26

I live in the UK, and I recently checked my credit score for the first time, on Experian. I saw that it is a number out of maximum of 999.

I was surprised to see that there is a maximum credit score to compare against. For example, imagine someone who has the maximum credit score, and they are 35 and earn a salary close to the average salary for their country, and don't own any properties or other assets. Then this person has obviously some reasonable potential to pay back any credit given to them, based on their income.

Now imagine instead someone of the same age who earns three times as much as the first person, and they also own two properties and a small business. Surely the second person has a higher potential for paying back larger amounts of money loaned to them? How is it not possible for the second person to have a higher credit score than the first person, because the maximum credit score is 999 (on Experian in the UK)?

My intuition was that there would be no limit to the amount that your credit score could increase, given your circumstances. For example, I imagine Bill Gates' potential for paying back any money loaned to him would be higher than either of the two people I made up.

16
  • 8
    I don't think this is going to be answerable. Those scoring mechanisms are highly proprietary, and I doubt anyone involved in creating them is a member of the site.
    – JohnFx
    Jul 23 at 12:04
  • 5
    Which country are you in? Jul 23 at 12:51
  • 22
    Literally any range of numbers can be mapped to the range 0-100, but that doesn't make it a meaningful percentage - a credit score of 730 doesn't represent 73% of anything. The real question here is why do credit scores have an upper limit. Jul 23 at 12:58
  • 32
    Billionaires are not necessarily good credit risks.
    – user662852
    Jul 23 at 13:03
  • 2
    @NuclearHoagie An infinite range could also be mapped to the range 0-100. This would make sense if you decide greater scores have smaller and smaller effects on risk, it stops having an effect at some point or there exists little to no people with a score above some value, thus you map it to a finite range to more clearly show this. Or maybe the best scoring function they could come up with just directly maps to a finite range; that wouldn't really have an explanation other than "that's the best function" (unless you go into the exact details of what that function looks like and why it's good)
    – NotThatGuy
    Jul 23 at 22:33
54

A person's credit score is related to their risk of defaulting on a loan. When evaluating if your scoring model is any good, though, you can't evaluate the likelihood of default from one person, since every borrower either does or does not default. In order to calibrate a credit score model, you need to aggregate data from many individuals to determine whether the score accurately reflects risk or not. Due to this, it's not really useful to have an infinitely extending range of credit scores that apply to vanishingly small proportions of the population. Having a score range that is too wide does not give sufficient ability to "bucket" borrowers and evaluate aggregate performance.

You can give Bill Gates a unique credit score and predict his likelihood of default, but it would be impossible to tell whether your score model is accurate or not - whether or not Bill defaults, you have no means to evaluate whether his assigned credit score and predicted default probability was reflective of his true risk. You need many other people with the same credit score to make that determination.

Furthermore, it seems that some people defaulting is quite unpredictable from the data used to compute credit scores - roughly 0.1% of people with the highest possible credit score still default on their loan. Empirically, it's actually a slightly higher rate of default than people with a slightly lower credit score (see Fig 1 here, which also contains more information on how credit score ranges are defined). As you get to the very top of the credit score range, it becomes far less predictive of actual defaults. Once you're over a score of 930, having a 940 or 950 actually doesn't decrease your risk of default.

The upper end of the credit score range is capped because 1) higher scores would apply to few people, be counterproductively fine-grained, and have indeterminate reliability, and 2) higher scores would likely not be additionally predictive of default risk anyway.

8
  • The reference you posted was also instructive and helpful for me.
    – Joe
    Jul 23 at 14:40
  • 13
    Perhaps Donald Trump would be a better example than Bill Gates (who AFAIK pays his debts) for credit scores not being an absolutely reliable predictor.
    – jamesqf
    Jul 23 at 18:42
  • 12
    @jamesqf why would you think Trump has a high credit score?
    – Caleth
    Jul 23 at 22:40
  • 6
    @jamesqf Actually, Trump has had substantial trouble finding banks willing to loan him money; Deutsche Bank was one of the last, and they've dropped him. nytimes.com/2019/05/19/business/…
    – ceejayoz
    Jul 24 at 14:38
  • 2
    But Trump has no trouble finding ‘believers’ that not only loan him money but donate/gift it to him. He doesn’t need the banks…
    – Aganju
    Jul 25 at 4:06
18

The credit score is an evaluation of your credit history. It doesn't know how much you make. It looks at how much of your available credit you use. It looks at how long you have had credit lines. It looks at the number of on-time and late payments.

If you have a very bad score lenders will either refuse your loan application, or they will charge a higher rate. If you have a great score they might charge a lower rate.

As a part of the approval process a lender will look at how much you make, and how much you owe, and then determine how much you can afford to pay for the new loan.

If the loan you want is affordable at the interest rate they have to charge, and your score is good enough they will approve the application.

