I have received many questions about quantitative analysis/quant trading. This post is to address these FAQ I receive and point you in the right direction if this is something you are interested in!
If I missed any questions, please leave them in the comments and I will add an addendum!
Q: What is quantitative analysis/quant data?
A : Quantitative analysis is the practice of applying mathematics and statistics to stock trading data. It involves the process of data mining and drawing statistic inferences between related and unrelated variables to look for correlations in data that can be used to predict future stock movement.
Q: What is a “quant”?
A: There are two types of quants or quant traders. This is more applicable to hedge funds and banks who employ these people, but essentially, there are quantitative modellers and quant developers/programmers.
Quantitative modellers (which is essentially, what I am) are generally statisticians who have a degree in applied mathematics or statistics. They employ statistical theories to develop working mathematical models of stocks and attempt to quantify stock behaviour into mathematical formulas and determine probability of meeting certain conditions (i.e. price).
Quant developers/programmers generally have degrees in computer science or computer engineering and software development. They take these models from the statisticians and program them into software to create high frequency trading algorithms and longer-term trading algorithms. They will also use this data to develop software to manage and view risk quantitatively.
Q: Is quantitative analysis the same as technical analysis?
A: No. Technical analysts apply a type of qualitative data analysis. While technical analysis attempts to, loosely, base itself on mathematical principles, it is an attempt to qualitatively represent quantitative data. As such, technical analysis is slightly more susceptible to biases. Whereas one TA may view a Fibonacci level as indicating bullish movement, another may view the same level as indicating bearish movement. It is dependent on the TA’s own sentiment and their ability to recognize sentiment and context.
Contrast this to a QA, the range that one QA comes up with will likely be very similar to the range of the other QA. That is because QAs all apply the same statistical strategies and tests to identify the data and trends. Biases for QAs are generally counter-intuitive to the process. QAs should not care about what the context or sentiment is, they simply follow the algorithmic processes which are characterized as “If – Then” statements.
Q: How does “quant” trading work?
A: Traditional quant trading and the quant trading done by hedge funds and banks are accomplished through computers that execute algorithms directly with exchanges. They do not operate through brokerages, they have a direct link to the exchange where they can quickly enter and exit trades that have satisfied the algorithmic conditions.
For retail quants like myself, it varies. As I am a quant modeller and not a programmer, I must execute my own trades based on the conditions being met. This introduces the possibility of bias on my part and this bias has gotten me into trouble before!
However, other quant traders that are more on the computer programming side, develop their own trading algorithms that will automatically execute their trades, etc. To do this, you need a broker that allows third party integration, in order to integrate your trading platform directly with your developed software. I have no idea how to do this, but I know there are brokers out there that allow this to happen and I know quant retail traders who do, do this.
Q: What do I need to be a quant trader?
A: Generally, you need a solid understanding of statistics and/or computer programming. In order to effectively develop a working model of a stock, you really need to have a strong understanding of statistics; however, I do know some quants that apply machine learning to their modelling which works okay from what they tell me and can avoid the hassle of developing complex mathematical models of stocks (which takes a long time, I speak from experience!).
You also need software and to have a working understanding of a programming language (knowing Excel as a programming language is sufficient!). You need either some form of statistical analysis software or programming software. Software that I frequently see advertised being used at quant firms and banks (at least in Canada) include MATLAB, C++ and Python.
I personally use SPSS (in lieu of SAS and MATLAB) and Excel (in lieu of Python/C++).
Python is much more powerful generally than Excel and even MATLAB, equally as powerful as SPSS and SAS in its ability to analyze statistical problems and has the ability to actually do more critical appraisals of information than SPSS, SAS, MATLAB and Excel can do. However, for mathematical modelling, I tend to prefer SAS or SPSS combined with Excel but this is mostly because I am a statistician and this software presents the results in a way that I am familiar with (I’m an old dog with no interest in new tricks). A software engineer or programmer would most likely prefer Python. Specifically Anaconda has the same functionality as MATLAB (or so I am told).
Q: Is there a cost to the software?
A: So, Python is free! So if you know how to use Python or you are interested in learning, you can download it free online! It is open source and very powerful! If your novice, I recommend downloading Python Anaconda, it has everything you need!
