Event Studies in finance and economics 

Not really. The impact of COVID-19 can easily be seen simply by looking at the chart of market index itself. Typically, event studies a specific event (say a buy-back announcement), which occurs at different dates, to various shares in "event time" and calculates "abnormal returns" vs the index for each share etc. COVID was a market event, impacting all shares (although differently, as Amazon did well (say) vs AirBnB (badly)) and the (event) effect can be seen in the market index itself. Same would apply to the 9/11 event in the USA, which was a market event.

@@MichaelWardFinance Thank you professor, really appreciate your response.

@ Dr Hasan, May I have your details, I’m doing something similar in my dissertation and I’d do with some input from you

Yes, you can. This would be a single event, for one company (so you don't need "event time"), but you would need to estimate the Abnormal Returns (and cumulate these) around the event date.

Hi Matthew, many would argue that event studies are only valid over an event window of a few days (say -10 to +10 around the actual event dates). But, this will depend on how good your forecast methodology is of abnormal returns (if you have a good model to predict abnormal returns, you can lengthen the window - and there are ways of testing this). However, for a long horizon such as you plan, I would suggest Style Analysis (also known as "Buy and Hold" portfolio analysis). In your example, you would compare the performance of just two portfolios (horizontal and vertical acquisitions) over three years. Here's a link to our video on our "Style Engine" which explains more: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-6iTK784XAqE.html

We only use trading days, so an event on a weekend or holiday is the return (not price) on the first trading day that follows. So, eg for a weekend, return=ln (Price Monday / Price Friday). This measures the return from the closing price on Friday to the closing price on Monday and includes any event within the period.

@@MichaelWardFinance thank you so much for your valuable suggestions.

@@MichaelWardFinance so the t0 would be monday and t+1 would be on thursday and so on? (edited: i meant Tuesday)

@@rarararawr9931 t0 is always the day the "news" of the even hits the market. t+1 is the next trading day. So, if for example, the event was on Monday (t0) then the next trading day (t+1) would be Tuesday (unless Tuesday and Wednesday were public holidays?)

@@MichaelWardFinance sorry, yes i meant was Tuesday. So lemme confirm again, if the event day supposed to be on December 31, 2020 but it's on holiday. Is it true that t0 would be on Jan 4, 2021; t+1 Jan 5, 2021 and so on?

Of course COVID affected all listed shares, including banks. The easiest way to see the abnormal returns (which are the essence of event studies) for the banking sector would be to simply calculate the cumulative ARs for the banking sector index versus the market (S&P500) and plot these.

@@MichaelWardFinance how can I do it sir? It’s about my master thesis. Where can I find that banking sector index? So basically the banking sector index will give me a macro view about how lockdown impacted the stock price of banks but then I would like also to separately see the effect of both type of banks the G-sibs and the D-sibs. However, I have 81 D-sibs and 19 G-sibs banks. So how should I do it? And can you maybe tell me where can I download the adjusted close price of the banking sector index and how it’s called?