Koss Corporation has had a horrendous couple of years: a $30 million accounting embezzlement; its former Principal Accounting Officer sentenced to 11 years; and a related SEC Complaint in October alleging that the Company prepared materially inaccurate financial statements and lacked adequate internal accounting controls.
Things brightened a bit when Koss, and its co-defendant, CEO Michael Koss, agreed to settle with the SEC. They filed the settlement with federal Judge Randa in Wisconsin, no doubt hoping for swift court approval so the Company can get on with the business of selling headphones. That didn’t happen.
Not so fast, says Judge Randa.
Judge Randa is apparently a fan of the jurisprudence of Judge Rakoff, the federal judge who recently stunned many observers by rejecting the SEC’s proposed $285 million settlement with Citigroup and ordering the parties to trial. Recall that Judge Rakoff refused to approve the settlement “because the Court has not been provided with any proven or admitted facts upon which to exercise even a modest degree of independent judgment.” This was somewhat surprising because ordinarily one does not expect to find “proven or admitted facts” in a settlement. A settlement is an agreement to resolve disputed claims.
Anyway, like Rakoff, Judge Randa of Wisconsin has some troublesome questions for the SEC about the Koss settlement. He raised them by letter on Tuesday, virtually assuring that Michael Koss will have neither a Merry Christmas, nor a Happy New Year.
But before we take a look at the Randa letter, did you know that Koss’s computer accounting system was thirty years old? We learn this from the SEC complaint, which argues (rather persuasively) that Koss lacked “adequate internal controls” over accounting. Michael Koss, says the SEC, “knew that Koss’s computerized accounting system was almost 30 years old and he twice deferred proposals for a new system.”
Only twice? The SEC is being generous here. The company, which failed to notice $30 million missing, apparently looked on with disinterest as the tech world rolled out the 286, the 386, the 486, the Pentium, the Pentium Pro, the Pentium II, III, IV, and V, the Celeron, the Core 2, all the Apple equivalents, countless enterprises systems, and 12 versions of Microsoft Excel. Each new system apparently received a shrewd thumbs down from a cost-conscious management.
So Judge Randa has the task of evaluating Koss’s proposed settlement with the SEC, which requires (among other things) that Michael Koss disgorge his 2008, 2009 and 2010 incentive bonuses, and requires (by an injunction) that the company fix its inadequate accounting controls.
So what’s the problem? Well, like Rakoff, Judge Randa wants more facts, or as he puts it, a “written factual predicate:”
The Court requests that the SEC provide a written factual predicate for why it believes the Court should find that the proposed final judgments are fair, reasonable, adequate, and in the public interest.
The judge explains that he is troubled, for example, by the injunction against future violations because the order to maintain an adequate control system contains no specifics – no time frame for the improvements, no mechanism for implementing the changes, and no provisions regarding reporting, oversight or auditing of the agreed changes. This lack of detail, says Randa, would make it hard for the court to enforce the injunction if necessary. This seems like a fair criticism. One imagines that the lack of specifics is probably because Koss has already overhauled its accounting controls – a $30 million theft tends to put a pep in one’s step. The judge’s criticism can likely be addressed by simply inserting more specifics on how Koss will improve its controls, the timetable for improvements, and mechanisms for reviewing the changes to ensure compliance. Such detail in a regulatory injunction would not be all that unusual.
The bigger problem for Koss is the second thing troubling Judge Randa – the amount to be disgorged by CEO Michael Koss:
[W]ithout any factual predicate for how those disgorgement terms were determined and what more, if anything, could have been subject to disgorgement, the Court cannot assess their fairness and the extent to which they serve the purpose of disgorgement which is to deprive the violator of unjust enrichment and thereby further the deterrence objectives of the securities laws.
Here we may be crossing into Rakoff country. Ideally, in assessing the fairness of a disgorgement, the court would have before it a set of proven facts. Among those facts would be a dollar value representing the amount by which Michael Koss was unjustly enriched by his prohibited conduct. This dollar value would, of course, be a good place to start in ordering disgorgement.
But we don’t have proven facts here. Michael Koss disputes the SEC’s claims. That is, he denies liability for an embezzlement that was actively concealed from the company by its Principal Accounting Officer and its Senior Accountant. In this context, the disgorgement value is based on what the SEC thinks it could likely prove at trial. Some or all of the SEC’s factual allegations may be difficult to prove. Some or all of its legal theories may rest on shaky grounds. So the SEC must estimate the probability it will win at trial, then use that estimate to discount the amount it would demand at trial and arrive at a ball-park settlement figure.