Excel and SPSS (what I use) tend to be costly. Excel is the cheaper alternative, I think it costs me about 75$ a year (however, I am still a student so I get the student discount, not sure full price).
SPSS, MATLAB, SAS are extremely expensive. In excess of over 2,000 USD. There is an option to do an annual licensing agreement for less, but the price would add up.
Q: Do I need a degree in mathematics or computer programming?
A: NO! You don’t. You can learn this stuff from books and reading. Having a degree doesn’t even guarantee you that you will understand this stuff. I speak to some of my classmates about what I do, and they still don’t understand what I am doing (despite also having MScs in statistics hahaha). It all comes down to your critical application of knowledge! Education is very important, IMO, but its not everything and everyone has the potential to learn if they are truly motivated to!
My background was I started as a nurse with a bachelors of science. I fell in love with mathmatics and statistics in my undergrad and ended up pursuing higher education in mathematics, specifically applied statistics.
Q: How is quant trading different then technical trading?
A: So, as I wrote above, technical trading is the qualitative appraisal of quantitative data. I am not a technical trader and can’t speak too indepth about this process.
But I can contrast a little bit, which I will do below!
A technical trader may look to see that a particular price point was respected and not surpassed over a number of days. They would likely label this as strong support and would assume that a break of this support would lead to more sustained selling.
Contrast to a quant trader, I do not pay attention to any one specific price point. Price action tends to be more on the random side. So I rely on all of the data over many years of trading to develop working ranges and variances between the data. From this, I can determine the range that a stock likes to operate in (whether it be +/- 10 points or, if its TSLA, +/- 30 to 60 points). From here, I can use previous day data to predict a likely range for the next day. When I have that range, I can then express my hypothesis in conditional algebraic forms, like:
IF Condition 1 met THEN statement 1 correct AND statement 2 incorrect; or
IF Condition 2 met then statement 2 correct and statement 1 incorrect.
I then follow linear algebraic principles to identify those conditions and subsequents.
For example, for today, SPY opened around 420.28. The range that I calculated for SPY today was 415 to 427. So, the problem that I needed to solve mathematically was:
If Condition X met then SPY = 427; OR
If Condition Y met then SPY = 415.
Then I must use algebra and statistics to determine what Condition X and what Condition Y are.
If you read my ideas, you will notice that I express my ideas in linear algebraic form. For my post about SPY today, this is what I had wrote:
A break above 424 would indicate bullish sentiment and likely continuation towards 427.
A break below 418 would indicate bearish sentiment and likely continuation towards 415.11.
If you notice, this can be expressed as a conditional (algebraic) statement:
IF X > 424 THEN 427 is met; OR
IF X < 418 THEN 415 is met.
Now I don’t manually do this because it would be to labour intensive. Which is why I say you need to know a programming language. You can program Python, Excel, MATLAB, C++, etc. to do this for you and identify those ranges. But you need to have the theory in order to understand how to get there and how to give Excel, Python, C++ or MATLAB what it needs to solve the problem for you.
Q: Can you recommend books or videos on quant trading?
A: So, I have not found any quantitative retail traders on youtube. There are 2 quant developers that actively post on YouTube who have okay content, one being Trading Jesus and the other is Korean Yuppie (who is still kind of novice and hasn’t posted much). Both are from the perspective of quantitative developers; however, this is a completely different skillset from a quantitative modeller. But equally interesting and informative!
In terms of books, I would recommend general statistics books and books on programming language like Python or even books on Excel. Excel is generally an under-rated platform that is capable of quite advanced data analysis. Don’t under-estimate it! Excel is involved in my trade planning, execution and profit taking process. It is the thing that dictates what I should do and where I should enter/exit.
You also need a solid understanding of the market, how it is organized and how it functions. So general books about market theory and trading are also useful. I have no
real recommendations as I haven’t read any books, aside from The Trading Zone, which I found insightful but not helpful. Most of the information you need is available for free online. I wouldn’t invest a huge amount of money in books that are mostly fluff, especially books on day trading.
Equally, avoid courses! Don’t buy people’s courses and don’t trust trading “gurus” from YouTube.
Hope this answers all of your questions, again please let me know if you have any that I have not addressed!
