44
45 banks should be under particular pressure to adopt best-practice governance schemes which makes them interesting to consider in the light of institutional influences.
Now, using selection criteria developed in 2011 for the same panel going back in time calls for caution. First, we note that two of the banks (Banque Populaire CdE and Bank of New York Mellon) result from mergers that took place in the time period between 2003 and 2011, Banque Populaire did not merge with CdE until 2009 which means that only one observation is available for the bank as it stands today. Therefore, we exclude it from our sample. However, Bank of New York Mellon resulted from a merger between Bank of New York and Mellon Financial Corporation which was announced in 2006. Thereby, we have observations for the two most recent time periods and choose to keep the bank within the observation group to favour the total size of our sample. There is no reason to expect a radically different trajectory when it comes to corporate governance practices for Bank of New York Mellon than for the remaining sample group, which means that we can include it in the 2007 and 2011 observations without risk of biasing our results. Further, using 2011 criteria for selecting a sample group that we use also in 2003 and 2007 may bias the sample by not considering the systemic importance of each constituent at each point in time. However, by redefining the sample in each time period, our results would need to be adjusted for the difference e.g. geographical composition which would be difficult and potentially distort our findings. Also, since the 2007-2011 time period is of central interest to the paper, the systemically importance of the banks in 2011 most accurately reflects what we seek in our sample group: banks which are potentially subject to institutional influences in defining their governance regimes.
In brief, our final observation group include following banks (domicile) ; Bank of America (US), Bank of New York Mellon (US), Barclays (UK), BNP Paribas (FR), Citigroup (US), Commerzbank (DE), Crédit Agricole (FR), Credit Suisse (CH), Deutsche Bank (DE), Dexia (BE), Goldman Sachs (US), HSBC (UK), ING Bank (NL), JP Morgan Chase (US), Lloyds TSB (UK), Morgan Stanley (US), Nordea (SE), Royal Banks of Scotland (UK), Santander (ES), Société Générale (FR), State Street (US), UBS (CH), Unicredit (IT) and Wells Fargo (US).
4.3 Collection of data
In order to produce as reliable data as possible, we have manually extracted all information from company filings. What the primary source of data has been depends on the domicile of the bank. For example, given the strict and standardised filing requirement in the US, we have extracted reports such as 10-K and 14-A from the centralised SEC database for the most parts when gathering information on American banks. However, data on the European sample group comes predominantly from annual reports and other filings made available through the corporate webpage, such as compensation policy documents and general principles of corporate governance. In some instances, when we have not been able to find the information we seek in any of the methods outlined above, we have considered
46 registration documents and general meeting protocols for additional disclosure. Also, for capital markets related information on dual share classes, share price developments and leverage, and where company filings contained insufficient information on shareholder bases, we have used S&P Capital IQ and Bloomberg to obtain a fuller dataset. A full list of the documents consulted is available in 10.1 Appendix 1.
When we expand our analysis by introducing cases, several additional data sources are considered. In particular, news article are used together with public company communication releases to map the events surrounding a particular topic. Also, official documents are the main source for information relating to regulatory developments and the publication of principles and recommendations in the realm of corporate governance.
4.4 Description of observation group
Based on our definition of observation group, the sample includes eight US based banks, four UK based banks, two banks each from France, Germany and Switzerland, and four banks domiciled in Belgium, Netherlands, Sweden and Italy. Thereby, our observation group has a slight overweight of European over US based banks.
Beyond national differences, the banks included in the observation group have been affected differently by the financial crisis due to their individual business mixes and exposures to certain asset classes. Since financial results and stock market performance have direct reputational effects which in turn may change corporate governance practices, we consider individual performance data for each bank. We use three key metrics to analyse individual bank performance and the aggregate of the full observation group: (1) y-o-y stock price changes, (2) price/book value and (3) leverage. All data is collected from S&P Capital IQ17. Overall, we observe that banks’ stock prices (rebased at 100 at 2002-12-31) were severely affected by the financial crisis, on average declining from an average of 183,9 at year-end 2007 to 71,6 by year-end 2008, afterwards hovering around 65,8 to 99,9 until the end of 2011. Further, leverage, which proxies for risk taking, fell from an average of 22,6 by year-end 2007 to 16,9 in late 2011, indicating a reduction of the total risk exposure. Along with the falling share prices, the price-to-book value receded from a range of 2,4-2,6 in 2003-2006 to a 0,7-1,1 interval following the crisis18.
17 Annual observations for each bank, along with the total sample average, are presented in 10.2 Appendix 2.
18 Examining the banks individually, it appears, based on subjective evaluation of both stock price, leverage and price to book value, that Dexia, Royal Bank of Scotland, Lloyds Banking, Société Générale, UBS, Credit Suisse, Morgan Stanley, Goldman Sachs and Crédit Agricole were worst hit y the crisis. The banks that appear to have been more resilient to the financial crisis appear to be Nordea, JP Morgan, HSBC, Wells Fargo, State Street and Santander.
47