-
@Baron-Silas-Greenback said in British Politics:
@Kirwan if that is your new standard, l look forward to all your posts being expressed perfectly .
Just worry about your own posts.
-
@Baron-Silas-Greenback said in British Politics:
@Kirwan Which is exactly what I was doing until you jumped in to talk about my posts
Which takes us back around to you being unnecessarily aggressive and obnoxious. Take that point onboard instead of trying to deflect.
Not complicated.
-
@Baron-Silas-Greenback said in British Politics:
@JC said in British Politics:
@Baron-Silas-Greenback said in British Politics:
@JC said in British Politics:
@Rembrandt said in British Politics:
Just listening to Markus Meecham (Count Dankula) on a podcast. He's been trying to buy a house and despite meeting bank criteria he was decline from numerous banks because 'Nazi pug'. His mortgage broker says he's never seen anything like that in 20 years. He has finally found somewhere but it's pretty incredible. Worst crime in the world clearly.
TBF, it’s not a conspiracy, it’s because he has a conviction. Now it may be unfair that he has got one, but it’s on his file and that’s the end of it. FYI unless a loan is out of standard criteria his application will almost certainly not have been viewed by a human. Pretty much all banks use loan origination systems that connect directly to credit bureaux. The applications are scored based on correlation with a huge number of data points, and not all of them are intuitive. The “Scorecard” workflow generally assigns the applications to one of 3 buckets, approve, conditionally approve or reject, with actual people only looking at the middle one. With a conviction his will be going straight to reject. Harsh, but that’s the way it works.
I have never seen an automated loan system that rejected based on a non custodial sentence. And I have been involved in writing 2 such systems. One of them even a secondary loan system.
I am not 100% sure of many things, but I am 100% sure about this, You are wrong, I still actually have code for one of the processes I wrote for the sub prime company (don't tell them)I’m installing one now BSG. And it does deduct a sizeable amount from the score for a conviction. It’s entirely normal. To be clear it’s not rejecting based on the conviction, it’s scoring based on a statistical analysis that correlates convictions with defaults. Some of the correlations that are statistically significant are pretty strange. People with high assets are more likely to default that those with medium sized assets.
Yes I know how it works. And you are wrong. They distinguish on conviction type.
A custodial sentence for murder, rape or fraud is not graded the same as shoplifting or drunk driving.You do NOT get a binary response. And you certainly don't use a binary factor if convicted true/false
I don’t want to derail a British Politics thread into banking systems but given how strongly you obviously feel about this I’ll answer you.
I’m not wrong. When I go back into work on Tuesday I’ll continue working on implementing one of these. My fifth. This time I’m merging two existing ones.
The characteristics and attributes that are used for scoring the applications are selected from a universe of about 600 by assessing their ability to predict loan default behaviour. The ones that make the cut are weighted according to their correlation and built into a scorecard. The scorecard starts with an serviceability calculator and if that passes it applies a baseline credit bureau score (or composite if multiple bureaux are used) and adds or deducts points based on the characteristics. I’m sure you know this. The answer to the question “Do you have a conviction” (bearing in mind the constraints imposed by the Criminal Records (Clean Slate] Act 2004) is highly correlated to subsequent default. In the ones I have been associated with the type of conviction is not relevant, because it doesn’t provide any tighter correlation. The fact of imprisonment isn’t a factor at all, for 3 good reasons. First, asking for that data generates a requirement to protect it in line with the Privacy Act 1993, an update applying from this month of which requires declined applications to leave no footprint on applicants’ credit history. Unless the data provides additional information that would affect the decision, it is better not to hold it. Second, the credit bureaux already know this information and build it into the scoring, just as they do bankruptcies and defaults. Third, the effect of imprisonment that has relevance for a loan application is deprivation of income for a period, which puts the ability to repay at risk. Temporary cessation of income for any purpose is highly correlated but will be flushed out as a matter of course. Applicants are asked to provide consent to use a tool such as Proviso to access and scrape their bank account transactions to provide verification of income and outgoings. A period where an applicant has no income of any type for any reason (perhaps because they are in prison) is very easily picked up. It is used to adjust the score same as anything else, then the data is discarded.
The purpose of this is to create a final score that is used to place an application into a pricing band. In theory every loan in approvable providing the rate is sufficient to cover the risk. But in practice banks try to avoid applying the rates that would need to apply to a loan that is in the high risk category because they are concerned about their reputational risk. So they apply a hard cutoff at a floor score instead. For someone who does not already have a high credit score the negative scoring attached to a conviction of any type will generally be enough to place the score below the floor score. That will result in an automated rejection. Banks do not encourage their staff to second guess the decisions as it is time consuming and works against the model.
There are people in NZ who know more about this than me. But not many and I think I know them all. I’m still learning - it’s a fascinating field - but trying to characterise me as some kind of bullshitter is unnecessary. I’m happy to discuss this or anything else with anybody who’s interested and prepared to be civil. I will say, in my defence, that I learnt decades ago to be careful venturing advice on things I know little about. In this case I am entirely comfortable that I know a great deal more about this, and (given for many years I ran a team of many thousands of banking ops and technology people) about banking operations in general, than most.
-
@Baron-Silas-Greenback said in British Politics:
@JC
I am impressed you went to such lengths and typing. I was about to start writing a rebuttal, then realised that I just lack the willpower. Will respond next week hopefully.All good. Enjoy the rest of your weekend.
-
-
-
The Commons didn't give two-thirds ratification for a December 12th election yesterday.
So Johnson is coming back with a one-line Bill for a 12th December election.
Almost certain to become law, but might be amended for a slightly earlier election (Liberal Democrats favour 9th December) or to give 16 years olds the vote.
-
@sparky said in British Politics:
The Commons didn't give two-thirds ratification for a December 12th election yesterday.
So Johnson is coming back with a one-line Bill for a 12th December election.
Almost certain to become law, but might be amended for a slightly earlier election (Liberal Democrats favour 9th December) or to give 16 years olds the vote.
But 9th December is a Monday??
-
-
This post is deleted!
-
@sparky said in British Politics:
@JC No law saying a UK election has to be on a Thursday, just tradition since 1930s. Go back far enough and elections used to take place over several days.
An election probably can't be on a religious holiday or day of obligation.
That counts out Ramadan then.
British Politics