfer and Lucey [2014] studied the extent of operational and fuel hedging in airlines, identifying that neither financial nor operational hedging is effective and that fleet diversity matters greatly. Corbet et al.
[2020] investigated the effects of negative WTI on hedging and portfolio dynamics. Hsu et al. [2008] proposed a class of new copula-based GARCH models for the estimation of the optimal hedge ratio and compare their effectiveness with that of other hedging models, including the conventional static, the constant conditional correlation (CCC) GARCH and DCC-GARCH to find performance improvements in comparison to other dynamic hedging models. El Hedi Arouri et al.
[2011] used a generalised VAR-GARCH approach to examine the extent of volatility transmission between oil and stock markets in Europe and the United States at the sector-level to identify the existence of significant volatility spillover between oil and sector stock returns.

Data
The analysis initiated by constructing a concise list of airline disasters that can then be utilised in a thorough and robust methodological investigation with engine manufacturers corporate accounts and stock market performance. The authors establish as to whether there exist specific speculative financial punishment for engine manufacturers in the aftermath of an airline disaster. As earlier hypothesised, broad speculation based on the cause of such an airline disaster can manifest through many forms, but direct financial punishment due to investor perceptions can present a number of damaging side-effects for the broad aviation sector. To develop such a dataset, a number of strict rules are developed in an attempt to standardise the process across major international financial markets. The first implemented rule is that the specified company must be a publicly traded company with an available stock ticker between the period June 1, 1995 and May 31, 2019. This specific time period is identified due to the relative absence of concise financial market in the period before.
There are two specific methodological approaches developed within this research. Firstly, an investigative approach is instigated to determine if there exists a direct effect within the company as measured by its internal financial performance across a number of theoretically supported, robust methodologies, while also considering the manner in which the financial performance of the engine manufacturers change. Secondly, there is a focus on the financial performance and investor perceptions of the events through a thorough analysis of share price volatility of the same manufacturers, along with a thorough analysis of the contagion effects of such volatility as sourced from the airline industry upon the manufacturers of aviation engine technologies. The selected corporate account data is taken from Bloomberg. Stock price data is taken from Thomson Reuters Eikon. The second news selection rule is based on the source of the data. The authors develop on a combined search of LexisNexis, Bloomberg and Thomson Reuters Eikon, search for the keywords relating to aviation disasters. For added robustness of the developed dataset, the authors leverage upon that of the National Transportation Safety Board (available at: https://www.ntsb.gov), the International Civil Aviation Organisation, ICAO (available at: https://www.icao.int) and the Aviation Safety Reporting System (available at: https://asrs.arc.nasa.gov). To obtain a viable observation, a single observation must be present across each of the selected search engines and the source was denoted as an international news agency, a mainstream domestic news agency or the company making the announcement itself. Forums, social media and bespoke news websites were omitted from the search.
Finally, the selected observation is based solely on the confirmed news announcements being made on the same day across all of the selected sources. If a confirmed article or news release had a varying date of release, it was omitted due to this associated ambiguity. All observations found to be made on either a Saturday or Sunday are denoted as active on the following Monday morning. All times are adjusted to GMT, with the official end of day closing price treated as the listed observation for each comparable company when analysing associated contagion effects.
The associated summary statistics for the selected engine and engine component manufacturers are presented in Table 1. Three control variables have been selected to act as a representation of international effects. The Dow Jones International Average (DJIA) is used to control for financial market performance during the period under investigation. While considering a number of international airlines and engines manufacturers, the study also accounts for financial growth and crises during the period. This is further accounted for through the use of West Texas Intermediate (WTI) oil prices and the S&P500 Aviation Index. The aviation sector is found to present the largest mean returns during the period between 1995 and 2019. WTI is found to possess the largest standard deviation of returns (0.0231) of the selected independent variables along with the two most extreme one-day price increase and decrease of 17.83% and 15.25% respectively. The aviation index is found to be the most negatively skewed (-0.1486) and also possess the largest level of kurtosis (9.1538).
While focusing on the examined engine manufacturers, there are a number of particularly significant events and results. The largest mean return is exhibited by MTU Aero Engines (+0.0008), while Textron exhibits the largest standard deviation with regards to the selected companies (+0.0241).
