A Dissertation based on the FX Market

This is a discussion on A Dissertation based on the FX Market within the Economic & Fundamental Analysis forums, part of the Methods category; Hi all, I'm new here, but have traded a little FX myself. I'm currently an undergraduate in UK studying Economics. ...

Reply
 
LinkBack Thread Tools Search this Thread
Old May 14, 2008, 6:31pm   #1
 
4 Posts
Joined May 2008
A Dissertation based on the FX Market

Hi all,

I'm new here, but have traded a little FX myself. I'm currently an undergraduate in UK studying Economics. The time has come for me to select a dissertation title and do a bit of research.

As is always the case with dissertations, one requires inspiration, and I'm finding it hard to come by!

My main interest is the FX Market. Since I'm doing an Economics degree, it must be linked; I want to do a dissertation in this field - it will be an empirical (rather than theoretical) approach, and I would like to test for something. The obvious one would probably be whether some given currency pair follows a random walk. However, this seems very bread-and-butter and am looking for something slightly varied.

I thought this would be the best place to ask for ideas and suggestions. Does anyone have any broad or narrow areas they can point me to; something which might make an interesting topic for me to a) research, b) write a dissertation on and c) isn't an untouched area. Point (c) is important because this is an undergraduate dissertation and my knowledge, time and the importance of this dissertation are all limited (but fairly significant).

Thank you for your time and ideas.
N9ne is offline   Reply With Quote
Old May 14, 2008, 7:52pm   #2
Joined Dec 2004
Take a look at Benoit Mandelbrot's work. His book The (Mis)Behavior of Markets might give you some really interesting ideas for research paths. Be warned, though. He's not a proponent of classic theory where it relates to the markets and provides plenty of solid arguments against the likes of random walk, efficient markets, etc.
__________________
John Forman, PhD
Author - Trading FAQs, The Essentials of Trading
Rhody Trader is offline Training literature vendor (e.g courses / spreadbetting guides / eBooks)   Reply With Quote
Old May 14, 2008, 8:08pm   #3
Joined Mar 2007
Interesting topics you can find in the FX market.

A rather simple one would be the initial market reaction to news announcements in the economic calendar.
You would find some interesting stuff that is actually counterintuitve to the economic theory you were taught

It is simple, and easy to use correlations to prove the point. and you dont need much historical data....there are so many news announcements that you would not need more than 6 months of news announcements.....metatrader will give you enough.
kenobi is offline   Reply With Quote
Thanks! The following members like this post: Priceman , firewalker99
Old May 15, 2008, 12:06am   #4
Joined Jun 2006
Kenobi,

Good idea. If i may expand, I would suggest choosing and comparing the most watched data, eg CPI, Input/output prices for two countries - US/Eurozone for US/EUR - and examine how the currencies react to their own, and each others, indicators over the course of a year or two.

Grant.
grantx is offline   Reply With Quote
Old May 15, 2008, 12:29am   #5
 
nine's Avatar
Joined Sep 2003
Adding to the news idea: does the market regularly move in the direction of the news before the news (or in fact do the opposite)? What does this imply?
nine is offline   Reply With Quote
Thanks! The following members like this post: Mathemagician
Old May 16, 2008, 12:05am   #6
 
4 Posts
Joined May 2008
N9ne started this thread Thank you all for your replies. They've all been very useful and insightful!

To expand on this, I am thinking perhaps of something rather more applied to economics. I fear my dissertation title may be rejected if it lacks that economics element.

There's a broad area I'm interested in, not necessarily restricted to the FX Market. Perhaps an interesting dissertation would be to explore the relationship between oil, equities and FX. This would be tremendously interesting and would offer me the opportunity to research into the history and the mechanics of not one but three different markets.

However, what would my dissertation specifically be on? I would not be allowed to write a dissertation on merely the 'interactions' between those three very distinct markets. Is there something particular here, which I could explore in greater detail?
N9ne is offline   Reply With Quote
Old May 16, 2008, 12:44am   #7
Joined Mar 2007
Quote:
Originally Posted by N9ne View Post
Thank you all for your replies. They've all been very useful and insightful!

To expand on this, I am thinking perhaps of something rather more applied to economics. I fear my dissertation title may be rejected if it lacks that economics element.

