tldr; When we talk about diversifying the portfolio generally it means to have all assets, but if there is a correlation between assets then still it will have risk as all related assets tend to move in the same direction. example, HDFCBANK, and BAJAJFINSV should not be treated as two assets. Instead, they should get a weighted ratio by price to reduce the risk.
When we talk about diversifying the portfolio it simply means put eggs into different baskets so that risk of damage is reduced. You may not get optimal results due to a tradeoff between risk and reduced return. In a practical sense, it is like buy IT, Metal, pharma, bank all of them instead of investing all money into one. let us say a balanced portfolio and personal risk preferences you choose 40, 10,20,30 into these sectors. This is one way to choose the stocks however we can also use past data for finding a balanced portfolio. For example, if we say Gold increases when all stocks decrease or vice-versa then Gold could be a good asset to have it in the portfolio. We are looking for this -ve correlation between these two assets and reduce allocation for assets like Silver which might be correlated with gold.
If we consider two series of values that are correlated, they look somewhat like the below graph, both following similar directions most of the time. Below lines are having .99 correlation which means they are highly correlated, if they are opposite then correlation goes negative.
Scatter plot of the series gives a different perspective of the data:
Now let us look at NIFTY analysis for correlation:
If you carefully see the above graph, dark blue indicates a positive correlation. i.e, both stocks likely to move together (maybe because of similar customer base, products or context). In such cases consider all related stocks as a single asset and divide the allocation between them instead of allocating them as independent assets.
Consider the Red color stocks as a complement for Blue ones as both together are likely to manage the risk. Be cautioned about long term drags like sun pharma which just has -ve correlation with any stock that is doing good. Again this is just another indicator or quality factor to consider should be done with other checks like fundamental information and any ongoing issues with the company.
As usual learning every day and sharing what I understand. Please let me know your thoughts!