The scoring system is just part of the loan approval process. A multimillionaire can have terrible score, and a young adult a few years out of school can have a great score. They put an upper limit, because at the upper ranges there is very little difference in the credit history. You don't need to even reach the maximum score before you are viewed as a low risk of defaulting.

3
  • Thanks for your answer @mhoran, this is a good explanation for me about how credit scores work, and "at the upper ranges there is very little difference in the credit history" partly answers my question. I understand based on your response that this maximum value is somehow correlated with a credit history of someone who has always made their debt repayments on time. Does this seems reasonable to you?
    – Joe
    Jul 23 at 14:19
  • Just to confirm this Answer: credit.com/blog/… Jul 26 at 17:27
  • @Joe, there's a significant amount that goes into a credit score. This includes the length of loans, how much of your available credit you're using, how often and what type of credit checks are happening to your score, and much more. The credit scoring methods are different per credit bureau and are proprietary (and not public knowledge), so there's a lot of guessing as to how exactly credit scores are calculated. So, paying your debts on time are just one of many ways to increase your score. Jul 26 at 17:31
9

First of all, I'm not convinced that your credit score is a "percentage" (percentage of what?) Credit scores in the UA have maximums as well ,but they are less that 999 (and different depending on the provider). So I don't see how it can be translated as a "percentage".

But as for the rest of the argument - a credit score is a measure of your debt payment history. It has nothing to do with your salary or how much debt you can pay back.

So I'll propose two more scenarios to refute your argument. Suppose you have person A that makes barely enough to live on, uses credit sparingly, but always pays it back. Then you have person B who makes a lot more money, but uses credit exceedingly and often misses payments.

Which one do you think a bank will consider to be a "safer" borrower? Which one would get a better interest rate on their debt?

Now, when you go to get an actual loan, then the bank will look at the amount of the loan compared to the collateral, your income, etc. as well as your credit score to determine first if they will even make the loan, then will determine an appropriate interest rate (mostly based off of your credit history). Riskier borrowers (lower credit scores) get higher interest rates.

3
  • 1
    "percentage of what?" That may be what OP is asking.
    – RonJohn
    Jul 23 at 13:11
  • @D Stanley "I don't see how [credit scores rated out of a maximum value] can be translated as a percentage". The calculation is; (credit score)/(max credit score)*100, which is valid for any credit score which goes from 0 to some maximum value. @RonJohn is correct that I'm asking, "what does the maximum value mean"?
    – Joe
    Jul 23 at 14:09
  • @D Stanley, thanks for explaining how credit score is calculated I didn't understand that it was based only on previous debt payment history. This doesn't really explain to me why there is a maximum value though
    – Joe
    Jul 23 at 14:11
5

The question seems to reflect a misconception that a credit score is intended to reflect the amount that can be safely lent to someone. Rather, it is intended to reflect the likelihood of repayment for a loan of a given size relative to the borrower's income. It is natural to have an upper limit, which would correspond to (nearly) 100% likelihood of repayment.

The theory is that finances are approximately scale invariant -- a person with $30k/year income borrowing $5k typically has a similar repayment likelihood as a person with $300k/year income borrowing $50k. The latter person will tend to have 10x higher expenses, credit limits, and balances; their finances are essentially the same with an extra zero on everything.

So the credit score does not incorporate any absolute dollar amounts, but only percentages and ratios (such as credit utilization rate) and non-monetary quantities (such as number of late payments and length of credit history). It does not incorporate income at all. Lenders, in turn, scale the amount they are willing to lend (at a given interest rate for a given credit score) with the borrower's income.

Thus, equally responsible people at different income levels will tend to have the same financial ratios and the same credit score.

4

Consider the usage of a credit score: it is an assessment of how likely a borrower is to pay back debt as agreed, used by lenders to decide whether to loan money and how much to charge for the loan (interest rate). The useful range for this assessment is from "will almost certainly miss payments and/or default" to "will almost certainly make all payments on time", with variable "buckets" of risk along the way (where people in riskier buckets pay higher interest to account for the risk of loans in that bucket defaulting).

Once you get to the higher end of the score spectrum, where you trust that the person will most likely make all payments on time, it really doesn't add any value to further differentiate between borrowers.

3

Your question is based on a flawed premise. Even if you can make it look like a percentage, it is still a score.

Per Experian in the U.K., credit scores range from 0 to 999.

We consider a ‘good’ score to be between 881 and 960, with ‘fair’ (also the average) between 721 and 880. However, there’s no ‘magic’ number that will guarantee lenders will approve an application if you apply.

If your credit score is poor, you’ll probably find it harder to borrow money or access certain services. We consider a ‘poor’ score to be between 561 and 720, with ‘very poor’ between 0-560. But remember, lenders may have different views of what an ideal customer looks like to them, which will be reflected in how they calculate your credit score.

This is conceptually similar to sovereign/corporate credit ratings, which range from C to AAA.