The largest one-day fall in the share price of an engine manufacturer was experienced by that of Safran (-33.3%), closely followed by that of Textron (-31.6%). The largest one-day price increase was experienced by Textron (+48.9%), presenting evidence of the particularly volatility behaviour of this share price. The second largest one-day share price increase was experienced by Honeywell International (+26.2%). The most negatively skewed company returns are that of BAE Systems Insert Figure 2 and Table 2 about here   Table 3

Insert Tables 3 and 4 about here
Considering that the market for airline travel is an extremely competitive industry, any broad error of substantial loss of credibility and reputation can lead to quite a disastrous financial outcome. Table 4, represents a list of airlines that are no longer in operation today, in the aftermath of a substantial aviation disaster. It is also quite important to note that there are also a number of such companies that have actually experienced multiple events. Adamair is found to have had three substantial events leading to the lost of 102 passengers. Prior to the crash of Adamair Flight 574, it was the fastest growing low-cost carrier in Indonesia. On March 16, 2007, the Indonesian government grounded the carrier as it was deemed mismanaged which was further compounded by embezzlement proceedings. Hewa Bora Airways also experienced three separate incidents before operations were suspended after flight 952 crashed in 2011. This carrier had undergone a series of mergers and management changes (FlyCongo, Compagnie Africaine d'Aviation, Zaire Airlines, and Congo Airlines). The third company that experienced three incidents that forced its grounding was that of Rico Linhas Aereas S/A, which was a Brazilian regional airline authorised to operate scheduled passenger and cargo services in the Amazon region. Insert Tables 5 and 6 about here  Table 6, denotes the summary statistics for each major crash associated by engine manufacturer that was attached to downed airliner. United Technologies Corporation and Rolls Royce are found to be the leading engine manufacturer for 160 and 157 of the identified accidents respectively, obtained during the analysis. Furthermore, General Electric provide leading engines for use in the CF6 engine-series which were mounted on the wings of Airbus A300, A310, A330, B747 and B767, while the GE90 series are used in B777 and the next generation of engines, termed the GEnx series are fitted on the Boeing 787 and 747-8i, which collectively accounted for 30 significant accidents that resulted in the deaths of 1,269 passengers. Pratt & Whitney's main engines (United Technologies Corporation is the associated parent company) include the JT8D series which were the powerplants for early Boeing 737s, the JT9D series were attached to Airbus A300, A310, Boeing 767, 747, the PW2000 series thrusted the Boeing 757, while the PW4000 series were used in Airbus A300, A310 and A330 as well as the Boeing 767, 747 and 777, while the smaller PW6000 series powered the Airbus A318. Meanwhile Rolls-Royce produced the RB200 series engine which was used in Boeing 747 and 757, the Trent 500 series was used in later Airbus A340, the Trent 700 series powered the Airbus A330, while the Trent 800 series were used in Boeing 777, the Trent 970 series was attached to the Airbus A350 and A380, while the newer Trent 1000 series is used in Boeing 787s.
Between these three companies, there has also been a number of joint-venture collaborations that have been at the forefront of engine manufacturing development. Firstly, the CFM International joint venture between General Electric and Safran colluded to produce the CFM56 series widely used in Airbus A318/A319/A320/A321, early A340, and later Boeing 737s. Secondly, the Engine Alliance was a joint venture between General Electric and Pratt & Whitney which produced the GP7000 series that is used for the Airbus A380. Finally, one of the largest Joint ventures was based on that of International Aero Engines, who formed a consortium that encompassed Pratt & Whitney, Rolls-Royce, Japanese Aero Engine Corporation and MTU. This engine was used in the IAE V2500 series and powered the Airbus A319/A320/A321. Other significant engine component manufacturers that experienced severe aviation disasters were that of BAE systems, Lockheed Martin and Textron.

Methodology
While it is of course expected that the airline that has been directly affected by a major disaster would have substantial financial ramifications, both in terms of immediate internal monetary damage through loss of earnings through reduced passenger flows and by the exceptional costs associated thereafter, without any direct evidence of potential causality, the engine manufacturers would be expected to be somewhat buffered from such attributed responsibility. Any evidence of a substantial financial shock to engine manufacturers without firm evidence provided would be considered to be generated by highly speculative investor behaviour. This very situation could potentially generate broad financial repercussions for the aviation sector at large through increased costs, regulatory misalignment and increased exposure to external shocks due to excessive competitive pressures.