There's a broad area I'm interested in, not necessarily restricted to the FX Market. Perhaps an interesting dissertation would be to explore the relationship between oil, equities and FX. This would be tremendously interesting and would offer me the opportunity to research into the history and the mechanics of not one but three different markets.

However, what would my dissertation specifically be on? I would not be allowed to write a dissertation on merely the 'interactions' between those three very distinct markets. Is there something particular here, which I could explore in greater detail?
ok mate, you come to a BB to get a topic for yer dissertation. cool. think ye already got a topic. ye just need to read again. wot has been posted by several of us is pretty economical to us.

right, probably it wasnt explained.

in yer economics course you were told....

a) if BOE raises rates....then X happens to the £
b) if Inflation numbers are up........then X happens to the £ ......

and so on.

i dont really know how good yer Uni is.....but will give you a hint.....use rational expectations and market efficiency theory........

a) test against the first 15 minutes following a news release
b) test against the the 240 minutes following a news release
c) compare to economic theory and what the new info brings into the valuation

you will have so many topics to isolate

well, cant get more "economics" than that.

good luck.
kenobi is offline   Reply With Quote
Old May 16, 2008, 11:41am   #8
Joined Jun 2006
What about, What will President Bush do if the Saudi's don't play ball re increasing oil production (hint: he'll form an alliance with Venezuela and Iran).

Grant.
grantx is offline   Reply With Quote
Old May 19, 2008, 12:31am   #9
Joined Mar 2008
I just finished my undergraduate economics dissertation, and yes, it kicked ass. It is so important that you pick a good topic. If you pick a good topic you walk straight into a first as long as you can explain what you have found.

My advice to you is to think about data availability right from the start. The people from my class who have written good pieces are the ones who had access to a really good data-set. Pick a topic that uses data you can get easily.

FX data is obviously quite easy to get from a variety of place but it will be quite hard, for example, to get historical data that will tell you how FX prices react to news announcements on a time frame of minutes (which is the only interesting result really). However, if you still have a long time to go before you have to write it, you will be much better off collecting the data in real-time, i.e sitting in front of a chart when a news story breaks and measuring the reaction.

If you do go with the FX market reaction to news topic, it is interesting but econometrically it could be a bit troublesome. Anyway, it would be interesting to test to what extent the market predicts the outcome of the announcement etc. You would certainly need to get a lot of observations in order to come to a firm conclusion (we are talking hundreds of announcements) and this would be very time consuming to do. All depends how much time you want to put into it.

My final piece of advice is not to do a topic JUST because you are interested in it. Think about the conclusion that you are going to be able to come to before you start. The dissertation will only be as interesting as the conclusion. Some people in my class pursued topics just because they were interested in it and didnt consider the data and conclusion and have written **** dissertations.

I, for example, have no interest in equities. Yet I wrote a dissertation using data from the FTSE 100 (easy to get) that was awesome, because I knew what I was going to find before I started.

Hope this helps

C
CHRISTO9HER is offline   Reply With Quote
Old May 22, 2008, 10:38pm   #10
Joined Nov 2007
Christoph makes a good point;

If you want to study the reactions of currencies to economic announcements, it might be tricky to draw any solid conclusions from it; you could compare the "textbook" result (e.g. High inflation => weaker currency) to the actual market reaction, but a driving factor of this will be how firms (with size to move the market) are positioned for the release and their subsequent effects on price (e.g. short covering) - something you have no hope of being able to find out. Combine that with the possibility that the #(expected reactions) = #(unexpected reactions) and it might be difficult to draw conclusions with any level of confidence.

Maybe that's a dissertation right there; an investigation that shows there are statistically significant drivers of price (read speculation) other than macro (read inflation, the only one I can think would be easy to quantify).
MrGecko is offline   Reply With Quote
Old May 22, 2008, 11:13pm   #11
Joined Jun 2006
Chris,

What was the subject of your dissertation?

Hello Mr Gecko, how's it going, mate?

Grant.
grantx is offline   Reply With Quote
Old May 23, 2008, 7:09pm   #12
Joined Mar 2008
Quote:
Originally Posted by grantx View Post
Chris,

What was the subject of your dissertation?

Hello Mr Gecko, how's it going, mate?

