With the economy tanking one of those comments you hear from time-to-time is that China, whose holdings of US debts amount to $585 billion, or over 4% of our entire economy, will suddenly call in our debt, crushing our economy.
Is this scenario plausible? Might China actually do this? I can't say it more strongly: no.
Unless I'm missing something, I can base this conclusion on
OK, I'm going to kind of gloss over number two because I did not follow this in my intermediate macro class, but I'm pretty sure this issue of debt is tied into trade deficits. So for China to call in the debt would mean that China would also not be able to continue selling us more than we sell them. And I don't think they want to do that. (If anyone can correct me on this one, please do)
But the third one is also pretty key. Let me add here that Japan holds almost the same amount of debt that China does, and combined they hold about 9% of our debt, or $1.15 trillion. So, I was going to assume that the treasury rate was about 4% returns on this debt. Unfortuantely, I just looked it up and realized it's closer to 0.4%. This really kills my argument, but I'll keep going anyway. So, 4% of $1.15 trillion means a return of $46 billion per year on this investment. By contrast - think about how much fuss there was over the $25 billion for the automakers. Even at today's rate of 0.46%, this results in annual interest of $5.29 billion. And that's not a bad take. OK, you're right it kind of is. Dammit.
Posts
It doesn't matter how shitty the return is because all the alternative investments out there right now are even worse. That said, however, if the holders of the debt are distressed they may be forced to sell if they have no better asset to liquidate.
ReplyDeleteI enjoyed this article in this week's NYT Magazine. What about that 1% chance...
ReplyDeleteI haven't even gotten past the first two paragraphs yet and I can't help but comment.
ReplyDeleteParagraph 1: Really? No consensus on whether qualitative or quantitative assessments of risk are better? Go talk to a psychologist.
Paragraph 2: Risk assessments are not the primary cause with this debacle. The activities the banks involved themselves in were downright fraudulent. Hell, I experienced that myself, when a friend and I were trying to find out if buying a cheap house could make financial sense. The loan officer required us to disclose our student loan amounts so they could count as income (when I said I didn't have them on hand, she just asked me what they were). So yeah, apparently I was making $45,000 my junior year.
Alright, I finished the first page (wow this is long - probably won't read the whole thing) and I'll ease up a little. But in fairness, even the article itself says that the error wasn't statistical so much as that the model was "fraud."
ReplyDeleteOK, I'm on page 5. This guy said what I've been thinking way better than I did.
ReplyDelete“Obviously, we are big proponents of risk models,” he said. “But a computer does not do risk modeling. People do it. And people got overzealous and they stopped being careful. They took on too much leverage. And whether they had models that missed that, or they weren’t paying enough attention, I don’t know. But I do think that this was much more a failure of management than of risk management. I think blaming models for this would be very unfortunate because you are placing blame on a mathematical equation. You can’t blame math,” he added with some exasperation.
OK, so I did read the whole thing. I admit I was a bit harsh at first but in the end, this was more a problem of human error than mathematical error. Two thoughts.
ReplyDeleteOne, the article once mentions the 95% VaR vs. the 99% VaR (that is, the most money you could lose ignoring the worst 5% of circumstances and the most money you could lose ignoring the worst 1% of circumstances). To not look at both is, to me, astounding.
Two, the VaR has a simple inherent bias. What if I told you that you had the opportunity to invest in a security that had 500 possible outcomes.
In outcomes 1-400, you would make $1 million.
In outcomes 401-475, you would make $2 million.
In outcomes 476-498, you would make $10 million.
In outcomes 499-500, you would lose $500 million.
Under this scenario, when every outcome has come up once, you will have lost a total of $220 million, or $440,000 per transaction.
Yet with VaR, under the best 99% of outcomes, you make $777 million, for an average of $1.57 million per transaction.
But that last 1% consists of 3 gains of $1 million and 2 losses of $500 million. So your VaR calculation ignores the 1% of the time in which you lose back $997 million.
This is what is meant in the article when they refer to "asymmetric risk."
Now, one simple fix (and I couldn't tell if this is what VaR does or if it does what I did above) would be not to include the 99% best scenarios, but the 99% most likely. Under this situation, asymmetric risk is still a problem, but when it isn't, unexpected huge losses will occur just as often as unexpected huge gains.