# Risk Perceptions on Hurricanes: Evidence from the U.S. Stock Market

^{1}

^{2}

^{3}

^{*}

## Abstract

**:**

## 1. Introduction

## 2. Data and Sample

## 3. Methodological Background

- Identifying the event of interest and the timing of the event;
- Specifying a benchmark model for normal stock returns behavior; and
- Calculating and analyzing abnormal returns around the event date.

- Estimation window: It is the period between T
_{0}and T_{1}. This period comprises 150 trading days. In this period, the Market model is applied for estimating normal returns. - Event window: This period ranges from T
_{1}to T_{2}and t = 0 (the hurricane landfall) is situated in the middle of this period. This is composed of twenty-one trading days: ten before the event date and ten after the event date. - Post-event window: which comprises the period from T
_{2}to T_{3}. This period will be used to prove and check if the daily returns of the companies selected go back to the previous situation before the hurricane.

_{0}and T

_{1}and is composed of 150 trading days. This period, usually designed as benchmark, is necessary to study the normal behavior of the stock prices before the event date.

_{it}is the return of the stock i on Day “t”; R

_{mt}is the return of the market portfolio on Day “t”; α

_{i}is the constant term; β

_{i}is a measure of the sensitivity between R

_{it}with respect to R

_{mt}; and ${\mathsf{\epsilon}}_{\mathrm{it}}$ is the random disturbance term.

_{it}is the daily return P & C Company, and R

_{mt}is the daily return of the S & P 500. The parameters α and β are estimated by Ordinary Least Squares (OLS):

_{mt}is the daily market index (S & P 500) return. $\widehat{{\mathsf{\alpha}}_{\mathrm{i}}}$ and $\widehat{{\mathsf{\beta}}_{\mathrm{i}}}$ are OLS estimates of the regression coefficients. Since stocks returns can exhibit autorregressive conditional heteroscedasticity, we have computed the quasi-maximum likelihood covariances and standard errors as described in [24]. The model is estimated under the assumption that the errors are conditionally normally distributed.

_{0}reflects the observed proportion of positive returns for a given time window. This statistic is distributed as a normal law of variance 1 and mean 0.

## 4. Research Design

_{it}) is by subtracting the current return minus the expected return (see Equation (1)). Then, to compute the global sample, we create a matrix, AR, composed of the abnormal returns of the P & C Insurance Companies for the event window E (−10, +10). The informativeness of the analysis is greatly improved by averaging the information over the sampled firms so the unweighted cross-sectional average of abnormal returns is considered [17]. Thus, we compute the average abnormal return (see Equation (5)). Reaching this point, we conduct a normality test to check whether the sample follow a normal distribution or not. Anderson–Darling (A-D) or Kolmogorov–Smirnov (K-S) tests can be suitable for such analysis. According to Engmann [27] the A-D test requires less data than the Kolmogorov Smirnov test to reach sufficient statistical power. They state that the A-D test is more sensitive to the tails of distributions and it is more reliable than the K-S. Consequently, we have conducted the Anderson–Darling test on the average abnormal return (AAR) of our event window.