Other factors play a role as well, not the least of which is the SEC’s scarce trial attorney resources. The SEC must be selective in the cases it brings to trial to ensure maximum impact and deterrence. Taking the Koss case to the hoop, so to speak, may not be the best allocation of scarce resources.
So what exactly does Judge Randa want? If he is looking, like Rakoff seems to have been, for “admitted or proven” facts supporting liability, this may be a dead-end street. If the SEC aggressively negotiated the settlement, the disgorgement amount was pushing right up against the maximum amount Michael Koss is willing to pay. Any significant amount more and presumably he’d rather take his chances at trial. A requirement to admit facts, in effect, demands more from Michael Koss. This is because admitted facts can often be used by plaintiffs to prove liability in private cases. So if the deal was aggressively negotiated in the first place, we shouldn’t expect Michael Koss to be admitting any important facts. Note that the “neither admit, nor deny” settlement format used by the SEC can actually increase the settlement amount paid to the SEC because the format allows the SEC to capture some of the value the defendant gets in retaining the ability to defend private lawsuits.
But there is another way to interpret the Judge’s request. Perhaps, in asking for a “factual predicate,” the Judge is really looking for more detail on why the SEC believes the disgorgement amount is appropriate, given the SEC’s view of its likelihood of success at trial. This would involve the SEC explaining not only the amounts it believes could be subject to disgorgement, but also the strengths and weaknesses of its case. This view comes with its own problems, however. For example, the SEC may be reluctant to provide a detailed account of the weaknesses of its case. Such an account would be interesting reading material for Michael Koss if the settlement were scuttled and a trial ordered, as happened in the Citigroup case.
The SEC has until January 24th to respond to the Judge’s concerns. We can expect to see the settlement modified to add specificity to the injunction.
We may also see some attempt by the SEC to explain its thinking on the settlement (including perhaps some acknowledgement of weaknesses in the case). It seems unlikely, however, that the SEC will deliver a set of facts admitted by Michael Koss. At least not any important facts. Perhaps he will admit that the computer accounting system had roughly the same oomph as the Atari 5200, released during Ronald Regan’s second year in office.
UPDATE: The SEC filed a response to Judge Randa on January 24, 2011. Judge Randa accepted the SEC’s explanations for the settlement and he approved it. As anticipated above, Koss had already undertaken the appropriate steps to remedy its inadequate controls. For example, the SEC noted that Koss had already: terminated all responsible accounting employees; dismissed its public accounting firm; hired a new CFO, Controller and Credit Manager; changed its internal banking procedures; hired new accounting personnel; segregated accounting functions; engaged a consulting firm to improve internal controls; limited the officers who could act as signatories; and many other things. On the intriguing enigma of the 30-year old computer system, however, the SEC notes only that “Koss is also integrating a new computer system into their business to remediate the account, access and monitoring internal control issues.” Perhaps the delay in integrating the new system arises from having to transfer 30 years of accounting data by floppy disk.
Also as anticipated above, the SEC elected to provide some acknowledgment of the weakness in its case:
If this case were to proceed to trial, the Commission would likely also seek to require MJK to reimburse Koss for $1 million in bonuses received by MJK after Koss filed its financial statements for the fiscal years ended June 30, 2005, June 30, 2006 and June 30, 2007, as well as additional incentive-based options received during that period. However, the company did not formally restate its financial statements for those periods. Thus, MJK might argue at trial that restatements were not “required” for those years within the meaning of Section 304(a)[of Sarbanes-Oxley] . . . . The Commission disagrees with such an interpretation of Section 304(a) and submits that it would not be supported by a plain reading of the statute. Nevertheless, the Commission considered the possibility that a court might find otherwise in determining to forego reimbursement for the earlier years in the instant settlement.
Given the current judicial climate, we can expect to see this type of frank explanation for settling made more often. It simply makes public some of the reasons for agreeing to a particular settlement that were previously kept internal to avoid: (1) tipping off the defendant to factual or legal weaknesses in the SEC’s case, and (2) public disclosure of the SEC’s method of allocating scarce regulatory and trial resources. These are, of course, legitimate concerns for the SEC. No doubt a balancing of these interests with the judicial desire to provide meaningful review is in order.