Manipulation of prediction markets

Nice post on manipulation of prediction markets: ...manipulation can improve (!) prediction markets - the reason is that manipulation offers informed investors a free lunch.

Nice also to see our old friend Hanson.


... the first casualty is causality

This week, like everyone, I had some thoughts about the $700B financial industry bailout. I put them in an email to some friends a few ago... Here they are (with some edits):

This bailout is looking more and more like a scam.   The fear-mongering about jobs sounds a lot like when the same folks were saying Saddam was about to nuke the USA in alliance with Al Qaeda. Just because it's scary, it's not automatically true, and it's not automatically the case that the proposed solution will avoid it. 

$25B for Detroit automakers slipped into the bailout? WTF? They would surely have been turned out if they had asked for it alone, but piggy-backed into the $700B... Recall how after Sept 11 all the airlines were begging for bailouts, and many of them got turned down. Imagine how many companies will try to jump onto this bailout like hungry vultures on fresh kill. 

Also, note that the relationship between Paulson and Goldman Sachs is eerily similar to the relationship between Cheney and Halliburton. Nothing against Paulson per se, I'm just pointing out the parallel. In any case, even if he's a saint, is any government capable of disbursing a freestanding $700B fund without massive corruption? Visions of scams with this bailout that'd make Halliburton's no-bid contracts for Iraq look like chump change....  In fact, this whole mass psychological phenomenon  looks more than a bit like an accelerated rerun recent history... With Fannie Mae-Freddie Mac-Lehman-AIG, in the role of Sept 11, and the Big Bad $700B as the Iraq Invasion of 2003,  the US congress as itself,  and special appearance by the Great Depression II as Saddam's WMD. I don't mean to compare the events of course, fundamentally different crises. But the the political and business dimensions sides are similar.

I'm also reminded of Russia in the early 1990s. The experts kept screaming that if this or that privatization didn't happen right away, the consequences would be severe, Russia would be come communist again, the 3rd world war would start immediately, etc etc. Guess what? Behind all the deadlines, and crises, rescues, etc., there were some dudes rigging privatization auctions. By the time people woke up to the game, seven gazillionaire oligarchs had ownership of more than 50% of the GDP of the ex Soviet Union, and the standard of living had dropped, life expectancy had dropped, the economy was in ruins... Basically the entire country got jacked by a few dozen people.

Back to USA 2008. The crisis  exists of course,  and ok let's agree something must be done [NOTE: this was written on Wednesday Oct 1, before the Trouble Asset Recovery Program (TARP)  bailout was approved by congress today].  But I'm skeptical about TARP.

So let's go to the root causes - not because I presume to know something more than the experts have been saying, but just to help me understand by summarizing. The crisis exists because:
1) we don't know how many more homes will foreclose, and what they will be worth when they do; which leads to
2) uncertainty in value of mortgage backed securities (MBS) which leads to
3) uncertainty about balance sheet of banks with lots of MBS, which leads to
4) unwillingness to lend i.e. credit crisis, which leads to loss of savings, loss of jobs, homelessness etc.

Now instead of injecting money at #3 which is what TARP (the current proposed bailout) does, why not inject it at #1? I think the government should just say it will buy the houses that are being foreclosed on rather than the securities derived from them. This seems better to me than TARP because:
a) the taxpayer is buying something whose valuation is better understood, and therefore taking less of a risk;
b) by having a buyer of last-resort for the houses and a known worst-case price, you're putting a floor under the value of these MBSes... this clears up the bank balance sheets, everyone will know which banks are solvent and not, and those that are can resume lending and that stops the credit crunch
c) the socio-political side-effects are better -- i.e. any bailout would not be for insolvent banks and poorly managed companies but for defaulting homeowners, by making them tenants instead of homeless, and the opportunities for corruption are greatly reduced
d) worst case in the long term, the government gets stuck with $700B of new public housing that's hard to sell (but has tenants), whereas the current bailout's worst case is just burning $700B on paper that ends up being worth $0.

I'm no expert but the more I read about this, the more I believe that it's better to solve the problem at the source, i.e. #1 and not #3.

I learned later that this is similar to something done in the 1930s called the Home Owners' Loan Corporation (HOLC), and has been advocated in the current context by Roubini. Some politicians have also called for it, but in addition to TARP!  The Washington  Post too  has a column proposing something along those lines.

In any case, TARP was approved today, so we'll see. If it gets implemented, I predict scandals and a blue ribbon inquiry comission circa 2010.  Hopefully it will work in spite of that, because alternatives are probably eliminated.