## 5. Findings and Results

## 6. Conclusions

## Acknowledgments

## Author Contributions

## Conflicts of Interest

## References

- Fink, J.D.; Fink, K.E.; Russell, A. When and how do tropical storms affaect markets? The case of refined petroleum. Energy Econ.
**2010**, 32, 1283–1290. [Google Scholar] [CrossRef] - Horwich, G. Economic lessons of the Kobe earthquake. Econ. Dev. Cult. Chang.
**2000**, 48, 521–542. [Google Scholar] [CrossRef] - Luo, C.C.; Wu, D. Catastrophe risk analysis: A financial perspective. Hum. Ecol. Risk Assess.
**2013**, 19, 1372–1384. [Google Scholar] [CrossRef] - West, C.T.; Lenze, D.G. Modeling the regional impact of natural disaster and recovery: A general framework and an application to hurricane Andrew. Int. Reg. Sci. Rev.
**1994**, 17, 121–150. [Google Scholar] [CrossRef] - Ewing, B.T.; Hein, S.E.; Kruse, J.B. Insurer stock price responses to hurricane Floyd: An event study analysis using storm characteristics. Weather Forecast.
**2006**, 21, 395–407. [Google Scholar] [CrossRef] - Lamb, R.P. An examination of market efficiency around hurricanes. Financ. Rev.
**1998**, 33, 163–172. [Google Scholar] [CrossRef] - Angbazo, L.A.; Narayanan, R. Catastrophic shocks in the property-liability insurance industry: Evidence on regulatory and contagion effects. J. Risk Insur.
**1996**, 63, 619–637. [Google Scholar] [CrossRef] - Hallegatte, S.; Dumas, P. Can natural disasters have positive consequences? Investigating the role of embodied technical change. Ecol. Econ.
**2009**, 68, 777–786. [Google Scholar] [CrossRef] - Born, P.; Viscusi, W.K. The catastrophic effects of natural disasters on insurance markets. J. Risk Uncertain.
**2006**, 33, 55–72. [Google Scholar] [CrossRef] - Strobl, E. The economic growth impact of natural disasters in developing countries: Evidence from hurricane strikes in the Central American and Caribbean regions. J. Dev. Econ.
**2012**, 97, 130–141. [Google Scholar] [CrossRef] - Hewitt, E.R. Examining Market Response Following Hurricane Landfall: Does the U.S. Stock Market React Efficiently to Hurricanes? Economics Honors Thesis, Macalester College, St. Paul, MN, USA, 2012. [Google Scholar]
- Campbell, C.J.; Wasley, C.E. Measuring abnormal daily trading volume for samples of NYSE/ASE and NASDAQ securities using parametric and nonparametric test statistics. Rev. Quant. Financ. Account.
**1996**, 6, 309–326. [Google Scholar] [CrossRef] - Kothari, S.; Warner, J.B. Handbook of Empirical Corporate Finance; Elsevier: Amsterdam, The Netherlands, 2007. [Google Scholar]
- Fama, E.F. Efficient capital markets: A Review of theory and empirical work. J. Financ.
**1970**, 25, 383–417. [Google Scholar] [CrossRef] - Bowman, R.G. Understanding and conducting event studies. J. Bus. Financ. Account.
**1983**, 10, 561–584. [Google Scholar] [CrossRef] - Kothari, S.P.; Warner, J.B. Handbook of Corporate Finance, 1st ed.; Eckbo, B.E., Ed.; Imprint: North Holland, The Netherlands, 2007; Volume 1. [Google Scholar]
- Peterson, P.P. Event studies: A review of issues and methodology. Q. J. Bus. Econ.
**1989**, 28, 36–66. [Google Scholar] - Brown, S.J.; Warner, J.B. Using daily stock returns. J. Financ. Econ.
**1985**, 14, 3–31. [Google Scholar] [CrossRef] - Martin Curran, M.; Moran, D. Impact of the FTSE4 GOOD Index on firm price: An event study. J. Environ. Manag.
**2007**, 82, 529–537. [Google Scholar] [CrossRef] [PubMed] - Sturm, P. Operational and reputational risk in the European banking industry: The market reaction to operational risk events. J. Econ. Behav. Organ.
**2013**, 85, 191–206. [Google Scholar] [CrossRef] - Fiordelisi, F.; Soana, M.-G.; Schwizer, P. The determinants of reputational risk in the banking sector. J. Bank. Financ.
**2013**, 37, 1359–1371. [Google Scholar] [CrossRef] - Sharpe, W.F. A Simplified model for portfolio analysis. Manag. Sci.
**1963**, 9, 277–293. [Google Scholar] [CrossRef] - MacKinlay, A.C. Event studies in economics and finance. J. Econ. Lit.
**1997**, 35, 13–19. [Google Scholar] - Bollerslev, T.; Wooldridge, J.M. Quasi-maximum likelihood estimation and inference in dynamic models with time-varying covariances. Econom. Rev.
**1992**, 11, 143–172. [Google Scholar] [CrossRef] - Wilcoxon, F. Individual comparisons by ranking methods. Biom. Bull.
**1945**, 1, 80–83. [Google Scholar] [CrossRef] - Strong, N. Modelling abnormal returns: A review article. J. Bus. Financ. Account.
**1992**, 19, 533–553. [Google Scholar] [CrossRef] - Sonja Engmann, D.C. Quantitative methods in audit and control. J. Appl. Quant. Methods
**2009**, 4, 408–429. [Google Scholar] - Luoma, T. Nonparametric Event Study Tests for Testing Cumulative Abnormal Returns. Master’s Thesis, University of Vaasa, Vaasa, Finland, 2011. [Google Scholar]
- Lynch, M. Reactions to Hurricane Katrina; Bank of America Corporation: New York, NY, USA, 2015. [Google Scholar]
- Baker, C.R. Breakdowns of accountability in the face of natural disasters: The case of hurricane Katrina. Crit. Perspect. Account.
**2014**, 25, 620–632. [Google Scholar] [CrossRef]

