anon_
26d ago
No, they assumed it was uncorrelated and that you could lower risk by bundling large tranches of mortgages.
That works until there is a large correlated event that impacts all of them at once.
anon_
26d ago
No different than any other public obligation or liability. Might as well shaft municipal bondholders too on this theory,
anon_
26d ago
Right, and I think the OP is trying to shoehorn "the defendant didn't trust the police/state" into that set, as some kind of faux-historical analog of "things that can't be held against you in court".
anon_
26d ago
The problem with the '07-08 crisis wasn't with the returns, it's that the loans were packaged into MBS and sold to investors under a false bill of health. The lessons weren't that you can't have high-risk/high-return assets, only that you must not try to pawn them off as low risk with fanciful assumptions. And that buyers of those collateralized debt must do more diligence.
anon_
26d ago
Why is "civilizationally load bearing" a relevant yardstick? Civilization advances by making all that stuff trivial so people can focus on doing other shit. That's the the measure of civilization, not who produces corn.
And no one that has ever worked in textiles would dare assert that it's in any way worse than an office job. That shit destroys your body.
anon_
26d ago
To be fair, most government-paid civil engineers take a much lower salary than they could in the private sector. So it's not quite that this guy figured out a hack here, only that he sacrificed early to reap a larger return later. Similarly the State took the inverse deal: pay him less now in exchange for more later, in order to make their budgets temporarily look better.
I think the solution for both is actuarial integrity -- defined benefit plans need to be run in such a way that the State pays in year X the expected future costs of all (incremental) future liabilities accrued during year X. The only real exploitation is that voters in X accrue liability for year >X without paying for it, another intertemporal transfer of wealth.
A non-solution (afaict) is for governments not to hire competent civil servants and instead farm that stuff out to McKinsey consultants and others. Not because the McKinsey consultants aren't smart, but because it's a diffusion of accountability that ultimately costs Idaho more than paying competitive salaries for in-house expertise.
[ One astute commenter noted that one good that McKinsey does produce is laundering the low status of working for bumfuck Idaho into PMC-respectability. An excellent observation, if something of a tangent here. ]
Presumably they have a net positive effect on GDP when measured on the spending side, and if we ASSUME they don't declare bankruptcy, or renege and duck out on the debt, or just die early (not something I wish on them), they're helping the engine of Capitalism in this country sputter along.
Eh, bankruptcy is (in expectation) priced into the transactions. Lenders make out fine charging these two 7% interest on their HELOC and car note. It's not like dumping it on the fisc.
Their retirement on the public dime, OTOH, will certainly be dumped on us.
anon_
27d ago
I don’t think “the defendant burned the bodies” would ever fall in those narrow categories.
anon_
27d ago
Should a defendant's distrust of the police be held against him in court?
It certainly should be presented to the court as a matter of fact.
There are a very small number of exceptional situations in which factually true things cannot be said in court -- for example, a prosecutor cannot comment on the defendant's refusal to testify in his own defense. Those exceptions are few and far between, and they prove that the rule applies in all other cases.
I agree that the tranch setup distributed risk as advertised. The problem was that much of market (including CDO sellers) believed in their models that used a gaussian copula, which vastly underestimated the tails as compared to a power law.
This wasn't "sellers pull a fast one on buyers" -- that's too simplistic a model. They got into the business because they convinced themselves of an overall model that didn't put enough weight on the tail risks. Then they kept the most junior tranches because of that belief in the low overall risk.
[ There's a related problem which is that gaussian models are extremely sensitive to parameter changes in ways that power laws aren't. When you get to the CDO-squared (a CDO of CDOs) then you the output of one model being fed as an input to another, with the expected impact to accuracy. ]
See, e.g., this excellent summary: https://www.hks.harvard.edu/sites/default/files/centers/mrcbg/files/Barnett-Hart_2009.pdf
More options
Context Copy link