There are several ways through which the unfortunate aviation disaster experienced by a company could have an impact on that company's financial variables, both on the book-value, and more immediately in the financial markets. In particular, perceptions of reduced future sectoral growth by the airline sector would manifest in quite a negative outlook for the engine manufacturers, which would be subsequently reflected through decreased profitability and leverage on the financial statements of the company. Moreover, due to the substantial negative publicity and perceptions of wrong-doing, which could possibly be present without due cause, the company's share price might experience dramatic decreases in the short through medium term. This could be expected to manifest in sharp increases in stock market volatility and interactions with other market variables through the spread of financial market contagion. Contagion effects are also considered as this can be the source of inter-sectoral risk transfer between the aviation sector and that of engine manufacturers. All these arguments will be analysed thoroughly in the rest of the paper using the following methodologies.

The impact of aviation disasters on firms' profitability and financing structure
In this part of the analysis, the authors aim to investigate as to whether engine manufacturer's next term profitability has directly been influenced by aviation disasters. Furthermore an examination seeks to determine if the firms have had to manipulate their use of financial leverage so as they can mitigate such financial difficulties. Both of these scenarios are investigated with emphasis on the fact that the engine manufacturer has not been held responsible in any way for the related airline disaster with the exception that they had supplied the downed airline with its engines and is perceived by investors to be potentially responsible. The motivation for these hypotheses comes from the fact that the company is potentially about to enter a highly speculative and sensitive period of time where blame could indeed be attributed. Therefore, profitability of the company might fall substantially and its access to debt financing might become substantially more difficult to obtain as lenders observe substantially elevated levels of risk. To work on the above mentioned hypotheses, quarterly balance sheet and income statement data are used for the period between Q3 of 1995 and Q2 of 2019, adding up to twenty-eight quarters in total. All data comes from Thomson Reuters Datastream. F L t represent financial leverage calculated by company's debt to equity at the end of quarter t and N I t represents the net income in quarter t. Control variables are constructed subject to three sets of stock characteristics. All variables are updated every quarter unless otherwise stated. To examine the effects of such incidents on a company's profitability (H 1 ) and financial leverage (H 2 ), the authors run the following panel regressions in equation (1) through equation (3) using two-way clustered standard errors: In this model, Y is either the N I or F L and X t is the vector of 12 stock characteristics described above, as separated as β 1 Hist t for the group relating to historical return patterns, β 2 Liq t which represents the variables analysing liquidity and transaction costs, and β 3 P rud t which analyses corporate prudence. On the other hand, Incident t is i) a dummy variable taking the value unity in the period after the accident in quarter t. In each of the twenty-eight quarters, the authors winsorize all variables at the level of 95% (both dependent and independent) with the exception of the index and incident dummies in the cross-section by two levels from both upper and lower tails to eliminate the outlier effect without removing any observations.
3.3. Does there exist an engine manufacturer share price discount due to aviation disasters? and ). The analysis is partitioned by each publicly traded company that provided engines to the airlines of interest. It is also of interest to examine as to whether there exists persistence of any identified share price discounts in such circumstances. We select a variety of time-frames on which to analyse such effects. The authors investigate the 1-, 5-10-, 20-, 30-, 60-, 90-and 180-day periods after each identified events to identify as to whether such engine manufacturer share price discounts persist while the investigation into the cause of the airline disaster takes place, or indeed, as to whether such effects are found to dissipate within this time-frame. Evidence of such a share price discount is particularly damaging across a number of areas. First, it must be noted that engine manufacturers are heavily reliant on external financing while researching and developing future engine technology. The impact of such financial pressures and reputational damage in the aftermath of such airline incidents can greatly reduce the availability of such financing. Further, the existence of such a share price discount would also strongly indicate that engine manufacturers have been greatly influenced by speculative investors, who are most likely attempting to invest on 'potential' developing stories rather than on a factual basis. Such speculative investment is by no means illegal, but it has been dealt with in a variety of manners over time, such as the implementation of 'short-sale bans' on banking shares during the most stressful and volatile periods of the most recent international financial crisis.
