No. of Recommendations: 43
Some of us just maybe do not elaborate in our writing, maybe because we are not so much interested in doing so or possibly because when we do it doesn't come out so well.
Agreed, mungofitch writes well, but he also rather consistently presents (and has here presented) sound logic, and facts - and this time a one-off bonus of some original empirical research. I have find in my own backtests that ROE is often pretty reliable in offering benefits as far as any single measure, other than momentum which for technical reasons (pun not intended) is about the only mechanical investing variable that cannot be arbitraged away with heavy widespread use.
Are you implying that he merely wrote elaborately in his last post, but in some sense rhetorical such that there was a weakness in the actual argument? If so, can you clarify the logical or factual weakness rather than blaming just his clear style of writing as the problem?
Your central argument against Return on Equity:
It is meaningless for multitudes of reasons, buybacks can flush equity fast when the stock is selling well above book for any number of reason, logical or not; write-downs of assets too can run up returns on equity.
You also mentioned dividends disturbing the ROE. If you are referring to the Return on Equity for firms paying dividends, their earnings are not marked down when dividends are paid, so the ROE figure should not be effected by the dividend. If you meant, however, the ROE on firms like Berkshire Hathaway that owns dividend paying companies, then the Berkshire's ROE also doesn't change substantially whether or not dividends are paid by the public subsiduaries - if s dividend is paid, there will be additional earnings from the dividend, but if it is not paid then there is a higher capital gain, and as Berkshire uses GAAP reporting (where the gains and losses are marked to the earnings) the earnings are also marked up by the higher capital gain. Either way, the ROE is about the same.
Your other criticism of ROE was the wite-downs. Write-downs indeed lower the equity, but they also mark the earnings down. Let's say you pay $50,000 for some asset and find it to be worth $30,000 later, then that -$20,000 should be represented as a loss, and your ROE that year should be reduced. Indeed the equity is also marked down, but it is the larger denominator so doesn't change the ROE much. For example, if your equity is $1 million and your earnings $100,000 then your ROE is 10%, but with the $20,000 write down your earnings are $80,000 and your equity $980,000, so your ROE is now 8%. The write-down has lowered your ROE from 10% to 8% (despite the lower equity), as it intuitively should in the economic sense. We are worse off for the write down, and our ROE gets punished correspondingly.
Sure, ROE will be a little higher later from the reduced equity if earrings are maintained, but deservedly so if it can keep on earning from the lower capital base.
Mungofitch already discussed the effect of buybacks upon ROE as being benign, but I would go further. Buybacks simply do not mark the earnings down, so are practically irrelevant to the ROE. If a company earns $100 million and the company buys back $30 million of their stock, the reported earnings doesn't change to $70 million but stays at $100 million. The cash flow is lowered by the buy-back, but we aren't looking at the cashflow/equity ratio.
Mungofitch's backtest of high/low ROE and high/low ROA shows their efficacy in the past. You didn't even acknowledge the backtest as contradicting your post's conclusion (stating ROE almost meaningless).
To the extent that reported earnings are representative of a typical year, the ROE is not meaningless. It is very much what business owners think when either buying a business, or expanding a business. When you start a venture, or otherwise lay down cash to make purchases, you want the best return upon the cash laid out. Whether or not your capital is scarce, what matters is the income produced relative to capital deployed. If you see an Return on Equity of 50%, you will throw as much equity in that direction as will allow.
Interestingly, you did not mention what I think is the obvious vulnerability with ROE, which is that earnings fluctuate from one year to the next. A better metric might be the five-year average inflation-adjusted earnings divided by this year's equity. I did some backtests with this, however, and didn't get better results than the straight forward ROE. The other problem is that you can elevate earnings just by borrowing more, so either use Return on Assets instead, or just make extremely sure the the debt is both sustainable and historically typical.
Writing well and investing well are two completely different skills.
This is straw-man argument as no-one has claimed good writing and good investing as the same skill. If someone (not mungofitch) has poor investment results, but presents good arguments and ideas, then I recommend to focus on critiquing their ideas rather than changing the subject to questioning their investment results which has no relevance to the arguments in the post.
- Manlobbi