Thank you, take care and as always, trade safe!
If I missed any questions, please leave them in the comments and I will add an addendum!
Q: What is quantitative analysis/quant data?
A : Quantitative analysis is the practice of applying mathematics and statistics to stock trading data. It involves the process of data mining and drawing statistic inferences between related and unrelated variables to look for correlations in data that can be used to predict future stock movement.
Q: What is a “quant”?
A: There are two types of quants or quant traders. This is more applicable to hedge funds and banks who employ these people, but essentially, there are quantitative modellers and quant developers/programmers.
Quantitative modellers (which is essentially, what I am) are generally statisticians who have a degree in applied mathematics or statistics. They employ statistical theories to develop working mathematical models of stocks and attempt to quantify stock behaviour into mathematical formulas and determine probability of meeting certain conditions (i.e. price).
Quant developers/programmers generally have degrees in computer science or computer engineering and software development. They take these models from the statisticians and program them into software to create high frequency trading algorithms and longer-term trading algorithms. They will also use this data to develop software to manage and view risk quantitatively.
Q: Is quantitative analysis the same as technical analysis?
A: No. Technical analysts apply a type of qualitative data analysis. While technical analysis attempts to, loosely, base itself on mathematical principles, it is an attempt to qualitatively represent quantitative data. As such, technical analysis is slightly more susceptible to biases. Whereas one TA may view a Fibonacci level as indicating bullish movement, another may view the same level as indicating bearish movement. It is dependent on the TA’s own sentiment and their ability to recognize sentiment and context.
Contrast this to a QA, the range that one QA comes up with will likely be very similar to the range of the other QA. That is because QAs all apply the same statistical strategies and tests to identify the data and trends. Biases for QAs are generally counter-intuitive to the process. QAs should not care about what the context or sentiment is, they simply follow the algorithmic processes which are characterized as “If – Then” statements.
Q: How does “quant” trading work?
A: Traditional quant trading and the quant trading done by hedge funds and banks are accomplished through computers that execute algorithms directly with exchanges. They do not operate through brokerages, they have a direct link to the exchange where they can quickly enter and exit trades that have satisfied the algorithmic conditions.
For retail quants like myself, it varies. As I am a quant modeller and not a programmer, I must execute my own trades based on the conditions being met. This introduces the possibility of bias on my part and this bias has gotten me into trouble before!
However, other quant traders that are more on the computer programming side, develop their own trading algorithms that will automatically execute their trades, etc. To do this, you need a broker that allows third party integration, in order to integrate your trading platform directly with your developed software. I have no idea how to do this, but I know there are brokers out there that allow this to happen and I know quant retail traders who do, do this.
Q: What do I need to be a quant trader?
A: Generally, you need a solid understanding of statistics and/or computer programming. In order to effectively develop a working model of a stock, you really need to have a strong understanding of statistics; however, I do know some quants that apply machine learning to their modelling which works okay from what they tell me and can avoid the hassle of developing complex mathematical models of stocks (which takes a long time, I speak from experience!).
You also need software and to have a working understanding of a programming language (knowing Excel as a programming language is sufficient!). You need either some form of statistical analysis software or programming software. Software that I frequently see advertised being used at quant firms and banks (at least in Canada) include MATLAB, C++ and Python.
I personally use SPSS (in lieu of SAS and MATLAB) and Excel (in lieu of Python/C++).
Python is much more powerful generally than Excel and even MATLAB, equally as powerful as SPSS and SAS in its ability to analyze statistical problems and has the ability to actually do more critical appraisals of information than SPSS, SAS, MATLAB and Excel can do. However, for mathematical modelling, I tend to prefer SAS or SPSS combined with Excel but this is mostly because I am a statistician and this software presents the results in a way that I am familiar with (I’m an old dog with no interest in new tricks). A software engineer or programmer would most likely prefer Python. Specifically Anaconda has the same functionality as MATLAB (or so I am told).
Q: Is there a cost to the software?
A: So, Python is free! So if you know how to use Python or you are interested in learning, you can download it free online! It is open source and very powerful! If your novice, I recommend downloading Python Anaconda, it has everything you need!
Excel and SPSS (what I use) tend to be costly. Excel is the cheaper alternative, I think it costs me about 75$ a year (however, I am still a student so I get the student discount, not sure full price).
