Differences

This shows you the differences between two versions of the page.

--- public:on_uea_gig_ticket_price_and_ftse_100 [2019/11/18 02:22] – [Analysis] fangfufu
+++ public:on_uea_gig_ticket_price_and_ftse_100 [2019/11/24 00:26] (current) – removed fangfufu
@@ Line 1: / Line 1: @@
-====== On UEA Gig Ticket Price and FTSE 100 ======
-<note tip> Please don't take any of the findings too seriously! </note>
-===== Abstract =====
-By analysing The Gig List [(gig_list>UEA LCR Gig History 1963-2018 Complete https://issuu.com/ruthselwyn-crome1/docs/uea_lcr_gig_list_for_website)], I discovered that the average ticket price ticket to a UEA SU's gig has been increasing on average 7.57% every year, since sterling decimalisation [(sterling>Decimal Day https://en.wikipedia.org/wiki/Decimal_Day.)] This is higher than the inflation rate in most years. I also compared the price inflation of the gig ticket against FTSE 100. I discovered that between 1985-01-01 and 2017-01-01, FTSE 100 on average has grown by 309%, while average UEA gig ticket price has grown by 936%. This means that if UEA SU was a commercial organisation, it might have been more effective to invest in UEA SU than an index fund linked to FTSE 100.
-===== Background =====
-On Friday evening, I played Scrabble with AV. I threatened to write a program to solve Scrabble, because I was so bad at playing that game. Writing a Scrabble solver is a two part problem - first you need an anagram solver, second you need to search through the board to find the highest scoring solution. But then I discovered Scrabble solver on the Internet, which I gave me a massive boost, and I managed to defeat AV in one game. However, cheating makes the game so boring. Since the problem solving part of my brain was active, I was actively looking for problems to solve.
-After AV has left, I arrived at UEA SU shop to get some snacks. I discovered a book named The Gig List, which contains the list of gigs played in UEA venue between 1966-02-22 to 2018-07-20. It cost £6. I immediate realised the historical significance of this book. It is a good tool for economics research I came up with the idea of investigating how the price of gigs changed over time.
-===== Objective =====
-  - Convert The Gig List from the paper format to Excel spreadsheet
-  - Analyse the spreadsheet.
-  - Disseminate findings from the analysis.
-===== Converting The Gig List from paper format to Excel spreadsheet =====
-==== From paper to image ====
-Initially I planned to sacrifice my copy of the Gig List. I started by cutting off the spine of the book using a pair of scissors, and scan the pages using the departmental scanner, which has a document autofeeder. However, after 20% in, I realised that this task was too painful for my delicate hands. I then discovered that there is a digital version of the book. However the downloading of that document had been disabled. I had to use a third party tool. The downloaded PDF file does not include any of the textual information. Only images were downloaded. This meant that I still had to perform OCR on the resulting image. I suppose the good thing is that the quality of the input image for OCR is guaranteed, the bad thing is that I wasted my £6.
-==== From image to text ====
-To convert the downloaded images to text, I ran the following command:
-<code bash>
-pdfimages Gig_list.pdf -all ./Gig_list
-for i in {000..088}; do tesseract Gig_list-$i.jpg Gig_list-$i -l eng -c preserve_interword_spaces=1 ; done
-cat jpg/*.txt | grep ^[[:digit:]]?*[[:alnum:]]?*  | grep -E 'p$'\|£ > Gig_list_semi_valid.txt
-cat Gig_list_semi_valid.txt|sed -r -e 's/^.{2}/&£/'|sed -r -e 's/^.{7}/&£/'|sed -r -e 's/^.{10}/&£/' > Gig_list_delimited.txt
-</code>
-=== OCR ===
-I tried various OCR engines, including ''Ocrad'', ''Cuneiform'', ''GOCR'' and ''Tesseract''. Tesseract produced the best result, probably because it is made by Google.
-Note that ''preserve_interword_spaces'' was needed to preserve the formatting of the table.
-=== grep and sed ===
-I agree that the way I used grep was particularly ugly. One regular expression search for strings started with number, then alphabets, the next regular expression searches for the letter ''p'' and the  ''£'' sign. It can definitely be constructed more elegantly, for a starter, these two regular expressions can be combined as one. I agree that I don't fully understand how regular expression works with grep. If someone can properly teach me how it works //in person//, that would be much appreciated. I don't use it enough to spend time learning about it. I know one day this will come back and bite me, but whatever.
-And, no I do not understand how to use sed at all. I basically copied and pasted the code from Stackoverflow. I basically decided to use ''£'' sign as the delimiter, and I decided to use sed to insert those delimiters.
-==== From text format to Excel spreadsheet ====
-Now a bit of manual processing is required - invalid lines needed to be removed. There weren't that many of them. Then open up the text file using Excel, set the delimiter to ''£'' sign.
-Excel is pretty much GUI, it is pretty easy to use. So it needs no further explanation. However it should be noted that when plotting the chart, exponential trendline is required, as inflation is an exponential growth.
-==== Calculation in Excel ====
-Initially I wasn't sure which model should be fit onto the ticket price. It certainly showed a curve. I then realised that exponential growth curve should be fitted, because inflation is an exponential growth.
-After obtaining the equation for the best fit line, I thought about converting that into annual inflation rate by changing the base of log. However I realised that would be too complicated. So I opted to just substitute timestamp back into the equation of the best fit line.
-The equation for the best fit line happened to be:
-$$ y= 0.0048e^{0.0002x}, $$
-where $x$ is the number of day has passed since 1900-01-01. (This is the way Excel's calendar system works.)
-It turns out that the average ticket price ticket to a UEA SU's gig has been increasing on average 7.57% every year.
-===== Analysis =====
-It is clear that the inflation on UEA SU's average gig price is higher than the CPI inflation rate for most years [(inflation>Historic inflation Great Britain - CPI inflation
- https://www.inflation.eu/inflation-rates/great-britain/historic-inflation/cpi-inflation-great-britain.aspx)]. This means that UEA SU's gig division has a steady increase in revenue in real terms over the years. There are multiple potential explanation for the high growth rate:
-  - UEA's social standing has increased over the years. Higher profile bands are willing to travel to Norwich to perform.
-  - UEA SU has decided to picked more expensive bands to play at UEA.
-  - The cost of making music has increased over the years.
-Personally I believe the reason behind the price increase is 1 and 2. To ascertain the reason, we can either look at the financial health of the music industry as a whole, or look into UEA SU's financial account over the years.
-However, it should be noted that since 1985, the average ticket price has gone up by 936%, while FTSE 100 has only gone up by 309% [(FTSE100>FTSE Index - FTSE Index Delayed Price. Currency in GBP https://finance.yahoo.com/quote/%5EFTSE%3FP%3DFTSE/history/)]. The average price of a UEA gig grows faster than London stock market! This means that if UEA SU was a for-profit organisation, it might have been a great investment opportunity. It might have been more effective to invest in UEA SU than an index fund linked to FTSE 100.
-However, we all know that profit does not necessarily link to revenue. It would be interesting to see if the profit created during those gigs have grown in at the same rate.
-===== Resources =====
-The Excel spreadsheet in question can be found at {{:public:uea_gig_ticket_price_analysis.xlsx}}.