[You can download the Diversification (Fundamentals) Excel from the Resources page]
We've all heard the advice to "diversify", but what exactly is it? This question stumped me when I first started on this journey. After all, if I hold more different stocks, wouldn't there be a higher chance that something bad happens to a stock I hold? It took me a while to realize that I was thinking of (expected return) = (gain/loss) x (likelihood), and that is a gross oversimplification of the problem. If you go back to the basics and think about the problem from the ground up, it's not as abstract as people make it out to be.
Consider the following example: you invest $256,712.82 (your bank account balance) into DBS. It's a solid company (or so you think) and the chances of it crashing is miniscule. Two days pass, and you finish using that $27.50 in your wallet, and you must either starve, or liquidate the investment. You can maybe skip a meal or two, so that gives you about 24 hours to sell at the best price you can get within that timeframe. Now for some reason, some investment fund decided to reduce their holdings on that exact day, pushing its prices down. You KNOW it will rebound once the selling is over, but you don't have a choice, cuz you have exactly half a packet of milo left on the shelf and are about to faint from hunger. You take the L and sell, crawl to the nearest ATM to withdraw the cash, and curse that fund manager. This is why people say to diversify across multiple different stocks. A company's CEO can get arrested, tax audits can fail, buildings can collapse; there are a million and one ways an otherwise healthy-looking company can fail. The issue isn't how unlikely, it's that IF it does happen to your stock, you have zero wriggle-room.
Okay, so that wasn't a good idea. Now what if you split that bank balance into UOB, DBS, and OCBC? All you need is one stock whose prices are fine, and you can sell those first, and you'll have a few more days' worth of food while you wait for the prices for the other tickers to recover. Well, maybe MAS released a statement that morning that Singapore's banks will be put under more stringent oversight due to a series of high-profile phishing scams / hacks. Boom, now all 3 stocks are hit. Again, it's sell or starve, so now you curse yourself for concentrating your position into stocks that are affected by the same piece of news. This is why people say to diversify across sectors.
"Fine, I'll buy DBS, Capitaland Integrated Commercial Trust, and Singtel. That's banking, real estate, and utilities. Now we're good, right?" Well, last I checked the majority of these companies' revenue come from Singapore, and guess what? MTI releases a report that Singapore's economic growth is expected to be bad for the next few years, and boom all 3 stocks are hit again. This is why people say to diversify across geography. They don't necessarily mean to buy stocks listed on overseas exchanges. If you buy Capitaland China Trust on SGX, Ping An Insurance on HKEX, and China Eastern Airlines on NYSE, you're still very concentrated in China. If China's economy crashes, they're all gonna be hit. On the other hand, you can have a basket of stocks all bought on SGX, and still be geographically diversified. In fact, many tickers on SGX have at least about half their revenue based NOT in Singapore.
"Okay I'll buy the index, and I know you'll cook up a scenario where somehow the fund gets into trouble, so instead of buying the Vanguard STI ETF or the SPDR STI ETF, I'll just manually buy every stock in the Straits Times Index!" Nice try. Russia invades Ukraine, inflation is sky-high, the US FED says they can no longer promise a "soft landing for the economy", and ALL stocks crash (sounds familiar?). Now, if only you didn't have to sell within a day. What if you had invested only half of your bank balance instead of the whole thing, and you can afford to wait a few years for prices to (hopefully) recover, and then sell? This is why people say to diversify across instruments (cash is an instrument, too).
Now, let's go back to the thing about (expected return) = (gain/loss) x (likelihood). By holding cash and 30 different stocks across different geography and sectors, anytime a piece of bad news is announced, you're pretty much guaranteed to get hit. However, you do not need to realize that loss! That's what's being done in the examples above: it doesn't matter if a stock/sector/geography/instrument is hit, as long as you have something that is NOT hit, you can sell/use that instead. I think this is an important tenet of diversification that often gets glossed over even though it's very relevant to the average investor. After all, you can talk about "theoretical long-run returns" all you want but for the average investor (not fund manager), when you need to sell, you need to sell.
What all of the above boils down to, is essentially the following: "If/when a piece of bad news appears, in whichever stock/sector/geography/instrument/etc, how much of my portfolio is affected?" In my Excel, I assume "how much" is based on the company's revenue breakdown for that category (for stocks and bonds). It's a simple Excel with tables and charts to help illustrate and visualize how "diversified" your portfolio is, without any complicated formulas (it's all just percentages). The hardest part is reading the annual report for every company you're working with (which takes some effort), and using appropriate inputs for your expected use-case (which may be very subjective).
In this example, you can see that I have some very dubious inputs (e.g. Splitting SIA's "East-Asia" revenue 4-ways between Singapore, China, India, and Japan). This was done because my end goal was to figure out how much attention I should pay to various news articles. If I had unlimited time and energy, I would of course strive to understand everything that's going on in the world. Unfortunately, I do not. If I had a portfolio that's 90% based in the US, I would not only read news articles about the US, I would make sure I follow every Fed meeting, understand how each state/electorate/mayor works, etc. In the example, most of the exposure is in Singapore, US (through the USD), China. In a pinch, I could probably drop US since cash shouldn't be as volatile as stocks. In the example, though, I think any bad news in any of those 4 countries will affect SIA's passenger volume, which will in turn affect its revenue in East-Asia, so when I read the news, I will want to divide my attention between those 4 countries. Hardcore analysts and purists are probably foaming at their mouths right now, but hey, it works for my purposes (i.e. to decide how I will split my attention when reading the news).
As it turns out, I could probably focus on understanding the economies of Singapore, US, and China for a start, and follow what's happening in those 3 countries. In terms of sector breakdown, I could probably read up more on the Finance and Real Estate sectors and ignore the rest (the 46% "Others" is, as we know from our input, simply due to the SGD and USD cash holdings, which shouldn't be too volatile). In a pinch, I may not even bother learning too much about how any of the sectors work, and instead use the time to write more blog posts.
With that, I'll wrap up and say that this is by no means a definitive method or tool to help you "diversify". As explained, it ultimately depends on your end goal (avoiding being overly vulnerable to certain types of news, deciding how to split your time and attention, etc) and your individual circumstances (Do you need/want to sell your stocks within a day or a year?). This is simply meant to throw a few ideas out there, provide an easy-to-understand method/tool (albeit possibly somewhat flawed), and let you figure it out. Have fun!
Comments