Every financial institution has a different application process and different procedures for assessing your application. AMEX uses household income.
Amex are primarily concerned with the ability to repay.
So if the applicant has a high household income, despite having a low personal income, Amex is comfortable that the card will be repaid.
It is this same thought process which is why they so heavily push supplementary cards.
Ditto as to why getting new Amex cards is so easy for existing Amex cardholders. If you have a good record or meeting repayments that is more important to Amex than income. So getting new cards is easy.
There is a lot of data out there about lending patterns and the bank rely heavily on these in profiling credit risk, in that they look at a number of factors and have statistical evidence of which one help them in assessing risk. As an example you might think having a mortgage means you have greater financial commitments and hence are less likely to pay other loans but the data says differently.
Similarly there is a (inverse) relationship between household income and failure rates. You may well be correct that in theory Amex can't access a spouses money but the evidence is quite strong that if there is money in the household spouses do feel some sense of commitment to help their partner out with their debts. I may be wrong as to the motive, but what I am very sure of is that lenders look at a number of indicators to assess peoples ability to pay, if there is a correlation they will include this in their assessment. And household income is one of these factors.
Supplementary cardholders are included on card insurance so that in the event of the cardholders death or serious accident, there are funds to pay the final balances.
burmans is correct there are many external factors which contribute to the internal amex credit worthiness scoring.
Items which contribute include: -
- Where you live and ownership status (median HH income for that area, also amex can see if you own/rent that property)
- Where you live is also an indicator for job security, there are numerous studies which show low employment in areas which are not as highly connected via roads and public transport.
- Telco meta-data can be up to 50% more effective in determining credit worthiness than historical banking data at a transaction level. Some European telcos make as much as 16% of their revenue by *selling your data* to third party corporations.
(further reading:
http://mitsloan.mit.edu/media/Lo_ConsumerCreditRiskModels.pdf and
http://alo.mit.edu/wp-content/uploads/2015/07/ML_FAJ_FINAL_003.pdf)
- Personality data also factored into the equation. Repayments are all about trust - ie: the bank trusts you will repay it. Some types of people have lower trustworthiness and therefore represent a slightly higher credit risk.
- Much much more
It also wouldn't surprise me if QF and VA were selling member data to banks to assist with credit applications.
Payslips are totally redundant these days, especially as we move to a data-driven society.
Veda scores only address the 'how much revolving credit does this person have', and anyone in data science will tell you this is only a minor factor in assessing true credit worthiness.
So while Veda is useful for banks now - their power will greatly reduce over the coming years until they move to a credit holding body. Eg: Veda wants to become the credit body where instead of individual institutions assigning you a credit limit, your personal limits are held with Veda, and you can carve it up and assign different limits to individual credit providers. So you may have 100K revolving credit available and assign 20K to an ANZ card, 20K to Citi etc. In this sense, you will be able to switch your credit from one provider to the next.
The next few years will be interesting.