I think you're right that news orgs misrepresented that you'll be punished for having a greater credit *score*, but if we take size of downpayment as an indicator of creditworthiness, then it is true that there are many situations where the more creditworthy customers pay more than the less creditworthy. For example, the LLPA for the bottom credit score tier, putting no money down, is less than the LLPA for the next two tiers putting 80% down.
Good catch. It is weird. That's why I included mortgage insurance fees. High LTV loans have to have mortgage insurance, and, I don't know why, but the fee schedules assign more fees via mortgage insurance and less fees in the LLPA. When you add them together, the low down payments generally pay more.
Maybe there's some way, actuarially, to think about it that makes sense. I don't know.
I liked your article, thank you. I’m a loan officer in Milwaukee and I saw many in my network share/post the Fox News screenshot. Very misleading.
In practice, my response has been to help clients get into the “affordable” programs Fannie and Freddie offer, there’s an income based program (not even for first time homebuyer) and also a first time homebuyer program. Based on either <80% area median income and <100% respectively.
Both of those programs have a full waiver of ALL LLPA adjustments. 660 score gets the “same rate as grandma” with a 40% DP and 780 score. I call it “grandma’s rate” with my clients.
It’s not necessarily fair, but they appreciate it. When we as lenders are forced to use a GSE under conservatorship, you don’t really get an option besides to play with the rules we are given.
If there is a national finance market for X, and part of that market is a small fee to offset risks, and the level of the fee rate is based on credit scores, and people with worse credit see a decrease, while people with better credit see an increase, it's necessarily the case that one of two things (or both) is true:
1) the people with better credit are subsidizing people with worst credit
Or
2) rates are inflated and there's enough slush in the system to re-jigger them in all kinds of ways - maybe even some that are political.
I haven't written about this much on substack, but I agree with your intuition here. Basically, after 2008, private mortgage lending with any substantial default risk were regulated to oblivion. FHA and the GSEs have the "Qualified Mortgage" patch, which means that a lot of the new regulatory liabilities for mortgage originators go away if the FHA & GSE underwriters agree to purchase your mortgage. This basically gives the federal agencies a monopoly on much of the mortgage market, and they charge for that monopoly, transferring billions in profits over to the government each year. So, you are right. These fees can be fairly arbitrary, because there is no legal market to arbitrage to a reasonable risk/reward level. If it was legal, private lenders would be making mortgages to borrowers with a moderate amount of default risk at much lower rates and/or fees than what is currently available.
Ok but that's not what you're arguing in your article... The Biden administration made these changes and to a certain extent they are subsidizing the less creditworthy by increasing rates on the creditworthy - aka penalizing them in other words. The GSE underwriters are the housing market for almost all single family homes.
It seems like you wrote this piece to take the air out of headline readers who might think that bad creditors pay less than good creditors. That's not the case, but it totally appears that the good are subsidizing the bad here, and by design.
There is no evidence that these fees net out to a loss on those borrowers that needs to be subsidized. These articles led to an onslaught of reactions that were a gross mischaracterization of the policy. That was the motivation for my post. There is no need to debate the marginal differences in gains and losses various actuarial assumptions might suggest, relative to the new fees. The onslaught of mischaracterizations wasn't about actuarial estimates. It is just one more drop of poison in the well against reasonable mortgage policy, and I felt I should address it.
"""I’m pretty sure what they’ve done here is cherry pick the low credit score that had the largest fee decrease. Then, they reported the total fee of a higher credit score. So, a low down payment 620 score has a fee that went from about 6.75% to 5% (when mortgage insurance is included). And, also, the fee for a 740 score went from 0.25% to 1%. (plus a 0.25% mortgage insurance fee). Why didn’t they just say that fees for 740 scores went up 0.75%? It would still get their partisan point across. It would still be weird, because it would be describing mortgages with two different down payments. And it would hide the fact that the 620 score still has a fee that is more than 3% higher than the 740 score. But, at least it wouldn’t be mixing levels with changes.'""
How about for accuracy's sake we use those numbers and report that the 620 fees dropped by 26% of interest charges, and the 740 score went up 400% when mortgage insurance is included for both?
Nope, just someone who can do math. My point was twofold. First, if Fox News wanted to be over-the-top in their assessments, the real numbers are pretty shocking. They didn't even go that way.
Second, your implication about the overall point being incorrect is flat-out false, just based on the numbers you shared in here. It looks like mortgage rates did go up for people with better credit ratings, and down (sometimes significantly) for people with worse credit ratings.
I guess you could retreat to your motte and just point out that you have no idea what the rationale behind the change was, but in that case, why write the article?
I mean, the numbers speak for themselves on that graph. Note how the new 620 line decreases and the much more-creditworthy 720 line goes up significantly. Isn't that exactly what people are saying the problem is?
Look at the before and after for the 21% down.
780: Roughly unchanged at 0.5%
720: Roughly doubled from .7% to 1.3%
640: Decreased from 3% down to 2.2%
620: Decreased from 3% down to 2.8%
For 19% down it's even more stark
780: 0.4% (weird, lower than with a bigger down payment) to 0.5%
720: 0.6% doubles to 1.3%
640: 4.5% drops to 3.8%
620 5.2% drops to 4.8%
The craziest one is for no money down. Highly creditworthy borrowers came up a little bit (although a big percentage of interest payments); 640 dropped by over 1%, and 620 dropped by almost 2%.
Alternatively you could just respond by going ad hominem and bashing people who read your article and do math, something something Fox News! Rawr!
Well, there are 2 levels here. First is what the appropriate steepness of these fees should be. Whatever the answer to that question would be, it doesn't justify the rhetoric in these articles. But, it turns out these fees didn't exist at all before 2008.
