Anyway, I was just musing over a key criteria in evaluating companies. Among other things, it must have:
1. High consistent ROE, > 15% over 10 yrs or more, or if 5 yrs, have a higher margin of safety
2. High consistent ROA, >7%, same time criteria as above
ROE is calculated by taking the net profits divided by the equities portion. You can find net profits in the income statement, usually right at the bottom of it (it's literally called the bottom line). Equities is found in the balance sheet. Equities is what you are left with after subtracting total liabilities from total assets. ROE itself changes depending on how much debt you use to leverage to increase your earnings, so it's very important to consider the debt profile of the company together with ROE and not just treat ROE as a stand alone matrix.
You can break up ROE into the following 3 components to better clarify what exactly drives the ROE.
ROE = Net profits/equities = Net profits/Revenue x Revenue/Assets x Assets/Equities
Net profits/Revenue is the net profit margin. It measures how many dollars you book as profits after subtracting all the costs of selling $1 worth of goods/services. It's important to realise firstly this might not necessarily translate into cash straight away because this is just an accounting profit. The actual cash might come in later (or not) and can be seen in the statement of cash flow. Obviously, the higher the net profit margin, the better it'll be.
Revenue/Assets is the component that measures how efficient your conversion of the assets to revenue. It varies widely from a company that sells goods to a company that sells services. Generally, a company that sells services are lighter on assets, so their revenue/assets ratio will also be higher than say, a company that sells shoes. Again, the higher the better.
Assets/Equities measures the gearing (some sort) of the company. Assets (A) = Equities (E) + Liabilities (L). If the company has a lot of debts, their liabilities portion will be substantial, thereby raising this ratio. Out of the 3 components, this is a ratio that might not be good if it's higher. Given the same ROE, a company with low Asset/equities ratio is generally in a better position than one with higher Asset/equities. With debts, you can leverage a higher profit, but it goes both ways too when the economic situation becomes dire. You can lose more than you owe. Short-term wise, with debts, generally ROE will be raised (but at what cost?) because it's just a product of the three components.
But I want to introduce ROA into the picture as well. ROA is defined as follows:
ROA = Net profits/Assets
It measures how efficient the company is at converting its assets into net profits. Generally, the higher the better. However, because the denominator is Assets, and Assets = Equities + Liabilities, essentially:
ROA = Net profits / (Equities + Liabilities)
Now compare this with ROE, which is:
ROE = Net profits / (Equities)
and you'll realise two things:
1. Mathematically, ROE is greater than or equal to ROA, because the denominator for ROA is larger or equal to that when calculating ROE
2. ROA approaches ROE when Liabilities approaches zero. For a company with zero debt, there should be no difference between the ROA and ROE
That was my hypothesis anyway. It formulated when I was at a coffeeshop waiting for my seafood hor fun and I got myself thinking after looking at the 2 criteria stated above (high ROE and high ROA) for good companies. I knew I got to test it out, so let's see it in real life cases.
It's instructive to compare 2 companies with similar ROE and then compare their ROA, after which we see their Debt/Equity ratio.
1. The two green columns - ST Engineering and Sheng Siong - have very similar ROE at 26+%. That's very high. But notice also that while ST Engineering has a much lower ROA, Sheng Siong has a much higher ROA. If we look at the Debt/equity, we notice that ST Engineering has a lot more debts than Sheng Siong. This by itself isn't a red flag, but it's just something we have to take note of. Debt is only a problem if they have a problem paying it off. We can see that the high ROE of ST Engineering is partly fueled by its debts.
2. The two purple columns - GDS and Vicom - again have very similar ROE at around 22 to 25%, which is very admirable. This time, both have very low debt/equity ratio and thus their ROA matches very closely to their ROE. GDS especially, is a company that I didn't know existed until I tried finding a company listed in SGX with the lowest Debt/equity ratio. Having a high ROE (25%) and a high ROA (19%) plus a good yield of about 3.8%, it's one of those small caps that might be worth while for investors to dig deeper.
3. The last 2 red columns - Sembcorp Ind and Hock Lian Seng - also have pretty good ROE of around 16.8%. But we can immediately see the difference in their ROA. Sembcorp Ind has still a decent 8% ROA but Hock Lian Seng's ROA failed the criteria of having ROA being 7% or more (it's 4.8% in 2013). And we can observe why - the debt/equity for Hock Lian Seng is much much higher than that of Sembcorp Ind.
I think if we track the ROE and ROA trends of a company across years, we can uncover something interesting. ROE by itself is just a number; we have to check the quality of that number. If it's maintained by higher and higher debts, we can see it in their ROA trends too. Basically you don't want a situation where the ROE keeps increasing and the ROA keeps decreasing. A divergence of this two matrices means only one thing - the debt is piling up.
But of course, I still highly recommend the elegant Dupont's analysis of ROE that I've blogged about at length and summarised here. It clearly shows which of the 3 components are the ones that brings the bacon home and maintains the ROE.