Theoretically, you could have an infinitely high credit score, if it were based on income/assets. Conversely, there's no reason it should be floored since some people have large amounts of debt/are facing bankruptcy.

Credit scores are used to categorize borrowers based on their creditworthiness, not to compare individuals by saying person A is twice as creditworthy as person B based on their age/income.

3
  • The idea that credit scores are not intended to be used to compare individuals is interesting, I had not thought about this. Do you have any evidence (eg. references) that they're not designed to be used this way, or that they don't function well in this way?
    – Joe
    Jul 23 at 14:23
  • 2
    I don't have a reference handy but think of it this way: banks don't extend large amounts of credit to few people, they lend relatively small amounts to a large number of people. They group these people by creditworthiness to assess how many of them will default on average, so that they can charge the appropriate interest rate for it to be profitable in aggregate.
    – 0xFEE1DEAD
    Jul 23 at 14:26
  • Put differently, borrowers are assigned to a category based on the range their individual score falls into, e.g. 721-880 or 880-960, rather than being ranked against all other credit scores.
    – 0xFEE1DEAD
    Jul 23 at 14:32
2

tl;dr:

Beyond the max doesn't provide any value. Decisions that are made based on credit scores are almost always "gates" based on a minimum score (oftentimes layered in tiers). Once you pass the gate, having a higher score isn't necessary.

Details:

First a minor correction: note that income and assets are not relevant in credit scoring models. (US, UK, and other countries too.) There are 2 different concepts at play here:

  1. Your credit score is (designed to be) mainly an indicator of how likely you are to pay back debt on time.
  2. Your income in conjunction with your current expenses (oftentimes called debt to income ratio or DTI) is used to determine how much you can borrow.

Therefore, higher incomes, or even whether you have any income at all, should have no bearing on your credit score. Note that you might have a maxed out credit score, but if you have no income or assets, you may still not be able to get a loan.

But we can tweak your question to make it relevant: Take 2 people with a maximum credit score. 1 person has more good history than the other, such as a longer history, more of each type of credit, and a much higher max limit on all of their current lines and a theoretical better utilization percentage any given month. Shouldn't that person have an even higher score?

It's simply not necessary because decisions based on credit scores are qualifying "gates". For example, in order to qualify for a loan with rate I you need a minimum score of X. If you qualify, you pass that gate and move on to the next gate, which probably looks at your income and determines how much you can borrow. Whatever the highest score tier is, once you pass that "gate" you're through; it doesn't matter if you're barely over the minimum or way over it.

As a silly analogy, when I was in HS, in order to make the varsity soccer team, you had to run a 6 minute mile or better. There was one person who ran 4:57 and he... made the team. (There was no concept of really making the team1.) And, there were a few people who got under the 6 minute mark but still didn't make the team, because their skills weren't strong enough.

1 There was an exception made for the one kid who was a superstar from Ukraine; he had made the team before he arrived in the US, since he was brought here specifically to play soccer. I'm pretty sure he made the six minute mile anyway though.

1

Credit scores are explicitly risk assessments. How do you score "higher" than "has had many types of credit, has no debt, and has never been late or defaulted".

While you can take this maximum value and then treat it as a "percentage", that's not what it is intended for -- basically it's a banded range of risk, very low, low, medium, high, very high, and are you crazy (or possibly unable to say).The credit score itself is broken down into sections, how long you have had credit, how much of your credit is being used, defaults, etc.

The lender then looks at that band as well as the individual sections of your credit report, and uses it as part of their assessment for a particular loan, in particular it frequently affects the interest rate, the greater the risk, the greater the interest rate.

0

Credit scores have the values they have because somebody decided it was a good range for people to look at and understand, while still having enough ability to differentiate at the level of accuracy needed to make the decisions.

In the US, most scoring models mimic the base FICO (Fair Isaac and Company) score, because that's what people are used to - the original VantageScore (made by some of the credit bureaus to compete with the various FICO) didn't quite mimic it, and it caused confusion as people didn't understand why they didn't match, so later variants largely use the identical range. FICO scores range from 300 to 850, which is just as arbitrary as the score you ask about; presumably the reason to choose a base of non-zero is for psychological reasons - if someone tells you you have a score of 0, that would feel very different from saying you have a score of 300, even though it means the same thing. The actual number doesn't matter - it doesn't mean anything in and of itself.

It's not really different from temperature scales, for example. Celsius and Fahrenheit measure the same thing, but a 50°C day would suggest a different choice of vacation spot, while a 50°F day would suggest wearing a light coat. They made different choices for 0 point and degree size for different reasons - Fahrenheit, to reflect on 0-100 the temperatures a person might normally feel, while Celsius was chosen to be more scientific - but the point is both were chosen essentially for psychological or usage reasons. The numbers don't actually mean anything, without the scale.

Not the answer you're looking for? Browse other questions tagged or ask your own question.