Your missing data

You'll remember that last time when discussing where to find your economic data, I mentioned that it was unlikely that you would find the interest rate data on the SARB online inquiry page. I also said that I would tell you where to find it, eventually that is.

Well its been almost a week since that post so here is a link to that missing data. Have fun

Your Economic Data

Data, the raw building blocks of your empirical investigation.

This entry will briefly discuss how your research question helps in the data collection process. Please note that here I am referring to secondary data, that is data that has been collected by a third party. I’ll write an article at some point discussing the different types of data available for econometric research.

As I discussed previously, your research question is of dire importance. Not only does your research question keep you focused, it also helps direct your data search. Lets revisit my thesis question and you’ll that even with minimal economic theory you’ll be able to identify what data you need for this inquiry.

So here’s that research question again:

Aim: 

To assess the impact of an exchange rate movement on a set of South African macroeconomic variables.

Specific objectives:

• To determine the effect of an exchange rate movement on the development of the South African Gross Domestic Product.

• To determine the effect of an exchange rate movement on South African consumer prices.

• To determine the effect of an exchange rate movement on the interest rate setting behavior of the SARB.

You should be able to identify data specific key words above, specifically; South Africa, exchange rate. Gross Domestic Product, consumer prices and interest rates.

From this we can see that in this instance the data that we are looking for South Africa, specifically GDP (income), exchange rate, price level (CPI) and interest rate data. The next question we need to ask ourselves is were is the best place to get your data. Fortunately all countries collect data on the performance of their economy, however the quality of that data is sometimes questionably.

For South African economic data the first places to look are the South African Reserve Bank (SARB) and Statistics South Africa (Statsa) websites. You can also obtain data from international sources like the World Bank, though be care and not to use to many sources as different agencies use different definitions and calculations, best be consistent.

The best place to look should be the SARB’s website as it manages South Africa’s monetary policy and all the variables mentioned in the research question also feature in the Taylor rule of an open economy. A Taylor rule is simply an expression that predicts the behavior of a central bank as function of a concise set of variables' deviations from their trend levels.

The next step is pretty simple, go to the website and look at the on-line statistical query page. Here you’ll be able to search their website for the economic data that you are interested in.

For my research I used the nominal effective exchange rate or NEER, real GDP in 2005 constant prices, CPI and the interbank rate. You could also use the REPO rate which is the SARB’s official policy tool (I’ll explain why I didn’t next time). I also used a couple other variables in the final version of my model but we’ll get there eventually.

You find that the interbank and REPO rates don’t appear when you search for them. This was a source of much frustration for me too, I promise that they are on the website however and I promise I’ll publish a link next time but for now consider this homework. I will also eventually provide the raw dataset once its time for us to start playing. You’ll also find that the best place to find CPI data is actually Statistics SA.

So have fun browsing the site for data, look at the various options available to you such the data frequency and time periods available. Can you see any issues on the horizon?

Until next time

The Research Question!

When embarking on the roller-coaster ride of emotions that is your thesis, it is of fundamental importance to have a clear idea of what you are attempting to accomplish. "But that's blindingly obvious!", I hear you cry. Yes it is my friends, yet the number of tears that have be shed by students who have lost their way because of this lack of focus is innumerable.

A suitable research question normally requires hours or perhaps weeks of careful consideration to craft. Many of us start off with a research idea that is far too grand. Grand is good, but few of us have the time or finances to stay at university until our magnum opus is complete. Rather, think of your initial research idea as a chunk of wood, which then then needs to be carved  and sanded down such that only the most essential and beautiful characteristics remain.

So how do you craft a finely honed research question? Well the steps are pretty generic:

Choose a topic that you are interested in

This is very important. Choosing your research focus is like choosing a roommate or girlfriend. Choose correctly and working on your thesis won't always feel like work. Choose incorrectly however and your relationship with your thesis will be unpleasant.

I have always enjoyed macroeconomic theory, so naturally this is where I sought my thesis topic. Unfortunately I struggled to find a topic that had not already be done to death.

Do your background research

Naturally if you have an interest in a general area I suspect that you'll already be casually reading the work coming out from it. If you are not however,  you really must start, else you'll have no idea of the wealth of research opportunities that are waiting for you. Great sources here include journal articles, working papers from major universities as well as the research published by national and international organizations, such as the World Bank and International Monetary Fund (IMF).

Eventually I read a policy document that outlined the New Growth Path (NGP) for South Africa. This document identified various strategies intended to stimulate the South African economy. I latched onto an assertion mentioned in that document that the South African Reserve Bank (SARB) should seek to depreciate the South African Rand to stimulate economic growth and create more jobs. The NGP however went on to add that the SARB be should proceed to depreciate the Rand whilst maintaining low inflation and low interest rates. More on that just a little later.

Start asking questions

Well, I don't like that. Rather, always question, accept nothing that is said as truth. Only once it has passed all your logical tests, allow it to sit as only a working theory, and then when new evidence is available, re-evaluate.

Whilst reading the NGP policy document I found the assertion that the SARB should depreciate the Rand to stimulate growth whilst simultaneously maintaining the interest and inflation rates low at odds with the behavior of the SARB.  I believed this, as to the best of my knowledge the SARB has behaved as an inflation targeting central bank since 2000. Addtionally, whilst a depreciation of the exchange rate may stimulate growth via a positive effect on exports it would also impose inflationary pressures through something known as the exchange rate pass through (ERPT) effect. Finally, the SARB, practicing inflation targeting would then need to increase interest rates in response to the ERPT effect.

To focus you research right down into a neat concise package, to which you can refer back to for guidance, I would suggest distilling your idea down to a single aim and perhaps 3 to 5 specific objectives.  My thesis research question was thus something like.

Aim: 

To assess the impact of an exchange rate movement on a set of South African macroeconomic variables.

Specific objectives:

• To determine the effect of an exchange rate movement on the development of the South African Gross Domestic Product.
•  To determine the effect of an exchange rate movement on South African consumer prices.
• To determine the effect of an exchange rate movement on the interest rate setting behavior of the SARB.

See, nice and neat. Almost sounds easy. In theory it is and this blog is dedicated to taking you through all the steps required to complete a piece of empirical research using Stata.

I hope that you will join me next time as I discuss gathering data, inputing that data into Stata and how the research question focuses you into what data you are gathering.

The Way Forward

Hi everyone,

This first entry is simply to to inform you of the way forward that I have in mind. The series will commence with what is essentially a walk-through of the empirical work that became my Masters thesis. My masters thesis looked at how a movement of the exchange rate would likely impact upon South African real Gross Domestic Product (GDP), inflation as measured by the consumer price index (CPI) and the interest rate setting behaviour of the South African Reserve Bank (SARB). 

To undertake this analysis a vector autoregressive (VAR) model was estimated and structural analysis techniques were used to investigate the internal dynamics of the model. The structural analysis techniques that were used, and will be discussed, include impulse response functions (IRF), forecast error variance decompositions (FEVD) and dynamic multiplier functions.

This series of blog entries will take you through every step of the process behind the empirical estimation of the model. Everything from data gathering and graphing (I'll use Microsoft excel and Stata) to the complex issues of unit root testing (using various tests), high-pass filters, VARX estimation and impulse response functions.

I hope the topics mentioned above will be of interest to some of you. In the coming weeks I hope to also introduce my other research areas which include trade and energy. I've done some work using panel data models as well and error correction models and bounds testing and am eager to share what I can with everyone. 

Until next time