[mfs01]

This program would be great for a preapproval for the sales agent to pin the client down. But a TRUE approval cannot come until the documents, credit, etc.. are reviewed by an underwriter.
We used to have a "cookie cutter" underwriting system when I was in the mortgage business and it worked well for the sales agents and getting pricing, but in the end look what happened there.
This Automated system is a great Idea and will help the sales agents, but I don't think an computerized algorithm is enough to calculate the risk of a cash advance deal (at least not for me).

[Gini Co]

This program would be great for a preapproval for the sales agent to pin the client down. But a TRUE approval cannot come until the documents, credit, etc.. are reviewed by an underwriter.
We used to have a "cookie cutter" underwriting system when I was in the mortgage business and it worked well for the sales agents and getting pricing, but in the end look what happened there.
This Automated system is a great Idea and will help the sales agents, but I don't think an computerized algorithm is enough to calculate the risk of a cash advance deal (at least not for me).

If not an algorithm, how do you calculate risk or provide risk based pricing? Your good judgement? It's not that you don't have a human underwriter confirm and verify the inputs to the scoring/risk model; and interviews are part of that process. But to say that you can tell if a person "knows their business" from an interview and therefore is credit worthy is a joke. The merchant interview is a 0/1. If the guy is sketchy, you don't give him anything, otherwise you fund according to the model. I hope everyone is just kidding here and that you're not 'lending' money on your good judgement.

[Sean Cash]

But to say that you can tell if a person "knows their business" from an interview and therefore is credit worthy is a joke. The merchant interview is a 0/1. If the guy is sketchy, you don't give him anything, otherwise you fund according to the model. I hope everyone is just kidding here and that you're not 'lending' money on your good judgement.

I agree that it's a 0/1, the merchant doesn't have to be sketchy though, he can just be clueless. I've been on plenty of merchant interviews where the guy "sounded nice" but had no idea what the margins on his products were, didn't know what he was going to use the money for, had no idea why his sales had fallen, had just inherited the business from a family member or had been thrust into ownership and didn't know a single thing about his product.

Using good judgment from a good underwriter on a borderline deal is sometimes all you have. The deal looks pretty bad on paper, so let's talk to the owner. If he's clueless or sketchy, 0, if he's got all his ducks in a row, 1.

I know that Discover Bank does merchant/consumer interviews just to get approved for a credit card. If they don't like you, 0. If they like you 1. The underwriter is then judged the on their portfolio's performance. Discover pulls a lot of Data too and it still comes down to good judgment sometimes.

[Capital Stack]

Gini, I completely agree with you. Teaching the algorithm can be more resourceful and accurate than a human underwriter just using there best judgment. But I also agree with still doing the verification call as an important step in the final process.

If not an algorithm, how do you calculate risk or provide risk based pricing? Your good judgement? It's not that you don't have a human underwriter confirm and verify the inputs to the scoring/risk model; and interviews are part of that process. But to say that you can tell if a person "knows their business" from an interview and therefore is credit worthy is a joke. The merchant interview is a 0/1. If the guy is sketchy, you don't give him anything, otherwise you fund according to the model. I hope everyone is just kidding here and that you're not 'lending' money on your good judgement.