TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Business Microloans for U.S. Subprime Borrowers [pdf]

30 点作者 iamchmod将近 11 年前

4 条评论

throwaway4545将近 11 年前
I work for a technology company in this industry, and its really cool to see this on the front page. There are some really interesting technical problems in figuring out the credit algorithms for loaning to small business owners. Banks don&#x27;t have the incentive to care about $10,000 loans, but it makes a huge difference (as seen in the paper above and our own data) to these businesses.<p>Its definitely an example of an industry where technology is &quot;disrupting&quot; hard. Loans which used to take 2-12 weeks to be approved can be approved by a computer in literally minutes, and the number of businesses that can be approved efficiently is drastically higher.
mschuster91将近 11 年前
Credit rating by computers already is too prevalent. Who is a machine to decide if one is credit-worthy? In most banks, you can&#x27;t pitch your idea like e.g. a startup founder can with angel investors because bank regulations say &quot;computer-only decisions here&quot;.<p>Have a bad mark (e.g. you were in jail for pot in your youth) and don&#x27;t get a credit for the rest of your life. This is the brave new world of full automation.<p>edit: also, this introduces a total lack of accountability and basic rights: &quot;when the regulations say that I am only allowed to accept new tenants when they pass the computer check, I cannot give you a home. Even if I like you personally, you have 100K+ income, and that pot jail sentence is 20+ years past - but the computer says no because you are black and a drug offender&quot;. Just think about the scenario. It is totally possible today if you feed the computer with the wrong data.
评论 #7870527 未加载
评论 #7870873 未加载
JackFr将近 11 年前
This paper seems to be all over the place, including subprime mortgage lending, commercial lending and pay day loans, really to no seeming point. Then they conflate small businesses with startups. All to reach the conclusion that firms who obtain credit are more likely to succeed.<p>There probably is a very good kernel of analysis in there, but this paper really needs to be tightened up.
dalek2point3将近 11 年前
any ideas on what company it is that they&#x27;re working with?
评论 #7870814 未加载