Economics is a social science. As it is not a positive science such as physics or chemistry, laws of nature can not express it. Yet, it would be misleading to say that human beings are the only subject of Economics. After all, we know that among some animal species, a primitive version of barter economics exists. So, we can conclude that human beings and animal species that tend to be gregarious are the subjects of Economics. Below, I tried to explain all the important facts that an Economics student must consider.
1. Learn English
Crucial. As in every science, Economics is reigned by English as well. You have to improve your English well enough to be able to read academic textbooks and to fully grasp the idea that they want to deliver to you. As most of the academic and corporate content (e.g. research reports released by big investment banks) are in English, you need to have a brilliant command of English (academic and technical terms/theories are included) to be able to conduct qualified research and to develop new theories.
2. Learn Mathematics
Just as in positive sciences, mathematics is again the only tool we have in our hands to understand and objectively comment on economical phenomenons. When I was a high school student in Turkey, very important mathematical concepts such as Limits, Derivatives and Integrals were not taken seriously by our professors. Unfortunately, we have only learned how to apply them to mathematical problems but we had no idea what do they actually mean! Okay, the first derivative of $2x^3$ is $6x^2$. But, what does that operation actually mean? Why do we use it?...
These 3 topics are crucial for setting the base for your mathematical knowledge and providing a solid way to higher-level mathematical concepts, such as differential equations. You can not fully understand and solve differential equations without mastering these 3 crucial concepts.
So what do differential equations mean?
Differential equations help us to understand the interaction between two variables. For example, you can analyze how the change in wind speed affects the movements of a leaf on a tree. Or, to give an example in economics, you can analyze the interest rate - inflation relationship with the help of differential equations. Apart from that, mathematical concepts such as set theory, number theory, or analysis are also very important.
As the importance of mathematics in economics deserves much more attention, I am leaving the rest for another post.
3. Learn Psychology
Expectations rule economies. I deliberately underlined this, because expectations are shaped around crowd psychology. Let me give you an example:
Let's assume that we want to build an econometric model to estimate the price of the BIST 100 index (or any other you want) for next month. This is an econometric problem. As we are concerned with the magnitude of the index price instead of its direction, we must run a regression. Therefore, we have to consider important factors that we think BIST 100 price is affected by. Some of these indicators are going to be fundamental macroeconomic variables, such as exchange rates. Similarly, some indicators we will pick using technical analysis, such as MACD (Moving Average Convergence-Divergence. If it crosses the 70-level upwards, then it signals there is excess demand in the market, and the price is expected to go down. Also, if it crosses 30-level downwards, then there is an excess supply signal in the market, and the price is expected to go up). Keep in mind that technical analysis indicators provide information about the market players' expectations, so it is a subject of crowd psychology.
We collect data from more variables like that to make a prediction. Fair enough. However, the biggest problem is: We have no data in our hands regarding the investment plans/expectations of high-level investors, such as investment banks or billionaires. Therefore, our forecast is never going to be perfectly accurate. Unless we have the option of conducting psychological experiments to have a clue about large investment decisions, we have to analyze the limited psychological information that comes from technical indicators as well as possible and place them in our model.
4. Master Your Skills About Modeling Techniques
One of the easiest ways to deeply analyze the relationship between two (or more) variables in Economics is to create an econometric model of them. Therefore, you must learn as many modeling techniques as possible by heart. Some of the well-known techniques are OLS (Ordinary Least Squares), Logistic Regression, Random Forest, Extreme Gradient Boosting, Artificial Neural Networks, Naive Bayes, Support Vector Machines, etc. Of course, the type of algorithm you should use depends on the nature of the economic/econometric problem. For example:
You work in the loan department of a bank. A mid-size company operating in FMCG (Fast-Moving Consumer Goods) sector applies to your bank for a loan. In this case, a credit-rating model based on Logistic Regression may help you a lot to decide on whether to grant the loan or not. However, as Logistic Regression only works with binary outcomes (0-1, True-False, etc.), it does not tell you how small/big the loan should be if you decide to grant it. To have an idea about the magnitude, you should use a hybrid model by employing e.g. Random Forest along with Logistic Regression. I will write the details of these algorithms as well as how to make an econometric analysis in my upcoming posts.
5. Learn Coding
Let me be clear: You do not have to be an expert in all coding programs such as R, Julia, Python, etc. You even do not have to master any of them. You are going to become an economist, not a software developer. Keep in mind that your coding skills are just tools in your toolbox, tools that are going to be helpful for your future career. Pick the one among them that is the easiest for you to learn. And learn at least basic coding to collect and preprocess data, as well as to write algorithms.
10-20 years ago, programs such as Gretl, Eviews, SPSS were very popular both in the markets and academic world. However; even though the interfaces of these programs are very user-friendly, as time went by they started to be a little bit obsolete especially for running newly developed algorithms such as XGBoost. To run them, you need more sophisticated coding programs such as R.
In such coding programs, there exist specific libraries which makes things a lot easier. Such as caret in R and scikit-learn in Python. Most of the formulas that you need to conduct your analyses exist in these kinds of libraries by default. So all you need to do is mastering as many libraries as possible!
6. Pay Attention to Visuality
This is especially important for all presentations and written material you produce. After all, we are living in an era that (unfortunately) often favors visuality over quality. People would care about how your works look at least as much as how qualified they are. Make sure that you prepare your presentations as clear and simple as possible to deliver your results to the audience. In my opinion, classical Powerpoint presentations start to become outdated. Nowadays, presentation/data analysis programs such as Tableau have come into prominence. If you decide to learn one of these professional presentation programs along with the suitable coding program, it will simply be the strawberry on the cake!
The second part of this post can be found under the title "Advices For Economics' Students - 2". All comments are more than welcome.
No comments:
Post a Comment