SPSS, MATLAB, SAS are extremely expensive. In excess of over 2,000 USD. There is an option to do an annual licensing agreement for less, but the price would add up.
Q: Do I need a degree in mathematics or computer programming?
A: NO! You don’t. You can learn this stuff from books and reading. Having a degree doesn’t even guarantee you that you will understand this stuff. I speak to some of my classmates about what I do, and they still don’t understand what I am doing (despite also having MScs in statistics hahaha). It all comes down to your critical application of knowledge! Education is very important, IMO, but its not everything and everyone has the potential to learn if they are truly motivated to!
My background was I started as a nurse with a bachelors of science. I fell in love with mathmatics and statistics in my undergrad and ended up pursuing higher education in mathematics, specifically applied statistics.
Q: How is quant trading different then technical trading?
A: So, as I wrote above, technical trading is the qualitative appraisal of quantitative data. I am not a technical trader and can’t speak too indepth about this process.
But I can contrast a little bit, which I will do below!
A technical trader may look to see that a particular price point was respected and not surpassed over a number of days. They would likely label this as strong support and would assume that a break of this support would lead to more sustained selling.
Contrast to a quant trader, I do not pay attention to any one specific price point. Price action tends to be more on the random side. So I rely on all of the data over many years of trading to develop working ranges and variances between the data. From this, I can determine the range that a stock likes to operate in (whether it be +/- 10 points or, if its TSLA, +/- 30 to 60 points). From here, I can use previous day data to predict a likely range for the next day. When I have that range, I can then express my hypothesis in conditional algebraic forms, like:
IF Condition 1 met THEN statement 1 correct AND statement 2 incorrect; or
IF Condition 2 met then statement 2 correct and statement 1 incorrect.
I then follow linear algebraic principles to identify those conditions and subsequents.
For example, for today, SPY opened around 420.28. The range that I calculated for SPY today was 415 to 427. So, the problem that I needed to solve mathematically was:
If Condition X met then SPY = 427; OR
If Condition Y met then SPY = 415.
Then I must use algebra and statistics to determine what Condition X and what Condition Y are.
If you read my ideas, you will notice that I express my ideas in linear algebraic form. For my post about SPY today, this is what I had wrote:
A break above 424 would indicate bullish sentiment and likely continuation towards 427.
A break below 418 would indicate bearish sentiment and likely continuation towards 415.11.
If you notice, this can be expressed as a conditional (algebraic) statement:
IF X > 424 THEN 427 is met; OR
IF X < 418 THEN 415 is met.
Now I don’t manually do this because it would be to labour intensive. Which is why I say you need to know a programming language. You can program Python, Excel, MATLAB, C++, etc. to do this for you and identify those ranges. But you need to have the theory in order to understand how to get there and how to give Excel, Python, C++ or MATLAB what it needs to solve the problem for you.
Q: Can you recommend books or videos on quant trading?
A: So, I have not found any quantitative retail traders on youtube. There are 2 quant developers that actively post on YouTube who have okay content, one being Trading Jesus and the other is Korean Yuppie (who is still kind of novice and hasn’t posted much). Both are from the perspective of quantitative developers; however, this is a completely different skillset from a quantitative modeller. But equally interesting and informative!
In terms of books, I would recommend general statistics books and books on programming language like Python or even books on Excel. Excel is generally an under-rated platform that is capable of quite advanced data analysis. Don’t under-estimate it! Excel is involved in my trade planning, execution and profit taking process. It is the thing that dictates what I should do and where I should enter/exit.
You also need a solid understanding of the market, how it is organized and how it functions. So general books about market theory and trading are also useful. I have no
real recommendations as I haven’t read any books, aside from The Trading Zone, which I found insightful but not helpful. Most of the information you need is available for free online. I wouldn’t invest a huge amount of money in books that are mostly fluff, especially books on day trading.
Equally, avoid courses! Don’t buy people’s courses and don’t trust trading “gurus” from YouTube.
Hope this answers all of your questions, again please let me know if you have any that I have not addressed!
Thank you, take care and as always, trade safe!
Premium indicators and content have launched! Get access at: www.patreon.com/steversteves