Name | Event Date * | Category (By Saffir-Simpson Hurricane Wind Scale (SSHWS)) | Damage (in USD) |
---|---|---|---|

Katrina | 23–31 August 2005 | 5 | 148 Billion |

Rita | 18–26 September 2005 | 5 | 12.037 Billion |

Felix | 31 August–5 September 2007 | 5 | 850 Million |

Ike | 1–14 September 2008 | 4 | 29.5 Billion |

Igor | 8–23 September 2010 | 4 | 25 Billion |

Ophelia | 20 September–3 Octorber 2011 | 4 | 21 Billion |

Sandy | 22–29 Octorber 2012 | 3 | 71 Billion |

Ticker | Company | Total Revenues * (in USD) |
---|---|---|

MMC | Marsh & McLennan Companies | 12,261 Billion |

PGR | Progressive Corporation | 18,170 Billion |

ACE | Ace Limited | 19,261 Million |

ALL | The Allstate Corporation | 34,507 Million |

CB | The Chubb Corporation | 13,502 Million |

TRV | Travelers | 26,191 Million |

BRK-B | Berkshire Hathaway’s | 182,150 Million |

Category | Wind Speed | Characteristics |
---|---|---|

1 | 119–153 km/h | Very dangerous winds. Extensive damage to power lines and poles. Large branches of trees will snap and shallowly rooted trees may be toppled. |

2 | 154–177 km/h | Extremely dangerous winds. Well-constructed frame houses could suffer damages in roof and siding damages. |

3 (major) | 178–208 km/h | Devastating damage will occur. No water or electricity services available. Houses will suffer damage or removal of roof docking and gable ends. Trees will be uprooted. |

4 (major) | 209–251 km/h | Catastrophic events will occur. Damages on roof structures and some exterior walls. Trees and power poles downs. Power outages for weeks to months. The area will be uninhabitable for weeks or months. |

5 (major) | 252 km/h or more | Catastrophic damage will occur. High percentage of homes destroyed. Isolation of residential areas due to fallen trees and power poles. Area uninhabitable. |

