As I mentioned in the previous chapter there are two ways to pair trade. We will be starting with the correlation-based technique. This is a very common approach, as many traders use it to get their pair trading hands held.
Before we can begin to use the technique, it is important to understand a few jargons. This chapter will discuss tracking pairs. I want you to be able to identify what I mean at this point. As we move on, we will connect the dots.
Spreads The spread is the most widely used jargon in the trading industry. If you're scalping the market, the spread is the Rupee difference between the ask price and the bid price. If you're doing an arbitrage trade, the word spread is the difference in the prices of the same asset on two different markets. The word spread is used in pair trading (actually, it's a correlation-based technique). It refers to the difference in the closing prices for two stocks.
Spread is calculated as:
Closing Value of Stock 1 - Closing Value of Stock 2
Have a look at these -
If I take GICRE and ICICIGI together, then the spread will be -
Spread = 6.1-3.85
Both 6.1 and 3.85 represent a change of stock price relative to the previous close. Both numbers are positive. Let's say that the closing price for ICICIGI was 3.85. In this scenario, the spread would be -
The spread has been calculated for the past few trading days. This should give you an idea about how it 'runs'. The spread is also known as the historical spread by traders because it was calculated daily.
The spread changes on a daily basis, as you can see. Here is another interesting observation.
There are many other combinations that can lead to expansion or contraction of spreads. We'll get to this more later.
Differential The differential is a measure of the stock price difference. The absolute difference between the stock closing prices of two stocks is called the differential. Below is the formula.
Closing price of stock 1 - Closing price of stock 2
If a stock 1 closes at Rs.175 while stock 2 closes at 232, then the differential is -
175 - 223
= - 57
You can do this as a time series, and then calculate it on a daily basis. This is what I did for ICICIGI and GICRE.
You need to be aware of differentials. Spreads can be used to track pairs on an intraday basis. Spreads are not the best way to track pairs, so the 'differentials" is best used at the end of each day.
These things will be covered in more detail later. Let's not get into too many jargons.
Ratio I find this ratio quite fascinating. This ratio simply involves dividing stock price 1 by stock price 2. It can also be reversed.
Stock price of stock 1 / Stock price of stock 2.
Here is the result of my calculation:
As you can see, the Ratio is more consistent when it's calculated as a series of time. All three variables have been represented on the graph.
What are spread, differential, ratios, and their relationship to pair trading?
These are, as you can see, the variables that help us quantify or measure the relationship between two stocks. We refer to them as pairs. This graph shows us the relative movement of the stocks. If we look at the spread, for example, it expands when S1's closing value is positive, while S2's closing value is negative, and contracts if S2's closing value is also positive.
The ratio is the same. If stock prices drop, the ratio decreases. Stock prices rise. There are many other possibilities. Stock 1 can decline rapidly while stock 2 remains flat, or vice versa. Stock 2 may also increase in value relative to stock 1.
It's not hard to understand, is it?
We need to see the chart for the variable that we are following. It could be spread, differential or the ratio. The spread of the variable must be tracked and figured out whether it is expanding or contracting. These are the two next jargons.
Divergence – If the ratio between the stocks or spread is expected to change, or alternatively, if you expect the graph will move up, this is called a divergence. You can make money, or at least try to make money by creating a divergence trade if your variable is expected to diverge.
Convergence If the spread or ratio between the stocks is expected to move towards one another or you anticipate the graph to move downward, this is called a convergence. A convergence trade is when your variable is expected to converge. This can help you make money or at least try to make it.
Here's the big question: What makes you believe that the variable can either diverge or converge? What is the best time to start a trade? What are the triggers for a trade? What are the triggers?If the trade doesn't work ,What will happen? What is the stop loss for these trades?
Before we answer these questions, let's first ask how can we make two stocks a pair? Does it mean that two stocks are a part of the same sector? For example, ICICI Bank and HDFC Bank are both in the private sector banking category.
We need to use the old statistical measure known as the "Correlation" in order to classify two stocks as a pair. We have probably discussed correlation many times on stock market box lessons.Following is a short and quick explanation.
The correlation between two variables can give us an idea of the movement of two variables in relation to one another. Correlation can be described as a number that varies from -1 to +1. If the correlation between two stocks exceeds +0.75, it means that there are two things.
The above shows that a correlation of +0.75 indicates that the variables not only move in the same direction, but also tends to move closely together. The correlation does not indicate the magnitude of the move. It only suggests that it is likely to occur. If Stock A moves by 3% and the correlation between stock B and stock A is +0.75, it doesn't mean Stock B will move by 3%. The correlation only suggests that Stock B will rise in the same direction as Stock A.
There is a twist to this story: suppose Stock A and stock B are correlated at 0.75. If Stock A moves higher than its daily average returns of 0.9% and 1.2% respectively, then Stock B is more likely to move above its daily average returns of 1.2%.
A correlation of -0.75 means that the variables move in an opposite direction (-ve sign), but they tend to move in the same direction. Let's say stock A goes up by +2.5%. The correlation tells us that Stock B will likely fall, but it is not known how much.
One more thing about correlation while we're at it. This is for people who are interested in the mathematics part of correlation. Only if the data series are'stationary around their mean', the correlation data will make sense. What does this mean? It simply means that the average values should be maintained by the data set.
This line "stationary around the mean" should be kept in your head. Don't forget about it. This is what we will be referring to when we talk about the 2 and techniques to pair trade. It happens much later in this module.
As a way to determine how closely two stocks are linked, we will use correlation. We will be discussing two types of correlation in the next chapter.
Spread, Differentials and Ratios are important concepts that I will be covering in this article.