[sksk]

>> Why is a rich person with no debt and huge amount of assets more of a risk than an undischarged bankrupt?

Good point and as mseebach points out absence of information is a signal and these type of no-debt cases are very rare that is hard to build a model around.

At our start-up SimplyCredit, we look at things holistically, not just your credit score and that should help with your concern. However, it is harder to do this effectively without proper data at scale. I wish the govt. made different types of data about the borrower available to lenders (with permission from the borrower of course) -- for example, it will be nice to know electronically someone paid for mortgage outright or their income statements, etc. Right now people attempting to use this information have to resort to approaches that Mint uses (scraping, partnership with banks to not block them, etc.).

That said, credit scores actually work in a manner similar to what you suggest. Credit scores are 'centered' around a value and good things get you +score1 and bad things get you -score2, bad things you didn't do +score3, etc. It is not exactly what you suggest as no-hits don't automatically get neutral rating -- most case those apps will get rejected.

[danieltillett]

People without a credit score fall into three categories which really have nothing in common in regards risk. 1. The unbanked (i.e. very poor). 2. The very wealthy or the moderately wealthy with an aversion to debt. 3. Recent immigrants. While I suspect 2 is relatively small, both 1 and 3 are quite large pools. Treating all three as a greater risk than someone with a history of poor debt management seems perverse.

The risk profile of a recent legal immigrant with a PhD and a high paying job is very different to someone that has never had a bank account and who has worked for cash in hand. Why treat them the same? At the very least let someone with no history of credit establish a credit score baseline based on their international history, current income, and assets.