I'm basically an Ayn Rand cultist, and I think the article delivers. It is now obvious that the CEOs running Lehmans, Bear, and Citibank in 2007 had absolutely no idea of what their assets were or how they were funded. They were as far outside their "risk limits" and "authorized positions" as any of these "rogues". But their limits were phrased with far less clarity and with far more wiggle room. The criminality of these "rogues" is defined by their failure to comply with clear and easily adjudicated policies. No such criminality could be applied to the CEO's deviation from their duties. The difference between the CEO and the rogue trader isn't in their basic behavior but in the standards that apply to them.
That problem cannot be solved by law or policy governing behavior. The standards applied to these CEOs do not admit of the precise specification needed for clear differentiation between error in judgment, bad luck, malfeasance or dereliction of duty. Hofstadter talks about the "computability" problems addressed by Godel and such, trying to determine which problems are and aren't resolvable by computation. Governance at that level is a similar kind of problem.
The implications of this are profound. Since we can't legislate / regulate right behavior, we have to motivate it through incentive alignment. Thirty years ago these banks were partnerships -- if they failed, they wiped out most of the saved capital of their retired partners and senior employees. You can be very sure those folks were both willing and able to apply the qualitative judgments of risk unavailable to a policy driven review.
Those partnerships were converted to publicly held equity because this was less risky for those partners and allowed the companies to achieve larger scale, momentarily conveying cost advantages. But public equity holders don't, and can't, understand these entities as well as those partners did, and therefore cannot discipline management nearly so well. The real long-term costs of that dilution in risk management are now more apparent.
None of this would be any of my business if banking weren't a necessary component of modern money creation and thus inherently intertwined with the government. As such, any citizen has a stake and a voice in the stability of these institutions and their role in the creation of money. The latest regulatory revisions seek to obtain the needed stability by improving the foresight of the regulation, backstopping any failure with an implicit government guarantee on the system. Since the regulation _cannot_ be improved, we are headed for another bailout at some moment to be determined. Meanwhile we are essentially underwriting excess compensation for those managers and traders who can figure out the extraction of cash from the system before the equity again falls to zero.
There are alternatives. Each of these institutions should be a lot smaller, so the system can tolerate the collapse of any several. And they should be generally capitalized by investors who can understand their assets and liabilities, who are in a position to get the confidential information needed to the institution's specific balance sheets, and who _cannot_ hedge or diversify away enough risk to become indifferent to the institution's success. Publicly traded equity fails on all of those conditions.
The consequences of partnership equity are obvious. It rolls back the scale, and with it the momentary cost advantages, enabled by larger, dumber, indifferent public capital. Those large institutions managed with integrity (e.g., JP Morgan) will be f*ed. Capital costs will rise, in part due to induced scale and liquidity inefficiencies, but in part due to the reflection of the real risks of these institutions. But the policy problems of moral hazard and public subsidy of private speculation will be removed. We will be spared the logical contortions required to explain away the obvious immorality and idiocy of many of the people running these institutions. We will avoid the corruption of language and thought that necessarily follows from attempting the impossible task of policing an unspecifiable morality. Most important, we will remove dangerous political corrosion that follows from associating that intellectual corruption with such extraordinary potential for private gain.
tldr; "Rogues" differ from CEOs only in the ability to specify position limit violations. The inability to specify CEO position limits implies that financial stability is unavailable through regulation. Stability thus only available through alignment of management and equity incentives and knowledge. The only equity with proper incentive and knowledge would be partnership equity.
tldr(tldr;) CEOs differ from rogues only in the impossibility of specifying their "position limits", so CEOs can only be managed by equityholders as knowledgeable and focused as they are -- e.g., partners, not public equity holders.