Hurricane | Main Areas Affected |
---|---|

Katrina | New Orleans and Mississippi coast |

Rita | Texas, Louisiana and Florida Keys |

Felix | Netherlands Antilles and Nicaragua |

Ike | Caribbean, Texas and Louisiana |

Igor | Bermuda and Newfoundland |

Ophelia | Bermuda and Leward Island |

Sandy | Jamaica, Cuba and Bahamas |

Hurricane Katrina | |||||

Sample Size | 21 | ||||

Statistics | 0.41432 | ||||

Rank | 14 | ||||

Α | 0.2 | 0.1 | 0.05 | 0.02 | 0.01 |

Critical value | 1.3749 | 1.9286 | 2.5018 | 3.2892 | 3.9074 |

Reject | No | No | No | No | No |

Hurricane Rita | |||||

Sample Size | 21 | ||||

Statistics | 0.21532 | ||||

Rank | 16 | ||||

Α | 0.2 | 0.1 | 0.05 | 0.02 | 0.01 |

Critical value | 1.3749 | 1.9286 | 2.5018 | 3.2892 | 3.9074 |

Reject | No | No | No | No | No |

Hurricane Felix | |||||

Sample Size | 21 | ||||

Statistics | 0.52823 | ||||

Rank | 22 | ||||

Α | 0.2 | 0.1 | 0.05 | 0.02 | 0.01 |

Critical value | 1.3749 | 1.9286 | 2.5018 | 3.2892 | 3.9074 |

Reject | No | No | No | No | No |

Hurricane Ike | |||||

Sample Size | 21 | ||||

Statistics | 0.5101 | ||||

Rank | 19 | ||||

α | 0.2 | 0.1 | 0.05 | 0.02 | 0.01 |

Critical value | 1.3749 | 1.9286 | 2.5018 | 3.2892 | 3.9074 |

Reject | No | No | No | No | No |

Hurricane Igor | |||||

Sample Size | 21 | ||||

Statistics | 0.18979 | ||||

Rank | 5 | ||||

α | 0.2 | 0.1 | 0.05 | 0.02 | 0.01 |

Critical value | 1.3749 | 1.9286 | 2.5018 | 3.2892 | 3.9074 |

Reject | No | No | No | No | No |

Hurricane Ophelia | |||||

Sample Size | 21 | ||||

Statistics | 0.22014 | ||||

Rank | 6 | ||||

α | 0.2 | 0.1 | 0.05 | 0.02 | 0.01 |

Critical value | 1.3749 | 1.9286 | 2.5018 | 3.2892 | 3.9074 |

Reject | No | No | No | No | No |

Hurricane Sandy | |||||

Sample Size | 21 | ||||

Statistics | 0.34723 | ||||

Rank | 15 | ||||

α | 0.2 | 0.1 | 0.05 | 0.02 | 0.01 |

Critical value | 1.3749 | 1.9286 | 2.5018 | 3.2892 | 3.9074 |

Reject | No | No | No | No | No |

Hurricane | N | Mean | StDev | St Error Mean | 95% CI for the Mean | t-Value | p-Value |
---|---|---|---|---|---|---|---|

Katrina | 21 | 0.002529 | 0.011378 | 0.002483 | (−0.002651; 0.007708) | 1.02 | 0.3207 |

Rita | 21 | 0.019033 | 0.012403 | 0.002707 | (0.013388; 0.024679) | 7.03 | <0.0001 * |

Felix | 21 | −0.027852 | 0.017014 | 0.003713 | (−0.035597; −0.020108) | −7.5 | <0.0001 * |

Ike | 21 | 0.034971 | 0.034986 | 0.007634 | (0.019046; 0.050897) | 4.58 | 0.0002 * |

Igor | 21 | 0.015657 | 0.015056 | 0.003285 | (0.008804; 0.022510) | 4.77 | 0.0001 * |

Ophelia | 21 | −0.02419 | 0.024902 | 0.005434 | (−0.035526; −0.012855) | −4.45 | 0.0002 * |

Sandy | 21 | 0.003562 | 0.018347 | 0.004004 | (−0.004790; 0.011913) | 0.89 | 0.3842 |

Hurricane | N | Median | 95% CI for the Median | Achieved Confidence | Position | p-Value |
---|---|---|---|---|---|---|