3.3.1. The volatility effects of aviation disasters within the sector for engine manufacturers GARCH-based methodologies are found to be one of the primary mechanisms through which one can observe direct volatility effects in financial markets, allowing for both comparison and scalability of parameters. Developing from key analyses that focused on issues such as the GARCH (1,1) quasi-maximum likelihood estimator (Lee and Hansen [1994]) and the consistency and asymptotic normality of such models (Lumsdaine [1996]), Tse and Tsui [2002] were the first to propose a Such terrorism impacts were also found to be both significant and substantial when considering the persistence of their effects at both sectoral and national levels (Kolaric and Schiereck [2016]; Carter and Simkins [2004]; Kim and Gu [2004]; Corbet et al. [2018]).
In this analysis, we undertake specific investigation is undertaken to determine if there exists a substantial change in stock price volatility of the engine manufacturers due to their direct association with the airline company that has experienced a severe accident. Analysis is conducted to decipher if the associated volatility in the periods both before and after the designated announcement of a crash presents evidence of substantial change. Firstly, statistical tests are applied to determine if there is an increase in unconditional variance of the company stocks' daily returns (and the corresponding excess returns over the market they are traded in) after announcements for various time periods by utilising a common variance inequality test. Secondly, more penetrative scrutiny is enforced by building upon the GARCH-family to understand the volatility dynamics of crypto-exuberance based on naming behaviour in the conditional variances. At this stage, a number of goodness-of-fit testing procedures identified the EGARCH(1,1) model as the best selected to identify specific volatility changes in the companies' returns, thus we exercise our analysis using this model. 7 The variance 7 EGARCH exploits information contained in realised measures of volatility while providing a flexible leverage function that accounts for return-volatility dependence. While remaining in a GARCH-like modelling framework equation of our EGARCH model is expressed as follows: Here, an additional D t term is included in equation (4) in the analysis to provide a coefficient relating to the observed volatility in the subsequent days following each event for each of the investigated companies. Before proceeding with the EGARCH analysis, exogenous effects are migrated which can be completed through the inclusion of the returns of traditional financial products in the mean equation of the EGARCH(1,1) methodology as displayed in equation (5).
The volatility sourced in shocks that are incorporated in the returns of traditional financial markets are therefore considered in the volatility estimation of the selected structure. In equation (5), R t−1 represents the lagged value of the observed company returns. DJIA t is the returns of the benchmark index where the stock is traded, and represents the interaction between the selected company returns and the corresponding market index. W T I t represents the interactions between the selected company and the market for West Texas Intermediate (WTI) oil. Such interactions are very much of interest from a methodological standpoint due to the observed interactions between the market for oil and international crisis and instability. Therefore, the selection of WTI in this methodological structure allows the GARCH-family methodologies to incorporate and account for such instability. Similarly, the addition of the variable AV I t incorporates broad trends and periods of instability within the broad aviation sector.

Analysing potential contagion effects: A DCC-EGARCH methodology
The final stage of our selected analysis develops on the channels through which such airline sector volatility influences that of the engine manufacturers in the periods after such incidents.
While no particular causality has been identified throughout a number of the selected cases in our above analysis, it is very much of interest in this section to identify and observe as to whether engine manufacturers are found to experience direct contagion flows from the airline industry, and indeed, as to how long these particular periods of contagion persist. It is important to identify as to whether companies who had traditionally little or no direct involvement with such incidents had observed structural changes in their previously outstanding interactions. For example, a company who had changed their name from that which did not identify as a blockchain or cryptocurrencyrelated company to one that does thereafter, could therefore theoretically experience a sharp change and estimation convenience, the model allows independent return and volatility shock and this dual shock nature leaves a room for the establishment of a variance risk premium. In our selection, other competitive models included EGARCH, TGARCH, Asymmetric Power ARCH (APARCH), Component GARCH (CGARCH) and the Asymmetric Component GARCH (ACGARCH). The optimal model is chosen according to three information criteria, namely the Akaike (AIC), Bayesian (BIC) and Hannan-Quinn (HQ).
in dynamic correlations, perhaps also indicative that the company is being treated differently by investors in the aftermath of changed perceptions of corporate risk-tolerance ).
