Fascinating little work to be published in the Quarterly Journal on Economics.
http://tuvalu.santafe.edu/~snaidu/papers/coups.pdf
QuoteWe estimate the impact of coups and top-secret coup authorizations on asset prices of partially nationalized multinational companies that stood to benefit from US-backed coups. Stock returns of highly exposed firms reacted to coup authorizations classified as top-secret. The average cumulative abnormal return to a coup authorization was 9% over 4 days for a fully nationalized company, rising to more than 13% over sixteen days. Pre-coup authorizations accounted for a larger share of stock price increases than the actual coup events themselves.
However, we show that firms benefit not only from publicly announced events but also from top-secret events, suggesting information flows from covert operations into markets.
Executive tearsheet:
QuoteIn other words, insiders who knew about CIA coups in Iran, Guatemala, Congo, Cuba, and Chile were buying stock beforehand in United Fruit, Anglo-Iranian Oil and other multinationals that would benefit. Or they were leaking the information to other investors.
:lol:
Great paper. Great avatar too.
CDM - how did this come to your attention?
Sample size is an obvious problem here although the authors appear to have taken steps to address it.
Perhaps a bigger problems is that the coup decisions don't arise from thin air - there are usually preceded by publicly available events or information that take the country in question to the brink and make the coup feasible. The control used for this is simply the quantity of NYT stories, which is pretty poor, both because it doesn't involve qualitative analysis and because there are a lot of signs or other facts suggestive of potential coup activity that a well-informed observer (like a multinational company on the ground) could notice that might not make the NYT.
Quote from: The Minsky Moment on April 28, 2011, 11:34:06 AM
CDM - how did this come to your attention?
Linkypoo in the Baltimore Sun finance blogs.
I mean, yeah...to the cynical, it can proably be filed under "D" for "Duh", but I thought the statistical analysis angle was an interesting one.
Quote from: The Minsky Moment on April 28, 2011, 11:34:06 AMand because there are a lot of signs or other facts suggestive of potential coup activity that a well-informed observer (like a multinational company on the ground) could notice that might not make the NYT.
This is quite true... but the problem will always be to find a way to investigate such an argument. These "other facts" rarely make it into official corporate documents - for the very reason that they would also be susceptible to leakages in all sorts of other venues. So, either we abdicate such an inquiry to future historians (and it wouldn't be much easier in any case), or we do a more traditional journalistic approach, which takes incredible amounts of times and effort - usually for results easily brushed aside as partisan (especially in such a topic) or limited to one case. I guess one could correct this by looking at the specialist press (and even then...).
Four of the coups occurred on or before 1961. That year is significant because it also happens to be the year when the SEC brought the first major enforcement action in an insider trading case. That is, prior to that time, trading on insider information was basically unregulated.
The fifth coup is Chile, and interestingly, the study reports little effect in terms of CAR from that coup.
So my guess is that the study is capturing exploiting of private information by corporate insiders. I don't know a lot about how CIA used to operate in those countries in the 50s and 60s but my wild guess would be that to a certain extent they cooperated with friendly US-based multinationals who had people on the ground with key local contacts and knowledge. So even without assuming direct access to top secret classified memos, the execs of these multinationals might have been getting good information about coup prospects.
Of course. My point was not that there might not be other possible, plausible explanations, but any such analysis needs to be anchored in a number of grounds. Projections of models of corporate governance in Latin America over these precise cases can be one. The statistical analysis was another.