When it comes to investing, I love thinking in terms of standard deviation. It’s one of the quickest ways to assess the risk of a potential investment, and there is an established good benchmark standard deviation that I shoot for when picking investment strategies. There is just one small problem – the close. Statistics, though one of the most helpful maths in the investment world, is rife with biasedness. Whether that biasedness is intentional or not, it’s still a big problem. One such parameter that is heavily influenced by this bias is Standard deviation.
Here you see a screen where I look for the highest standard deviation’s in the market. What do you notice?
They are all huge companies! A lot of the companies that appear in this list are some of the biggest and most stable companies on the market, so why are they showing up on this list? I want risky!
Let’s think back to class a long time ago, how is standard deviation calculated?
If it’s been awhile since you’ve been in school, this is looking for the square root of the sum of the given value minus the mean squared for every number in the sequence over the total number of items in the sequence. And if that wasn’t clear, that’s no big deal, because we calculate the standard deviation for you in Equities Lab. You don’t really need to know the specifics of the math, just how to use it.
Here’s where the problem is. If we just look for the standard deviation of the close, we will return a dollar amount. The more expensive the company the higher the dollar amount of even a small move. As a result, when we change our screener just slightly to find the top .01% of companies based on their standard deviation we only get one result.
So, let’s remedy this situation by normalizing the data. We can do that by doing one little thing to the line of Equihack code, and that is by looking for the change of close rather than the close. The change of close is a value that isn’t going to be influenced by the closing price of a stock. Instead all stocks move X% over a period of the year and it doesn’t matter if the price of a share is $0.01 or $1 million, that percentage won’t change.
Here are our adjusted results. The screener went from returning household names to returning the names of companies I’ve never heard of whose standard deviations are at hilariously high levels.
I write this article because even though statistics are helpful, and our system is a wonderful way to go about computing values, it is meaningless unless you’re diligent and are making sure you are using correct parameters. This constant checking and rechecking of our math and data is why we believe in so much transparency between a company like ours and our users. If you didn’t have a chance to double check if the math is right, there’s no reason for you to trust the system. Thankfully for you, we allow you to dive into every calculation we do within the system when it comes to screeners. You also have the ability to adjust these calculations anyway you like to best fit your personal investment needs. If you are careful and diligent the software will be the perfect companion for your portfolio.