For example, it is important to specifically analyse if these investors, who could perceive that these companies have changed in terms of their perceived high-risk behaviour. To consider the contagion effects, the popular dynamic conditional correlation (DCC) model of Engle [2002] is applied. The authors firstly let r t = [r 1,t , ..., r n,t ] be the vector of financial time series returns and ε t = [ε 1,t , ..., ε n,t ] be the vector of return residuals obtained after some filtration. Let h i,t be the corresponding conditional volatilities obtained from a univariate EGARCH process. Assume is the conditional expectation on ε t , ε t−1 , .... The asset conditional covariance matrix H t can be written as where R t = [ρ ij,t ] is the asset conditional correlation matrix and the diagonal matrix of the asset conditional variances is given by D t = diag(h 1,t , ..., h n,t ). Engle [2002] models the right hand side of Eq.(6) rather than H t directly and proposes the dynamic correlation structure where Q t ≡ [q ij,t ], u t = [u 1,t , ..., u n,t ] and u i,t is the transformed residuals i.e. u i,t = ε i,t /h i,t , S ≡ [s ij ] = E[u t u t ] is the n × n unconditional covariance matrix of u t , Q * t = diag{Q t } and a, b are non-negative scalars satisfying a + b < 1. The parameters of the DCC model are estimated by using the quasi-maximum likelihood method with respect to the log-likelihood function, and according to the state two-step procedure.

Impact of airline disasters on firms' profitability and financing structure
The estimated coefficients are reported of equation (1) Table 7, it is noticeable that whether a sub-group of control variables is applied or all applied at the same time, there is consistency in the signs of the explanatory variables with a few exceptions for both net income and financial leverage analysis. This shows the reliability of the results. According to the M all analysis, except the insignificant market cap and short-term momentum variables, effect of all variables on profitability differ with respect to either sign or insignificance between the types of incide. In this analysis, the main variable of interest is the dummy variable γ and according to our findings, the coefficient of this dummy indicates that net income falls sharply in the aftermath of the months following a major aviation disaster that involves an aircraft that utilises an engine created by one of the engine manufacturers in our sample. Further, there is evidence of a dampening effect on financial leverage of the company. This is quite an interesting result, as in the majority of the cases sampled, there is no distinct evidence provided during this time period to indicate that the engines on the aircraft were in fact responsible for the accident. Regarding the determinants of financial leverage, both methodological structures present common factors. For example, according to the M all column, variables such as market cap, book-to-market ratio, short-term momentum, beta, stock price, age and volatility have no significant impact on the leverage; whereas turnover and Amihud ratio are both significant and have impact in the same direction.

Robustness tests based on profitability and financing structure results
To check the robustness of the finding above, the authors try alternative variations of equations (1) through (3) and start with estimating the following equations (8) through (10). The motivation is to bring a dynamic perspective to the analysis by focusing on not the levels but the change in the levels of net income and leverage.
The robustness analysis is taken to another level and instead of a separate analysis for the combined financial effects, an analogous regression is employed of equations (1) through (3) and (8) through (10) in the new equations (11) and (12), respectively. The results are presented in Table 8. These models allow for control of the multiple types of financial effects at the same time in the aftermath of the identified aviation disasters. Significant results would indicate that there exist corporate financial effects on engine manufacturers without any evidence or allocation of responsibility.
At this stage, to verify the existence of effects on the profitability and financial structure of engine manufacturer companies, the expectations of the dummy coefficient signs in the last two equations above are the same as in the previous analysis, and the findings almost perfectly fit. In the case of equation (11), without an exception, all dummy coefficients are significantly negative for both income and the leverage.
Further verification is provided through the inclusion of equation 12 in Table 8 which provides the full specification methodology using first-differenced net income and financial leverage data for the analysed engine manufacturers, which, despite presenting results that are somewhat below the same analysis results for net income and financial leverage (-0.711 and -0.455 respectively), pertaining to the regression testing and when comparing this to the results of the robustness testing procedure, -0.475 and -0.286 respectively, we find that both outcomes produce the same size and scale, providing substantial support as to the selected methodological procedures used in this analysis. The findings indicate that engine manufacturers have to content with substantial income and financial leverage issues in the aftermath of a major aviation disaster, irrespective of whether they have been identified as a causation factor in the incident itself.

