5 July, 2007
By Dee Smith
IT’S déjà vu all over again: this week, another hedge fund manager pleaded guilty to fraud. The activities of Keith Gilabert, the founder of a fund called GLT Venture Fund, cost investors about $14m (£7.8bn, E11.3bn). This was just a skirmish, however, compared to the spectacular collapse last August of Bayou Capital Management, which took with it $350m in investors’ funds. In both cases, the losses were not due to anything like high-risk derivative investment structures: the cause was old-fashioned, outright fraud.
In breaking the Bayou story, the Wall Street Journal found lawsuits dating back several years indicating a pattern of bad-faith dealing. In the case of GLT, investors continued to put money into the fund throughout 2004 without knowing that the parent company’s investment adviser registration had been revoked the previous year.
In both cases, the relevant data were available long before the funds collapsed. For those who knew where to look and how to analyse what they found, it was actually easy to avoid a loss in Bayou or GLT by never investing in the first place.
The truth is ‘out there’ in many investment situations. Risk can be mitigated by detecting and avoiding people and entities with discoverable negativities. Since one truly bad investment resulting in a total loss of capital can torpedo the cumulative return of many good investments, simply steering clear of such situations whenever possible can dramatically reduce overall portfolio risk and enhance return.
This can be done by exploiting the enormously expanding global information environment. Total information produced worldwide is increasing by about 30% per year: i.e., doubling every three years. Estimates vary, but a best guess is that somewhere over 90% of all information is now accessible somewhere through open source or restricted source data. ‘Open source’ is the technical term used in the intelligence community to denote data available without special access.
The inherent practical problem is that such data are not situated in one place, and thus the answer to a key business question like “am I dealing with someone likely to defraud or misrepresent” (or, for that matter, “what are the new product strategies of my biggest competitor”) cannot simply be Googled. Such findings must be assembled analytically from many pieces of information gathered from many sources: some obvious, some obscure. Increasingly, the requisite pieces are simply not available on the ‘open’ internet, but are housed in the hundreds of proprietary subscription databases that capture and store full text data from tens of thousands of sources. Such data are ultimately legally accessible if you know where to find them and are willing to pay.
There is a set of protocols and approaches for collecting and analysing such data to produce actionable findings, and it has become known by an interrelated family of terms including competitive intelligence and investment intelligence. These techniques are closely related to methods long used by government intelligence agencies to derive meaningful findings from masses of data. This ‘intelligence approach’, which is only productive with large arrays of data, has become relevant to the business world precisely because of the explosion of information (the private sector now has access to a level of data that not even national intelligence agencies possessed in the past).
How does it work? Imagine the following: a chief financial officer makes a statement about her company while on a panel at a conference in London on a Monday (which is recorded and transcribed), the chairman of the board says something to the press during an interview in New York on Wednesday (which is published), and the chief executive comments on a radio interview in San Francisco on Friday (which is transcribed). Even though none of these individuals has said much if taken in isolation, when you put the particular pieces together, they reveal far more about a situation of interest to you than was ever intended. The key, of course, is knowing where and how to find those specific pieces, and then how to analyse them when they have been collected.
In the specific world of investment intelligence, this kind of discovery from media sources is only one piece of the puzzle. One also deals with data relating to litigation, corporate affiliations and structures, licensing, financial filings, judgments, liens, bankruptcies, regulatory violations, links and associations (business, personal, and political), philanthropic involvement, possible criminal and money-laundering connections, and even address histories. All of these elements, when combined, can reveal much about the nature of the subject of investigation’s life and activities.
While any one piece of data may not seem important in and of itself, the overall pattern of data can provide a striking level of insight. And since the data array is filled with errors and inaccuracies, it is important not to rely on any one piece of information, or even on one type of data. This is where techniques that allow data corroboration, reverse-engineering, and identification of anomalies become very important.
At its best, the intelligence approach reveals hidden information by collecting data on the subject under study, patterning these data, creating hypotheses based on the patterns discovered, and then verifying or falsifying the hypotheses based on additional collection and analysis. An iterative cyclical process called ‘the intelligence cycle’ is often used, which typically includes project/goal definition, planning, collection, analysis, production (of findings), and dissemination to the client. The goal is simple: to provide insight that helps the client make better decisions.
Increasingly, a ‘disclosed’ investment intelligence modality is coming into use, and is even being required by investors as a condition of investment. This means that the subject of investigation knows he or she is being investigated and has signed and completed a disclosure/release form, providing legal access certain restricted data (such as credit reporting data), and also provides assertions that can be verified or falsified (education, work history, etc.).
Beyond this, monitoring through repeated investigations on a regular or continuing basis, called ‘indications and warnings’, represents a way to extend risk management.
One of the newest applications of these techniques, driven by anti-money laundering concerns and regulations, is intelligence on investors prior to accepting an investment.
Legality is a critical concern: for the client’s protection, all investigations should operate fully within the law and be conducted by a firm licensed in its home jurisdiction to undertake such operations. The difficulty of conducting investment intelligence varies widely, based on how ‘friendly’ the information environment within a particular country is. The US is easiest, followed generally by western Europe. In other areas, difficulty increases as access to data becomes more problematic. Cost will naturally rise with difficulty. However, because the global data array is now so rich, investment intelligence can be revealing even in the most difficult environments.
The applicability of the military intelligence maxim ‘trust but verify’ to the investment world cannot be overstated.
From Spectator Magazine, Business – 5th July 2007, by Dee Smith