Katrina | 21 | 0.0009 | (−0.0027000; 0.0109000) | 92.16% | (7; 15) | 1 |

(−0.0027979; 0.0118795) | 95.00% | Interpolation | ||||

(−0.0030000; 0.0139000) | 97.34% | (6; 16) | ||||

Rita | 21 | 0.0239 | (0.0116000; 0.0262000) | 92.16% | (7; 15) | <0.0001 |

(0.0113388; 0.0262653) | 95.00% | Interpolation | ||||

(0.0108000; 0.0264000) | 97.34% | (6; 16) | ||||

Felix | 21 | −0.0341 | (−0.0400000; −0.0219000) | 92.16% | (7; 15) | <0.0001 |

(−0.0404571; −0.0180474) | 95.00% | Interpolation | ||||

(−0.0414000; −0.0101000) | 97.34% | (6; 16) | ||||

Ike | 21 | 0.0229 | (0.0182000; 0.0278000) | 92.16% | (7; 15) | <0.0001 |

(0.0174491; 0.0288448) | 95.00% | Interpolation | ||||

(0.0159000; 0.0310000) | 97.34% | (6; 16) | ||||

Igor | 21 | 0.0192 | (0.0089000; 0.0225000) | 92.16% | (7; 15) | 0.0072 |

(0.0083450; 0.0231203) | 95.00% | Interpolation | ||||

(0.0072000; 0.0244000) | 97.34% | (6; 16) | ||||

Ophelia | 21 | −0.0163 | (−0.0345000; −0.0111000) | 92.16% | (7; 15) | 0.0072 |

(−0.0346306; −0.0095655) | 95.00% | Interpolation | ||||

(−0.0349000; −0.0064000) | 97.34% | (6; 16) | ||||

Sandy | 21 | 0.0014 | (−0.0041000; 0.0142000) | 92.16% | (7; 15) | 1 |

(−0.0077240; 0.0143632) | 95.00% | Interpolation | ||||

(−0.0152000; 0.0147000) | 97.34% | (6; 16) |

Hurricane | N | Median | 95% CI for the Median | Achieved Confidence | Wilcoxon Statistic | p-Value |
---|---|---|---|---|---|---|

Katrina | 21 | 0.00355 | (−0.00275; 0.00750) | 94.84% | 142 | 0.3662 |

Rita | 21 | 0.01935 | (0.01415; 0.02560) | 94.84% | 227 | 0.0001 |

Felix | 21 | −0.0275 | (−0.03770; −0.02055) | 94.84% | 2.5 | <0.0001 |

Ike | 21 | 0.0249 | (0.0185; 0.0486) | 94.84% | 231 | <0.0001 |

Igor | 21 | 0.017 | (0.00820; 0.02335) | 94.84% | 214 | 0.0007 |

Ophelia | 21 | −0.0232 | (−0.03565; −0.01135) | 94.84% | 16 | 0.0006 |

Sandy | 21 | 0.0032 | (−0.0058; 0.0127) | 94.84% | 132.5 | 0.5663 |

© 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Feria-Domínguez, J.M.; Paneque, P.; Gil-Hurtado, M. Risk Perceptions on Hurricanes: Evidence from the U.S. Stock Market. *Int. J. Environ. Res. Public Health* **2017**, *14*, 600.
https://doi.org/10.3390/ijerph14060600

**AMA Style**

Feria-Domínguez JM, Paneque P, Gil-Hurtado M. Risk Perceptions on Hurricanes: Evidence from the U.S. Stock Market. *International Journal of Environmental Research and Public Health*. 2017; 14(6):600.
https://doi.org/10.3390/ijerph14060600

**Chicago/Turabian Style**

Feria-Domínguez, José Manuel, Pilar Paneque, and María Gil-Hurtado. 2017. "Risk Perceptions on Hurricanes: Evidence from the U.S. Stock Market" *International Journal of Environmental Research and Public Health* 14, no. 6: 600.
https://doi.org/10.3390/ijerph14060600