Cumulative abnormal returns (CARs) for engine manufacturers in the aftermath of airline disasters
Next an attempt is formulated to establish how influential aviation disasters were upon the share price performance of the analysed engine manufacturers. To examine this effect we calculate the cumulative abnormal returns (CARs) which are presented in Table 9. First focusing on the overall average of CARs, we clearly identify that there exists a sharp one day loss of 1.64% in the aftermath of aviation incidents. This result is largely driven by the sharp initial one-day fall in the price of BAE Systems (-9.16%), Honeywell International (-2.84%) and Leonardo SpA (-2.38%). Substantial corporate instability is found to persist without the company being in any way responsible for the incident. Shortly thereafter, contagion effects increase as speculation diminishes and more factual evidence arrives.
Insert Table 9 about here There is evidence of a sharp deterioration in the calculated CAR approximately ten days after the event (-1.64%) and this excess loss over the market average continues up to one-hundred and twenty-five days after the event where the calculated CAR is -8.14%. Only in the period thereafter does the calculated CAR improve to approximately -5.4%, but in the subsequent three month period, the market performance deteriorates further in comparison to the market average. One hundred and eighty days after each of our analysed aviation disasters, the calculated CARs of the engine manufacturers of these airlines are found to have under-performed the market index in excess of 5%. While a broad variety of results exist, there exist two companies that out-perform the market index in the six months after the average of the incidents that are analysed.
Insert Figure 3 about here A graphical representation of the daily estimate of the CARs for the overall sample is illustrated in Figure 3. For the overall sample there are multiple distinct, yet sharp periods of deterioration in the performance of the companies when compared to the market index, the first in the forty-five days after the event which is found to result in a loss of 5% approximately. The share price performance falls quite substantially once again between four and six months after the event leading to a point in time estimated loss in excess of 8% when compared to the related domestic market index.

Firm-level volatility by type of airline disaster
The level of conditional volatility on a case-by-case basis is presented with significant results in Table 10. With regards to specific companies, General Electric, Rolls Royce, Textron and United Technology Corporation present evidence of the largest number of significant volatility events during the period analysed. There is also an interesting observation to be made in the wide-ranging reduction in the number of statistically significantly volatility events within the designated time period under observation. It is apparent that while each company with the exception of Rockwell Automation experience significant effects for a number of aviation disasters in an exceptionally short time-frame (such as the day on which the news was announced), the effects appear to subside in the weeks that follow, as evident in the reduced number of statistically significant volatility changes in the periods throughout twenty, forty and sixty days after.
Insert Table 10  evidence of detrimental effects in the day immediately after the incident, conditional volatility is then found to increase somewhat in the sixty-day period thereafter as more robust news continues to disseminate. On a case-by-case analysis, this finding also holds as the largest cases of conditional volatility are not found to be as frequent in the period after the rapid growth of this social media technology. This could indicate that the existence of better quality information could in fact be changing the manner in which volatility and contagion interact between companies.

Volatility transmission between the aviation sector to the engine manufacturer
to conclude the research, the paper focuses on the interaction between the returns of the engine manufacturers and the aviation sector in the aftermath of the aviation disasters. In these selected samples, there has not been any official announcement as to whether there exist a relationship between the engines on the aircraft and the cause of the incident itself. While the aviation sector at large can be susceptible to sharp decreases in valuation in the days following a significant incident, it is quite alarming from the viewpoint of a provider and developer of engine components to be susceptible to such volatility, particularly due to influence that such perceptions and reputational risk can have on the continued development of the associated companies.
