I was wildly speculating to myself what impact this mass event will have to credit scores, and the borrowing power of people in a post-COVID world where interest rates are expected to be low (in the short run, until they become anything but). Maybe I should have kept this to myself, but I’m curious who else is starting to run numbers?
I was refreshing myself on the models, and what information is in public domain (vs speculation, regression, or anything that might not be subject to nda)
Regardless of the FICO® Score version, the keys to obtaining favorable FICO Scores remain the same:
- Make payments on time
- Keep credit card balances low
- Open new credit accounts only when needed
What’s different in V8 vs V2
Keeping credit card balances low can help maintain or improve the score.
- Isolated late payments
If a lender reports to the credit bureau that you were at least 30 days late with your payment, it will likely result in a loss of points within all FICO® Score versions. If the late payment is an isolated event and other accounts are in good standing, FICO Score 8 is more forgiving compared to previous FICO Score versions.However, if the credit report shows numerous late payments, the reverse is true and the FICO®Score 8 will likely lose more points as compared to previous FICO Score versions.
So in the currently published version of the model, unless they roll out some kind of extreme change, which there are unlikely to do, anyone with two or more late pays is going to start losing points. While many banks and many car lenders are offering deferral’s, there are still a number of store cards and branded credit cards that probably won’t and will still report these late.
We will probably need 2 to 4 weeks of new unemployment data minimum to start modeling how many people might fall into this category, but I’m speculating about whether the model might skew as the distribution of credit scores changes drastically in a very short time.
Not the latest source but a quick illustration:
If you look at the distribution from 700 to 749, 750 to 799, 800 - 850: I could see any of those dropping by 20 to 50%, and if by some chance all of them dropped that much she would go from it being 40% of the population with the last hundred points of credit score to possibly 15 or 20%. If the distribution skew significantly, creditors will need to change their guidance or end up lending less money.
Also consider that many leasing companies price A-tier around a 720, which falls somewhere between 43 and 60% of the population. So more than 1/2 of adults with credit are eligible for A-tier.
So there are at least a few points:
- I don’t see anything in the most widely used credit scoring model that won’t penalize people for 2+ late pays. I also don’t expect FICO to change the current model. I also don’t think semantics in the model are the difference between hard times and financial ruin in the wake of hard times. To that end, people who default early will benefit least from low rates before they inevitably shoot up.
- credit scores will likely drop as a result of covid-19. how many, and by what %, hard to project yet. Deferments on bigger ticket loans (house/car) will likely help to reduce the number of late pays and score drops.
- risk is going to be more difficult to price, which I think will cause lease portfolio managers to pull back the volume of leases they offer (how? TBD: on fewer models? By reducing incentives?)
- I think all of this will see production return slowly (when it does) and every car company cut their projections by 10%+, meaning record car sales years are going to be a memory for a while.
We have a new normal coming in hot.
Edit: updated with this morning’s DOL unemployment chart
