Cases in point. JNP has TTM net income of 5.95 million and common equity of 34 million but has a TTM ROE of -15.82%. By my calculation, its ROE should be 17.59%. KRO has TTM income of 43.3 mil and common equity of 452 million but an ROE of -0.06% instead of 9.58%. DCO is even more baffling: income of $25 million, common equity of $201 million, ROE of -17.87% instead of 12.56%. EXK: P123 says its TTM ROE is -79.81% but its income is 3.91 million and its equity 88 million, so its ROE should be 4.43%. CSRA: income of $195 million, equity of $166 million, so ROE should be 117%, but P123 says it’s 76.49%. Other stocks with huge discrepancies include PJT, CARB, SRI, KODK (which has negative equity), GPRK, TCMD, etc.
Please let me know if I’m missing something vital here. I just don’t understand how the numbers can be so different.
But in the meantime, in looking at the P123 definition of how ROE is derived, they say in help “This value is calculated as the Income Before Extraordinary Items for the period divided by the Average Common Equity and is expressed as a percentage. Average Common Equity is the average of the Common Equity at the beginning and the end of the period.”.
So I think they use IncBXorForComTTM and not NetIncBXorTTM (or IncBXorAdjCSETTM). This seems to give the big swings for the numerator for JNP and DCO from positive to negative.
Could this be an ‘adjusted earnings issue’?
I am puzzled by both IncBXorForComTTM and IncBXorAdjCSETTM because they are supposed to be in P123 on a per share basis and yet when looking at the numbers, they seem to be non-share numbers whereas NetIncBXorTTM is just the dollar figure. So not sure how P123 is calculating these but IncBXorForComTTM does have a negative amount for some of the companies you mention…
I cannot figure out KRO either though…that seems to use the NetInc from the Cash Flow statement or something. Not sure.
There are literally dozens of ways to measure the bottom-line on an accounting basis. They all attempt to get at something unique. However, no accounting metric solves the core problems of accounting.
For the bottom-line, I tend to rely on:
a. NetIncBXor()
b. NetIncCFStmt() = NetIncBXorNonC()
By the way: NetIncBXorNonC() - NetIncBXor() = “Minority Interest - income statement (mii)”
In both cases, I understand what items Compustat analysts are adding and subtracting. I can recreate them exactly for about 97% of all stocks with constituent data items in P123. In fact, the actual Compustat data item for net income is not even available in P123 syntax (most people probably don’t want to use it anyway, though, since extraordinary items can really distort things).
Maybe it’s just me, but if someone shows me a derivative metric and I don’t have a good idea of what’s exactly on the numerator or denominator, then it’s just noise.
I’m trying to apply DuPont analysis. If income available to common is used to calculate ROE but net income is used for profit margin, the results make no sense. And which of the two is used for GAAP earnings?
Yuval - Equity is the same thing as book value. Return on equity is the same thing as change in book value.
I screened JNP as of the date of your original post.
BookvalTTM was 33.84. BookValuePTM was 40.47. A decrease of 16.38%.
The typical reason a company can have positive income and a negative ROE? It pays out more in dividends than it earns. Thus it has negative retained earnings. In this case, retained earnings TTM was -247.69.
I’m afraid that’s not how one defines return on equity. One has to subtract preferred dividends, but not common dividends. And it’s not change in book value, it’s net income divided by book value. If P123 is using a different definition, I don’t really know what it is.