Here’s a special case that needs to be discussed.
Here is what you need.
I investigated a pair trading algorithm and found an appealing trade. The following are the regression parameters –
What do you think? It’s a match made in heaven—ICICI and HDFC, two of India’s largest private sector lenders. Not only do they have a comparable business model, but their revenue systems are also alike, with the added bonus of being regulated by the Reserve Bank of India. Could this be the ultimate pair trade?
The Adf value of 0.048 signals that there is only a 4.8% likelihood of the residual not being stationary, which is exceptionally good news at 95.2%.
The residual value of +2.67 makes it perfect for initiating a short pair trade, i.e., HDFC gets sold and ICIC bought.
So, what are the price and lot size details? Here they are –
HDFC Fut Price = 2024.8
HDFC Lot size = 500
ICICI Fut price = 298.8
ICICI Lot size = 2750
We previously discussed position size in the last chapter. To calculate the number of shares needed for this transaction, we can look at the beta.
Beta 0.79 tells us that we need to buy 0.79 shares of ICICI (X), for each one of HDFC (Y) we possess – the lot size for Y being 500. Therefore, if we want to balance our portfolio with the beta value, 395 shares of X should be held.
Can you spot the issue? The lot sizes clearly aren’t compatible.
We cannot stick to the same trading pattern as seen in the TATA Motors and Tata Motors DVR example discussed in the previous chapter. Doing so would compromise its beta neutrality.
To make sure we get this right, we must be precise when it comes to the quantity of our lot sizes.
The lot size of ICICI is 2750, with a beta of 0.79; HDFC has a lot size of 500. What would be the lowest number of HDFC shares required in order to achieve beta neutrality for 2750 shares of ICICI?
To discover the answer, we can divide –
Since the lot size of HDFC is 500, we can round this up to 3500, giving us 7 lots of HDFC compared to 1 lot of ICICI.
Now that the position size has been determined, the question remains – should you undertake this trade?
It all looks great – ADF has desirable value and the residual is at 2.67 SD, correlations between the two stocks are high, and the two businesses are quite similar. What risks could be lurking?
Yes, from a closer glance, it appears fine, but on closer inspection the intercept paints a slightly different picture.
In order to grasp the concept, let us take another look at the regression equation.
y = Beta * x + Intercept + Residual
We are aiming to understand the stock cost of Y by accounting for the stock cost of X and its beta. The intercept is that part of the y’s value which the model cannot account for, with the residual being the gap between the anticipated and actual y.
From this, a significant intercept can be observed, implying that a large part of Y’s stock price is not captured by the regression model.
In this case, the intercept is 1626. HDFC’s stock price is 2024 per share – this implies that 80% (1626/2024) of it is beyond the explanation of the regression equation. This means that only 20% can be elucidated by the equation, which can be considered quite difficult to grasp.
This further suggests that trading this pair involves a low chance of success. As someone who looks at the risk before reward, I would rather steer clear of it than try my luck. However, some traders might see this as an interesting opportunity, so it’s really up to their preference and risk assessment.
Best of luck!