The problem with RoE (or RoA) is that the denominator e (Equity) or A (assets) is "path dependent". This means that in particular the intangible part of assets can have different values depending on what happened in the past (mergers, good will writedowns, accounting treatment of goodwill etc.) if you leave goodwill out entirely, you will get very high ROA/ROE values because the medical companies in questions (COV, BEC, BAX, MHS, DGX) need little hard assets to operate their business.
Of the two metrics ROE and ROA I always prefer ROA because it accounts for Debt orleverage. You may not care truly from a shareholder perspective and only say that owners earnings count, but I think debt is important because it a) increases the risk for the shareholder and b) and acquirer would only care about the EV (which includes debt).
So ROA only makes sense with a few disclaimers as to what is counted for as assets and how management has treated the intangibles in the past (how do you account for good will writedowns if they have occurred in the past ?).
I like EV/EBITDA because it is a number that is harder to manipulate then earnings and it's easy to calculate knowing that looking at this ratio alone will penalize a well run business that makes effective use of their cash flow,relative to mediocre or bad ones. No ratio is perfect and each of the approaches (ROA, EV/EBITDA) makes sense, imo.
As to medical companies, I consider BEC and COV with PE's of around 11.5 as attractive. on an EV/EBITDA basis COV looks a little more expensive (even thought he PE is similar to BEC) because it is domiciled in a tax efficient location, so it pays a lower tax rates which leaves more for the owners.
I pay a higher multiple for a manufacturing business like COV or BEC than a service business DGX because the latter has higher risk, imo. I think the whole bunch is starting to look increasingly attractive assuming that the fundamentals stabilize. |