Grant.
Hi

I'll describe what I wrote about and what I found and sorry if I go into more detail than people find interesting!!

My dissertation was to do with the Capital Asset Pricing Model, one of the most celebrated financial developments of the last few decades and still one of the most useful models to find the expected return on a share.

It basically says that:

Return on share = risk free rate + Beta*(return on market - risk free rate)

Where the beta of a share is a measure of the correlation between the returns to the share and the returns to the market. Essentially the model says that the only type of financial risk that should attract a return is market risk.

Anyway, Beta is found by regressing the returns to the share on the returns to the market for some past period. In the model, "the market" should actually be "all wealth" (i.e it should be equities, bonds, houses, etc etc) but in reality people use a market index as an approximation.

I showed that if you regress the returns to the share on a market index (ie the FTSE 100) and the share is also a component of that index, then the beta value that you get is a biased estimate.

To understand why, consider that if share that is a large component of the index e.g. Shell or BP or HSBC has a great day and makes a 10% gain, the FTSE will also make a gain attributable to that. This effectively increases the correlation between the share and the market and therefore boosts the beta value to an artificially high level.

I found that this bias can be up to 20% of the beta value for the most heavily weighted shares in the index!! This is extremely relevant as the beta value is considered pretty critical by many analysts.

I then went on to find that the size of the bias is affected by market conditions. When the credit crunch hit we saw a MASSIVE shift in the correlation between shares (just think about those days when every bank lost 10% because one of them uncovered losses). The bias has been strongly affected by this.

Anyway, I have attached the final version for anyone who is interested. If anyone actually reads it then let me know what you think.

And a note to N9ne, (the starter of this thread), dont copy it. I am not saying this because I care if you copy it (I actually dont care at all) but all universities are now co-operating to get undergraduate dissertations onto a single database to detect cheating.
But hopefully it might get you thinking about how you might be able to combine your enthusiasm for financial markets with a topic with good economic/econometric content.
Let us know what you decide to research!!

Anyway, cheers guys

C
Attached Files
File Type: pdf EC331-a1.pdf (536.9 KB, 2284 views)
CHRISTO9HER is offline   Reply With Quote
Thanks! The following members like this post: MrGecko , firewalker99 , grantx , GammaJammer
Old May 23, 2008, 7:17pm   #13
Joined Nov 2007
Much kudos to you Christoph for posting your dissertation!

I went to warwick (good few years ago now); studied Physics. Did you play Rugby?
MrGecko is offline   Reply With Quote
Old May 23, 2008, 7:38pm   #14
Joined Mar 2008
Quote:
Originally Posted by MrGecko View Post
Much kudos to you Christoph for posting your dissertation!

I went to warwick (good few years ago now); studied Physics. Did you play Rugby?
Good to see a fellow Warwick Alum on here!! (Or at least I will be an alumni in about two weeks)

As for the rugby question, I played in the first year but ended up getting too injured. But the playing was always secondary to the drinking anyway and I have certainly kept that up!! You really cant beat the rugby circles on sports night!
CHRISTO9HER is offline   Reply With Quote
Old May 23, 2008, 7:42pm   #15
Joined Nov 2007
Quote:
Originally Posted by CHRISTO9HER View Post
As for the rugby question, I played in the first year but ended up getting too injured.
pussy

Quote:
But the playing was always secondary to the drinking anyway and I have certainly kept that up!! You really cant beat the rugby circles on sports night!
WHACK OFF!!!!

I have just sung "OH, THE, CAPTAIN of the 1sts he is a fool! He's only got a teeny weeny tool....." at my desk. seriously.
MrGecko is offline   Reply With Quote
Thanks! The following members like this post: grantx
Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search

Similar Threads
Thread Thread Starter Forum Replies Last Post
Market related dissertation ideas? sak07 The Foyer 8 Nov 14, 2008 10:42pm
Market Depth based Scalper francoisLdn Discretionary Trading 26 Oct 28, 2008 9:55pm
stops based on "market" - Finspreads kool4caats Spread Betting & CFDs 11 May 30, 2007 10:35am
Gambling game based on the stock market in2uxs General Trading Chat 1 Nov 9, 2006 12:45pm
dissertation survey...need your opinions! sak07 The Foyer 4 Jan 3, 2005 3:11pm

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)