A simple yet highly scientific explanation of what is diversification, why is it important, and how should you achieve it?
What is diversification?
As the old saying goes “You should not keep all your eggs in the same basket”. Diversification is keeping your eggs in different baskets.
Diversification lowers risk. Risk can be measured by ‘variance’. Therefore, diversification reduces variance.
The below write up about variance looks a little mathematical to begin with, but trust me – it is not complex mathematics. If you read it a few times, you will understand it well. Even if you do not understand the math, continue to read and you will get the gist of this post.
What is Variance?
In most simple terms, variance is nothing but a measure of how far individual values are from the average value.
Look at two portfolios below – both have the same average return (as we will soon confirm) but Portfolio-1 has a higher variance: the returns in individual years are ‘spread’ further from the average (mean) of 5%.
Here is the data that I used to plot the above chart (please do not get discouraged by looking at so many numbers – I assure you that this is the most mathematical part of the post):
Portfolio-1 returns -20% in one year (20% loss) and gains 30% in next year, so on and so forth for the next 10 years. Average return= +5%
Portfolio-2 returns -5% in one year (5% loss) and gains 15% in next year, so on and so forth for the next 10 years. Average return= +5%
Both have the exact same average return of 5%.
Why does Variance matter?
Continuing with our example of the two portfolios, which one would you prefer? Any rational investor will prefer the one with lesser Variance (Portfolio-2).
For the same average return, rational investors would like a portfolio with minimum Variance.
How to minimize Variance?
Short answer: by adding more stocks to the portfolio. 1-stock portfolio almost always has more variance than a 2-stock portfolio. 2-stock portfolio almost always has more variance than 3-stock portfolio.
Long answer: In order to understand this, we will need to define one more term ‘Co-variance’. (Co-variance is the same as correlation).
In probability theory and statistics, co-variance is a measure of how much two random variables change together.
We want to make things as simple as possible: positive co-variance is when two variables move in the same direction. For example, outside temperature and electricity bill for a cooling device – when outside temperature increases the usage (and hence cost) of cooling device increases.
Another example is – sale of umbrellas is positively correlated with the number of days it rains. More the rain, more the sale of umbrellas. Lesser the rains, lesser is the sales.
Another example – an increase in car production will lead to an increase in Aluminum prices.
On the contrary, negative co-variance exists when two variable are likely to move in opposite directions. The more time Lisa spends at the mall, the less money she has in her bank account. Her spending time is negatively correlated to money in her bank account.
An example close to my heart: The higher the expense ratio of a mutual fund/ ETF, the lower the investment returns.
Adding co-variance (positive vs negative)
Let us continue on what I said earlier – portfolio returns variance can be reduced by adding more stocks to the portfolio.
To begin with, let us assume that there is only one stock in the portfolio. What happens when you add a second stock to the portfolio? The variance of the portfolio either goes down or stays the same.
It stays the same if both the stock returns are perfectly positively correlated -> Whenever stock-1 returns 0.94% more, the stock-2 also returns 0.94% more; and Whenever stock-1 returns 0.59% less, the stock-2 also returns 0.59% less. As you can see this is so rare that it is unlikely to happen in reality.
For any other correlation between them, a 2-stock portfolio invariably has lesser variance as compared to 1-stock portfolio. So adding a second stock to a 1-stock portfolio reduces risk/ variability.
Let us take an example, say you have $100:
Scenario – 1: You buy JP Morgan Chase bank stock with $100.
Scenario – 2: You buy JP Morgan Chase bank stock with $50 and you buy Bank of America stock with $50.
In which scenario do you think there will be less variability in returns? In other words, in which scenario do you think your returns over the next few years will be more stable?
Answer is Scenario – 2. Why? Because you are spreading your risk across two stocks. Let us say that a senior executive of one bank meets an accident and decides to take a year off OR the bank is found engaged in illegal activity OR the bank is fined a large sum by regulators -> the list is endless.
But the chances that both the banks are so adversely affected in the same period are less (than each one of them being affected).
But you may ask: Since both Chase and Bank of America are banks, what if there is a new law/ regulation in the market that affects both the stocks similarly. For example, if the Government decides to charge the banks 5% extra on their profits. then both the stocks will become less profitable. Similarly, if the Government decides to give big banks a tax break, then both the banks will be more profitable.
It is easy to visualize that the two bank stock returns are positively correlated.
On the contrary – if you buy $50 worth JP Morgan Chase stock and $50 worth AT &T stock, then the chances that they benefit/ hit by the same Government regulation are less likely.
So here are the three portfolios in decreasing order of risk/ variance:
Portfolio-1: $100 JP Morgan Chase stock
Portfolio-2: $50 JP Morgan Chase and $50 Bank of America
Portfolio-3: $50 JP Morgan Chase and $50 AT & T
So what should you do? You want to hold stocks that are not closely correlated (not positively correlated); simply stated you want to hold stocks that are not similar.
Portfolio returns variance can be reduced by adding more stocks ‘with a negative co-relation‘ to the portfolio.
We started with a single stock portfolio. We added another stock and the risk went down. We added a third stock and the risk went down even further. Scientific studies have shown that risk continues to go down till about 30 to 50 different stocks in a portfolio.
Now we conceptually understand how to diversify a portfolio. Now let us look at the same concept with respect to the asset classes.
From an earlier post, we know what is an asset class. We also know that the main asset classes are domestic large cap, small cap, developed nations, developing nations, bonds, and alternatives.
Since investments within the same asset class share the same riskiness and return, it makes most sense to diversify across asset classes.
The image below depicts a well – diversified portfolio.
“Inside out” in stead of “Outside in”
We earlier started with JP Morgan Chase and then added Bank of America or AT & T to the portfolio, and then discussed how adding more stocks increases diversification and reduces risk.
There is another way to look at the same image: Do not start with JP Morgan Chase. Start with the other end of the visual – who is at the other end? You 🙂
You are @ the center of the portfolio. You want to have exposure to Domestic Large Cap, Domestic Small Cap, Developed countries, Developing countries, Fixed income instruments (Bonds), and alternates.
For each of those asset classes, you want to have exposure to multiple entities. For example – for Domestic Large cap asset class, you would like to have an exposure to all the 11 sectors, namely Financials, Telecom, etc.
For each of the sector, you want to have exposure to multiple players. For example – JP Morgan Chase, Bank of America, Citibank, and other banks within the Financials.
Now come back to the center of the portfolio – you. You are affected by anything that happens anywhere in the outermost layer of the visual. If Bank of America makes money, you make money. If Gold prices increase, you profit. If Toyota files for a patent, you are like to make money.
Will all of them make money? No. Will most of them make money? Well, over long term, the worldwide GDP grows. So it is statistically likely that ‘more number of them’ will make money than ‘lose money’.
Some smart people always ask me – why don’t I just pick and choose the ones that will make money? Because you do not have a ‘vision ball’. You do not know which ones will make more money than others.
Will diversification eliminate all risk?
No – it certainly does not eliminate all the risk BUT it does eliminate all the risk that is ‘diverisifiable’.
Every investor should diversify. Investing without diversification is gambling.
With diversification, you might make ‘less’ than your friend (who does not diversify) in some years but it also means that you will make far more than your friend in other years.
What is important is the fact that your returns will be somewhat consistent.