Daily Speculations

The Web Site of Victor Niederhoffer & Laurel Kenner

Dedicated to the scientific method, free markets, deflating ballyhoo, creating value, and laughter;  a forum for us to use our meager abilities to make the world of specinvestments a better place.



Write to us at: (address is not clickable)

James Sogi on Corporate Corruption

Acknowledging the List writings of GZ and Tom Ryan, GZ recommended the Way of Go. Troy Anderson, author, suggests a method of analysis called "Reverse Forward" in which analysis of a situation does not progress along strictly temporal lines, rather uses a method called 'Tewari' which takes steps out of the usual order. Move backward to make forward progress or use reverse psychology. Solve a maze backward from the conclusion to find the best current move. Tom suggests 'turning the problem on its head".

You can't keep a good man down, and by the same maxim, Vic theorizes that you can't keep a good company down. In rooting out corruption that drives a good company's price artificially down, when the bad guy/news is ousted, the price bounces back. Try find good companies with minor corruption problems, buy when price is down, sell later for profit. The uncertainty arises in what the extent of the corruption is and whether it will be fatal. Here is where the sex life of accountants comes in. The accountants do the SEC 10K', 10Q ,8K's available on Edgar and Lexis/Nexis. Here, in theory, the dirt is exposed and , importantly, the amount of the dirt. Last year, when I first joined the list, back when I was bearish, I studied companies that delayed their 10Q's which materially affected their finances looking for nice short candidates hiding some bad news at a time when the the dirt is first exposed. When problems surface, the health of the patient is not known.

What are the factors. Take the search out of order, tewari style. When researching the law, one tewari technique is not to ask at first what is a winning argument or a winning concept, rather, find the winning case and look for the reasons the winning case used. Working backwards, out of order. The same might work for financial research. Resignations of the CFO or accountants appear in SEC Form 8k, . Examine financial. But to figure what to look for, go to the ones that had problems, and then work backwards to find the reasons, then reverse to expand to the full population, Tewari style.

For example, here is an excerpt from Enron's 8k announcing Lay's resignation. Even their own internal committee doesn't know what is going on.

"The Special Investigative Committee of the Board of Directors of Enron Corp. submits this Report of Investigation to the Board of Directors. In accordance with our mandate, the Report addresses transactions between Enron and investment partnerships created and managed by Andrew S. Fastow, Enron's former Executive Vice President and Chief Financial Officer, and by other Enron employees who worked with Fastow. The Committee has done its best, given the available time and resources, to conduct a careful and impartial investigation. We have prepared a Report that explains the substance of the most significant transactions and highlights their most important accounting, corporate governance, management oversight, and public disclosure issues. An exhaustive investigation of these related-party transactions would require time and resources beyond those available to the Committee.There were some practical limitations on the information available to the Committee in preparing this Report. We had no power to compel third parties to submit to interviews, produce documents, or otherwise provide information." And the dirt was a tsunami.

I used to think accounting was exact, but learned rapidly it is more like writing fiction. With modern derivatives, insurance, over the counter, consolidated accounting, partnerships, one can structure anything in the world, and like Fastow and Ebbers, FNM, hide billions.

The key is to work backwards to tailor the search query . Start with bad companies' links in Edgar or lexis/nexis . 8k 'other events' might be a good start announcing resignations. See what the key words are.

The 10K's list aggregate market value. Here is Enrons, " Aggregate market value of the voting stock held by non-affiliates of the registrant, based on closing prices in the daily composite list for transactions on the New York Stock Exchange on February 15, 2001, was approximately $60,207,479,342. That might be one screen, over x market value.

Then screen for 8k's announcing resignations or investigations generically known as "other events". Let's say we get a list of companies over x value, with "resignations" or 'restated financial statements" .or "SEC investigations". Then we can use quantitative tests to see if there is some useful correlation in the search terms with subsequent results. 233 companies in 2000 restated past years financial statements. That is a start. This process will be facilitated by Adoption of XBRL financial data standards. But til then, Lexis /Nexis or Edgar subscribers might have a actionable edge since the public filings are not easily text searched. * "*Aug 2001 Skilling resigns as CEO; Lay resumes position" This resignation would be announced and might have picked up the issue before it happened.

Some more sources for 8k's provides a start:
Source 1
Source 2

Alppublishing.com has the following:

Form 8-K is used by companies to file current reports on the following events:

Item 1. Changes in Control of Registrant. Item 2. Acquisition or Disposition of Assets. Item 3. Bankruptcy or Receivership. Item 4. Changes in Registrant's Certifying Accountant. Item 5. Other Materially Important Events. Item 6. Resignations of Registrant's Directors. Item 7. Financial Statements and Exhibits. Item 8. Change in Fiscal Year. Item 9. Regulation S Offerings.

The following companies have filed 8-K reports for the date indicated and/or amendments to 8-K reports previously filed, responding to the item(s) of the form specified.

We would look for codes 6,4, taking the list from the above source, and scanning with R and pull out results in the companies with 8ks announcing codes 4 or 6, maybe 3 & 5.. I don't have time right now to figure out how to do that in R, but will work on this start when I figure the R code to do the Wiz thing and keep the spaces in the table.