Warren Buffet once said “diversification is a protection against ignorance – it makes very little sense for those who know what they’re doing.” This is perfect for me, because I have no clue what I’m doing and have little interest in becoming an expert in whatever I’m buying into.
I’m assuming we all know about risk and everything. But I want to pad the word count on this a bit so:
- Unsystematic risk: the risk you take when you punt on a stock – it’s lower when you’re buying well-established gigantic companies with huge market caps and way higher when buying speculative pennies, however your rewards are usually reflected in that.
- Systematic risk: the risk of owning shares. You can’t do anything about this, you just have to accept that occasionally the entire market can and will shit its pants.
So you can’t do anything about systematic risk (other than be like hurr durr I’ll put my money in savings and lose it to inflation) but you can deal with unsystematic risk by diversifying your holdings. But at what point does diversifying become pointless?
Turns out there’s no exact point, but you hit diminishing returns in terms of risk reduction with every extra stock you add.
The way risk is measured in a numerical form is usually with “standard deviation” – basically how much your portfolio differs day-to-day versus its average performance over time. Normally you want to reduce standard deviation so you don’t have a heart attack every other day, but if you’re chasing really high returns on risky stuff you’d have made your peace with this.
So at what point does it become pointless? It depends, based on a few different studies. What I’ve done here though is mash up a table from the Institute Of Business & Finance together with one from page 57 of Modern Portfolio Theory and Investment Analysis (a math-heavy book that seems to get referenced a lot in these conversations) that found:
|Number of stocks||Standard Deviation (IBF article)||Expected Variance (book)|
I mashed these together because I couldn’t figure out which one was more “correct” as far as numbers go – the IBF page references this paper (written by the same authors as the Modern Portfolio Theory and Investment Analysis book) but it goes on about standard deviation instead of variance (which are two different things) – either way you’ll note that the difference between numbers in both columns really drops off beyond 20 stocks. So if I was buying individual stocks specifically to hedge against unsystematic risk, I’m probably wasting my time after the 20th – not to mention they all need to be different to each other: if my entire portfolio is 20 different BNPL companies then that doesn’t count as being diversified. Also this is a good time to stress once more that I don’t know what I’m doing and have only skimmed most of this. It’s interesting though!
So, beyond 20 stocks gets rapidly diminishing returns in terms of risk reduction. There’s also the the overhead of brokerage charges and having to keep up with the news on what every single once of those companies is up to. FUN!
The easiest way to get around this is to just buy a broad-based index ETF. VAS, A200, SPY, VGS or whatever – boom, buy one and then I indirectly hold stock in several hundred companies, thus instantly diversifying away risk without me having to think about it. I tell myself that my sensible approach is to have half my holdings in index ETFs and the other half in whatever floats my boat, however the reality is that I constantly get tempted to put more and more in individual stocks. At the time of writing, I am split 63/37 between stocks and ETFs. Will this come back to bite me in the arse or will I correct it in time? I’ll get back to you in a few years.