People don't want to publish the Aussie paper because it has missed the main point of LRZ. LRZ's paper is about persistence, not convergence. As for convergence, Chang and Dasgupta already show that mechanical mean reversion is a problem. Then what is the incremental contribution of the Aussie paper?
Capital Structure Convergence = BS
-
From LRZ's SSRN version: "The first notable feature in the figure is that leverage ratios exhibit a significant amount of convergence over time; firms with relatively high (low) leverage tend to move toward more moderate levels of leverage." Sounds like good enough reason to publish the Aussie paper.
-
^Yes, Chang and Dasgupta reinvented the wheel which can be found in an elementary Wikipedia entry. It looks like in sum, LRZ showed capital structures don't move much and converge slowly. Second point is mechanical, the other simply a descriptive stat of the data. If persistence is the main point -- then, most levels of financial market variables are persistent, volume, level of the Dow, cash balances etc etc. So, what is the grand insight?? This thing gets the Brattle? Well, it confirms that the profession is going to shit...
-
as dcc0 suggests all three publish insider shit from time to time...but this seems to reflect particularly badly on jf because the award winners are board members or have duke connections, and it looks like there were failures at multiple levels. personally, it seems rfs suffers the least from these problems...
-
Look some degree of bias/familiarity is almost unavoidable. One can argue that we have to live in a second best world. The problem with JF is that there are failures at multiple levels and they are repeated. Their award game looks like a complete inside game with some exceptions every now and then. The paper of this thread is a good example, but there are several others that fit the bill. Unless they change the hiding behind the Associate Editors model, it will continue to happen unfortunately for the profession.
-
This thread is seriously off point.
LRZ is of interest due to contrary claims in the literature. To be specific, according to Welch in the JPE ("Columbus's Egg" in the original version), equity follows a random walk. Debt is erratic and unrelated. So there is no force inducing reversion of leverage. To a first approximation leverage would be like a random walk plus a constant. If this claim were true, then over long periods of time leverage would not be stationary. But empirically over long periods of time leverage is stationary. LRZ is interesting because they show that the Welch claim, published in the JPE, is not the main driving force.
-
thanks a986 for this grand insight. but, have you even read the thread deeply? unfortunately, like dc70 says, lrz doesn't say jackshit about "reversion in leverage" or lack of it. the convergence result is not a result. so, your diction is better than your economics. it seems that in the desperation to show up ivo welch, people at all levels failed to catch even elementary flaws.
-
No, LRZ covers very old ground. Ground that has zero economic meaning. In fact, the reversion would seem to confirm the random walk hypothesis.
But don't take my word for it: go run the R code earlier in the thread. Create some white noise and get similar results. Then, ask yourself how you will be brushing up on your undergrad/masters level stats. **** me that my profession is full of idiots.
-
I have never run the R code (I hate R and all languages that enclose conditional statements in curly brackets), but I have written my own Matlab code that does the same thing. I'm not saying that LRZ is a good paper. Far from it. I'm just scraping around to try to give them credit for something...anything...at all. It's the "back to the beginning" part of the paper that is new ground, not the stupid convergence stuff. And I agree with you that the paper is utterly devoid of economic content. And as to your "**** me that my profession is full of idiots" comment, I could not agree more.
-
Matlab/Octave code? Excellent. Very well, carry on.
While I laud your attempt to be generous to the paper, I think you shouldn't have to stretch so much for a paper in JF, much less one that wins the Brattle Prize.
And yes, sadly, our profession is full of idiots. I know many people who couldn't organize their thoughts enough to write such code. The fact that you coded something up in Matlab that was more than a basic regression probably puts you in the top quartile of skill.