Insert Table 11 about here When considering the short term period after the individual event, a selection of multiple time frames is considered including that of the 1-, 5-, 10-, 20-, 40-and 60-day periods. 8 The final column presented in both Table 10 and Table 11 represent the entire period in the aftermath of aviation disasters. When considering the stock return's conditional variance, it can be seen that there are substantially fewer significant results in the period. When analysing the results of the interactions between the dynamic correlations between engine manufacturers and the broad aviation section, one finds evidence of both elevated and consistent relationships between the sample of incidents and the aviation sector throughout the entire sixty-day period after. The results indicate that while a broad number of companies exhibit elevated dynamic correlations, it is those relating to General Electric, Rolls Royce and United Technologies Corporation that are found to generate substantial sector-wide volatility effects. However, when considering only significant results it returns some very interesting findings. In the one-day period after the aviation incident, only 30 events, or 4.7% of engine manufacturer events analysed exhibit elevated levels of dynamic correlation with the aviation sector. Within the five-day period, this increases to 7.0% and then increases again to almost onequarter of the sample ten days after the event (9.7%). This figure is at its largest at twenty days after the event, representing 11.1% of the sample. One potential explanation for this outcome is based on the slow dissemination of news in the aftermath of the events. While the airline stocks themselves are sharply effected by the negative sentiment and reputational damage, it would appear that initially that the engine manufacturers do not experience the same influence. However, within ten days, this decoupling effect subsides and there is a substantial positive correlation identified between the engine manufacturers and the aviation industry. Moreover, this effect tends to be permanent as it does not disappear mostly after a quarter and even longer periods than that.

Concluding Comments
The continued evolution of the aviation sector relies on technological development, with particular emphasis surrounding the response of the industry to the growing calls to immediately deal with the carbon emissions disaster that has been escalating in recent years. While structural design and efficiency improvements will support such efforts, much of the weight of the task will be borne by engine manufacturers. However, one must understand their role within the sector when considering the presence of the effects of contagion risks due to accidents involving the airlines to which these manufacturers supply and the subsequent attributed responsibility for injury and fatality. These manufacturers are not responsible for the many identified shortcomings that have been identified in the sector in recents years with regards to poor maintenance which has been attributed to costcutting effects in an extremely competitive environment. However, when an aviation disaster occurs, there is widespread allocation of blame and responsibility until the true cause is identified. It is very much of interest to understand as to how such contagion effects influence engine manufacturers and as to how effects relating to asymmetric information and moral hazard can influence their profitability and potential for growth.
The results provide a variety of explanation based on the varying behaviour and interactions between these companies in periods of crises. First, when analysing the effects of airline disasters on engine manufacturer's profitability and financing structure, the findings reveal that net income falls sharply in the aftermath of the months following a major aviation disaster that involves an aircraft that utilises an engine created by one of the engine manufacturers in our sample. Further, there is evidence of a dampening effect on financial leverage of the company. These results are found to be robust across a variety of methodological structures, indicating that engine manufacturers have to contend with substantial income and financial leverage issues in the aftermath of a major aviation disaster, irrespective of whether they have been identified as a causation factor in the incident itself. The second stage of our research focuses on the calculation of cumulative abnormal returns that exist for engine manufacturers in the period after an aviation disaster. Focusing on the overall average of CARs, it clearly identifies that there exists a sharp one day loss of 1.64% in the aftermath of aviation incidents. However, the stock price remains substantially below the market average in the five days following the disaster. It is very much of interest to note that one hundred and eighty days after each of the analysed aviation disasters, the calculated CARs of the engine manufacturers of these airlines are found to have under-performed the market index by over 6.5%.
In the third stage of the analysis, the authors evaluated the changes in stock market volatility as a result of these disasters. Companies such as General Electric, Rolls Royce, Textron and United Technology Corporation are found to exhibit quite a substantial increase in share price volatility during the period immediately following an aircraft incident that included aircraft using their components. There is also an interesting observation to be made in the wide-ranging reduction in the number of statistically significantly volatility events within the designated time period under observation. One potential explanation for this outcome is based on the slow dissemination of news in the aftermath of the events. While the airline stocks themselves are sharply effected by the negative sentiment and reputational damage, it would appear that initially that the engine manufacturers do not experience the same influence. However, within ten days, this decoupling effect subsides and there is a substantial positive correlation identified between the engine manufacturers and the aviation industry. Moreover, this effect tends to be permanent as it does not disappear mostly after a quarter and even longer periods than that. When focusing on the inherent contagion effects of these incidents, there is evidence to suggest that the largest events relating to companies such as General Electric, Rolls Royce and United Technology Corporation, all of which possess significant links with Airbus and Boeing, are found to possess gradually elevated dynamic correlations with the aviation sector in the sixty day period after such aviation disaster. This suggests that the immediate price volatility particularly targets the designated engine manufacturer as news and speculation develops and spreads about the related tragedy. However, over a brief period of time, there is evidence to suggest that contagion effects take hold as speculation diminishes and more factual evidence arrives.