I am fascinated by attempts to replace Tobin’s Q as a firm performance metric with Total Q. Both metrics are not fit for purpose. One of the most interesting things is just how terrible the arguments in favor of Total Q as a firm performance metric are. I recently highlighted problems in Du and Osmonbekov‘s “Direct effect of advertising spending on firm value: Moderating role of financial analyst coverage”, see here. I thought it was a Total Q mystery how the paper got published in a usually good journal. Here is another Total Q mystery.
A Slightly Different Mystery
Leung and Sharma’s paper looks at the connection between R&D, innovation performance, and “firm performance”. To be honest the R&D analysis isn’t really my thing. As such, I don’t have much to say about the body of the paper. I’ll just concentrate on their measures of firm performance.
They explicitly follow Du and Osmonbehkov, citing these authors that Total Q is better than Tobin’s Q. (I agree Total Q is probably better — but not by much).
Leung and Sharma seemed to have found Du and Osmonbehkov convincing which suggests that they didn’t read the earlier paper properly. I would say that the tack they take is different. Du and Osmonbekov gave the reader a reasonable amount of detail — and thus provided enough rope to hang themselves with. Leung and Sharma seem to work on the principle of, ‘least said, soonest mended’. Their justification for their choice of two long-term financial performance metrics is a total of a mere 82 words. This total includes three citations (two of Du and Osmonbekov alone).
I am pushing for a new heuristic when reviewing a paper. If you can read the authors’ justification of their key dependent variable in a single breath then maybe sufficient detail on their thinking is lacking.
Two Metrics Are Better Than One?
The authors use both Tobin’s Q and Total Q. This is a mini-mystery. Basically, they justify using Total Q because Tobin’s Q doesn’t do the job well. So why keep Tobin’s Q in the paper? It seems to be designed to convince the reader that because two metrics work it must be right. But one of the metrics is acknowledged by the authors in the article to not be good enough.
What is more, they show the correlation between Tobin’s Q and Total Q. This is massive (0.63) because they are very similar metrics. Total Q is just a (very slightly) improved version of Tobin’s Q. They are highly correlated because they both have the same core.
Another Total Q Mystery
Despite not giving us much detail on the metrics the authors do reveal enough to show they are wrong. It is so easy to prove them wrong it seems almost cruel. But I feel like I have to do it. I’ve heard people defending Total Q and I’m determined to try and stop this nonsense before it infects and undermines the discipline as Tobin’s Q did. Tobin’s Q was biased to give publishable results and so a popular choice with academics. Maybe Total Q has a little less bias but has the same fundamental weaknesses. The changes between Tobin’s and Total Q simply don’t make the latter okay as a financial performance measure.
So how can I show that Leung and Sharma are wrong? The detective work isn’t hard.
Tobin’s Q, the ratio of market value to book value of tangible assets (Bebchuk, Cremers, & Peyer, 2011).
Leung and Sharma, 2021, page 84
I have discussed at length that this statement is wrong. Indeed, our Tobin’s Q paper had a section on this urban myth (see table A.2, here). Having read their comment it occurred to me to check their citation. Perhaps, Bebchuk, Cremers and Peyer had used an approach that I wasn’t aware of that actually excluded intangible assets. Or maybe Bebchuk and colleagues made an incorrect statement and have led Leung and Sharma astray. Nope. To be fair to Bebchuk and friends the mistake was entirely Leung and Sharma’s. Bebchuk was mis-cited as saying “book value of tangible assets” but had really said, “book value of assets” (page 212).
Here Is More Evidence
Bebchuk and colleagues made a large number of statements about Tobin’s Q. For example see also page 202, footnote 6, page 203, page 207 etc…. Leung and Sharma must have misread them all. Bebchuk even gives the Compustat item number so the reader can check if there was any confusion. Item 6 (the denominator of Tobin’s Q) Total Assets. Presumably, Leung and Sharma didn’t bother to check this.
| AT | 1 | 6 | iaitems | Assets – Total |
My point is that Total Assets and Tangible Assets are not the same things. You don’t need to be an accountant to guess that Total Assets includes all assets, both tangible and intangible. That is the reason why it is called total — because it includes all of the assets including the intangibles.
If you don’t believe me here is Bebchuk clearly stating what they mean by Tobin’s Q (see “data6 is the book value of total assets”).
Of course, financial accountants do tend to omit and understate intangible assets but that doesn’t mean intangible assets don’t exist in financial accounting. I’m pretty sure the reason why we have the tangible/intangible asset distinction is that accountants found it useful. This is because they do sometimes record intangibles. (This is ridiculously easy to prove, just look at INTAN, item 33, in Compustat data).
With Advocates Like These
Given lots of econometric manipulation, it can sometimes be hard to be clear what bias exactly remains in any analysis. The good news is that supporters of Tobin’s Q and Total Q continue to write silly things that make it obvious that they don’t know (or maybe care) what they are doing.
Why do reviewers let authors get away with a perfunctory explanation of dependent variables despite dependent variables being what we generally should care most about? How do so many people think Tobin’s Q does not include intangible assets when every version I have seen used in marketing has included them? Why do an analysis on Tobin’s Q if you think it is clearly a flawed metric? How can you justify use of Total Q as solving the problems of Tobin’s Q by adding in intangibles when you misstate Tobin’s Q as not including any intangibles when it clearly does include some of them already? This paper is another Total Q mystery.
For more on Total Q see here, here and here.
Read: T.Y. Leung and Piyush Sharma (2021) Differences in the impact of R&D intensity and R&D internationalization on firm performance – Mediating role of innovation performance, Journal of Business Research 131 pages 81-91
