Update: Thanks everyone for the continuing thought-provoking discussion. I intend to post my decision spreadsheet, and still am looking for suggestions on where to do so. It might come in handy come February. A discussion that I find interesting has branched off on the topic of technological progress versus Malthusian Crunch, and I started a new article on that over here.
I would like to kick off a discussion about optimal strategies to prepare for the event that the US government fails to raise the debt ceiling before the US Treasury Department's "extraordinary measures" are exhausted, which is estimated to happen sometime between October 17th and mid-November.
This is a risk *caused* by politics, but my goal is to talk about bracing against the event itself if it happens, not the underlying politics. If you want to debate Obama-care, who is at fault, or how likely a US default actually is, please start a separate discussion.
I consider this to be an indirect existential risk because if it kicks off a national or global recession, it will likely slow or halt research and philanthropic efforts at mitigating longer-term existential risks.
Since there are obvious associations between unemployment/poverty and crime, civil unrest, and poor health, a global recession is likely to be to some extent a personal existential risk to those living in the United States or countries that have trade links with the United States.
I notice that the markets do not seem to be anticipating a bad outcome. But I heard one analyst advance the theory that investors simply don't believe the government can (his words) "be that stupid". I imagine there is more than a touch of availability bias as well-- breaching the debt ceiling might, even for fund managers who harbor no illusions about the wisdom of politicians, be up there with science-fictional scenarios like asteroid impact, peak oil, grey goo, global warming, and terrorist attacks. Moreover, there may be a dangerous feedback loop as the politicians in turn watch the stock indexes and conclude that "the market says there is nothing to worry about".
So, I would like to hear what folks who are making contingency plans are doing. Especially people who have training or experience in economics and finance. What do you think the closest parallels in 20th/21st century history are for what the worst case scenario for a US government default would be like? Is there anything you would have done differently if you had known the date for the start of the 2008 recession with a +/- 2 week confidence interval, starting in two days? Or, if you did call it ahead of time, what are you glad you did?
Okay, some (anecdotal) evidence now in and it looks to be in support of... both models?
The insider-driven model you propose would predict that there would be a negligible reaction of the market to announcements of a deal trickling in this morning because the insiders would already know how it would go and would have traded accordingly. The non-insider-driven model would predict a rally after the deal was announced.
What we are seeing is W5000 (Wilshire 5000 index) closing last night at 18141, opening today at 18254, and rallying at just before 10am, when the first hints of the impending deal got on the news, to 18322 and staying roughly steady after that. S&P and NASDAQ behave similarly. So from last nights close to when this morning's rally tapered off it gained 181 points. Of them, 113 were gained in after-hours trading, presumably by insiders. The other 68 were gained in trading during the 10 minutes after the news first broke. So, if we do a naive estimate, 68/181 = 38% of the movement it attributable to traders with access to the same information you and I have. So if I'm interpreting this correctly, and if this one datapoint is a representative one (two huge ifs) then insiders do drive stock prices, but the influence of non-insiders is not negligible. That in turn implies that it is not hopeless to turn a profit trading ETFs.
Again, all this is on one datapoint and should not be taken at all seriously, but this might be a framework for how to actually test the insider theory on a larger dataset.
Also, this may imply that the insiders (at least the ones whose actions are observable in this time interval) did not have more than about 18 or so hours notice.
If there is an even earlier group of insiders, the data (try zooming out) might be consistent with one or more of the following from there not being sharp rallies earlier:
There are too few of them to stand out over the noise of the non-insiders
They did not get their insider information at the same time and no individual one of them was large or greedy enough to trigger a rally.
They had automated orders that would be triggered only under certain conditions.