3 ms·

It seems like a reasonable comment. I don't work in finance, but he seems to describe a feedback loop that's at least plausible. He links to a paper with a lot

It seems like a reasonable comment. I don't work in finance, but he seems to describe a feedback loop that's at least plausible. He links to a paper with a lot of numbers and charts in them, written by some Economics professors. It seems like a positive comment to have around. The only thing weird about credit agencies is that they're sometimes written into state and local laws as being the sole judge of whether some money can be invested in a certain way. The law shouldn't depend on private companies. The other thing is that because pensions are managed in aggregate, the individuals investing in them don't have much choice in what to invest in. It's the same problem with Social Security. It's a weaker but still valid concern with 401(k) retirement accounts, since your employer determines what you can invest in. It's just a bad idea to let other people manage your money for you, especially at scale. Nothing about the credit rating agencies needs to change(by force), but I'd say it's important that people not be forced into trusting them. Once that's in place, it's on you if you blindly trust a credit rating agency. They're just a private company, so it (should be) legally the same as paying my bartender for stock tips. I don't know - do credit agencies ever issue contracts guaranteeing their ratings? If you get one of those, the bartender analogy doesn't hold.

haberman 11y ago
I think the comment is informative, but I think the attribution of blame is indefensible.
- MichaelBurge 11y ago
  As a Scala quant, it sounded like the guy was thinking about the problem analytically rather than emotionally. Then, the phrase "the public caused X" would be more like how a software developer would interpret "this buffer overflow caused the attacker to gain a foothold in this 10-stage attack on some server". The determination of an appropriate scapegoat is more of a politician or PR guys' job, not a quant. Here's another one: You're a software developer. Software developers caused Heartbleed. Why do most software developers "live in a reality distortion field where it wasn't their fault"? Probably because they don't work on SSL libraries. Should I blame you for not taking responsibility next time Cloudflare goes down, even though you might have an outside opinion on what caused it? The user in question works at an online advertising company, not at a credit rating agency. Why do they think it 'was not their fault'? Probably because it wasn't his fault.
  - MichaelBurge 11y ago
    I can't edit it, but that should read "Given that he's a Scala quant". I'm not a Scala quant.
  - haberman 11y ago
    That is a very charitable reading of that person's comment. But look at the follow-up: https://news.ycombinator.com/item?id=5168773 https://news.ycombinator.com/item?id=5168773 > Nobody adjusted the models when they produced incorrect answers because the answers weren't incorrect. They weren't the answers you like in hindsight, but they were correct insofar as the model was vetted & the input was correct. This is 100% absolving the agencies of any responsibility for the brokenness of their models. According to this person, it's the public's fault for behaving in ways that led to unrealistic ratings, and not at all the agencies' fault for having models that didn't reflect reality.

haberman 11y ago

I think the comment is informative, but I think the attribution of blame is indefensible.

MichaelBurge 11y ago
As a Scala quant, it sounded like the guy was thinking about the problem analytically rather than emotionally. Then, the phrase "the public caused X" would be more like how a software developer would interpret "this buffer overflow caused the attacker to gain a foothold in this 10-stage attack on some server". The determination of an appropriate scapegoat is more of a politician or PR guys' job, not a quant. Here's another one: You're a software developer. Software developers caused Heartbleed. Why do most software developers "live in a reality distortion field where it wasn't their fault"? Probably because they don't work on SSL libraries. Should I blame you for not taking responsibility next time Cloudflare goes down, even though you might have an outside opinion on what caused it? The user in question works at an online advertising company, not at a credit rating agency. Why do they think it 'was not their fault'? Probably because it wasn't his fault.
- MichaelBurge 11y ago
  I can't edit it, but that should read "Given that he's a Scala quant". I'm not a Scala quant.
- haberman 11y ago
  That is a very charitable reading of that person's comment. But look at the follow-up: https://news.ycombinator.com/item?id=5168773 https://news.ycombinator.com/item?id=5168773 > Nobody adjusted the models when they produced incorrect answers because the answers weren't incorrect. They weren't the answers you like in hindsight, but they were correct insofar as the model was vetted & the input was correct. This is 100% absolving the agencies of any responsibility for the brokenness of their models. According to this person, it's the public's fault for behaving in ways that led to unrealistic ratings, and not at all the agencies' fault for having models that didn't reflect reality.