The only reason I wrote the article is because I was seeing and getting messages from several sources that clearly thought that they would now pay lower fees if they had worse credit and lower down payments. I wouldn't have written an article just to debate the slope of the fees. I wrote it because these articles clearly left a wrong impression for a lot of people, and it is a wrong impression that is relevant to my broader project here.
If you think current underwriting standards and mortgage fees favor or subsidize less qualified borrowers, I doubt that there is much for us to discuss.
But your 400% thing was brilliant, and it would have served Fox News' purposes much better than the numbers they used.
Do you have the data on total cost of a new mortgage for these bands, including fees? Do you actually know whether the all-in price of mortgages have gone up for more-creditworthy borrowers and down for less-, or the opposite?
That'd be an article that actually could make the point you're trying to make, if the data supports it.
The leading paragraph there starts by saying two true things, and then finishes by attempting the same conclusion you attempted here, which isn't supported by the evidence.
"The truth is no one with the same credit score will pay more for making a larger downpayment, and no one with the same downpayment will pay more for having a better credit score. Those with good credit scores will not be subsidizing those with worse credit scores."
The first two points are total straw men. Nobody that I've seen, including the articles quoted, have said that people with better credit scores will be paying more than people with worse credit scores, they've been saying that the rate changes make it so that people with better credit scores' payments are increasing, and it's subsidizing people with worse credit scores whose payments are decreasing.
The final point, "those with good credit scores will not be subsidizing those with bad credit scores" is something that you conclusively demonstrated isn't the case, and the way it's presented make it seem like it follows naturally from the first two points, which it definitely doesn't.
Later on we have "[Director Thompson] went on to explain that the GSEs’ Congressional charters “specifically include references to supporting low- and moderate-income families by earning returns on mortgages for these borrowers that may be less than the returns earned on other products.... your editorial fails to recognize that modestly higher fees will enhance Fannie and Freddie’s safety and soundness, which will benefit taxpayers.” Allow me to translate for the people who are mislead by the article's implication that this supports their point.
What Director Thompson said is that she has been directed to subsidize poorer credit scores, but that because she's charging higher credit scores more the net safety will be enhanced.
These two articles, yours and the one you linked, are both full-on gaslighting attempts, trying to tell people "this thing you're upset about isn't happening, and anyways, it's good that it's happening."
I am a retired mortgage loan officer. My recollection is that loan-level price adjustments were first introduced in 2008. In my opinion, they have no relation to actual loan losses and should be abolished.
I think you're right that news orgs misrepresented that you'll be punished for having a greater credit *score*, but if we take size of downpayment as an indicator of creditworthiness, then it is true that there are many situations where the more creditworthy customers pay more than the less creditworthy. For example, the LLPA for the bottom credit score tier, putting no money down, is less than the LLPA for the next two tiers putting 80% down.
I made this graph to help illustrate: https://imgur.com/wgcx9Cg
Is it possible to game this by putting down < 5% at loan origination and then paying off whatever you'd planned to put down originally?
Good catch. It is weird. That's why I included mortgage insurance fees. High LTV loans have to have mortgage insurance, and, I don't know why, but the fee schedules assign more fees via mortgage insurance and less fees in the LLPA. When you add them together, the low down payments generally pay more.
Maybe there's some way, actuarially, to think about it that makes sense. I don't know.
I liked your article, thank you. I’m a loan officer in Milwaukee and I saw many in my network share/post the Fox News screenshot. Very misleading.
In practice, my response has been to help clients get into the “affordable” programs Fannie and Freddie offer, there’s an income based program (not even for first time homebuyer) and also a first time homebuyer program. Based on either <80% area median income and <100% respectively.
Both of those programs have a full waiver of ALL LLPA adjustments. 660 score gets the “same rate as grandma” with a 40% DP and 780 score. I call it “grandma’s rate” with my clients.
It’s not necessarily fair, but they appreciate it. When we as lenders are forced to use a GSE under conservatorship, you don’t really get an option besides to play with the rules we are given.
Interesting.
Thanks!
If there is a national finance market for X, and part of that market is a small fee to offset risks, and the level of the fee rate is based on credit scores, and people with worse credit see a decrease, while people with better credit see an increase, it's necessarily the case that one of two things (or both) is true:
1) the people with better credit are subsidizing people with worst credit
Or
2) rates are inflated and there's enough slush in the system to re-jigger them in all kinds of ways - maybe even some that are political.
I haven't written about this much on substack, but I agree with your intuition here. Basically, after 2008, private mortgage lending with any substantial default risk were regulated to oblivion. FHA and the GSEs have the "Qualified Mortgage" patch, which means that a lot of the new regulatory liabilities for mortgage originators go away if the FHA & GSE underwriters agree to purchase your mortgage. This basically gives the federal agencies a monopoly on much of the mortgage market, and they charge for that monopoly, transferring billions in profits over to the government each year. So, you are right. These fees can be fairly arbitrary, because there is no legal market to arbitrage to a reasonable risk/reward level. If it was legal, private lenders would be making mortgages to borrowers with a moderate amount of default risk at much lower rates and/or fees than what is currently available.
Ok but that's not what you're arguing in your article... The Biden administration made these changes and to a certain extent they are subsidizing the less creditworthy by increasing rates on the creditworthy - aka penalizing them in other words. The GSE underwriters are the housing market for almost all single family homes.
It seems like you wrote this piece to take the air out of headline readers who might think that bad creditors pay less than good creditors. That's not the case, but it totally appears that the good are subsidizing the bad here, and by design.
There is no evidence that these fees net out to a loss on those borrowers that needs to be subsidized. These articles led to an onslaught of reactions that were a gross mischaracterization of the policy. That was the motivation for my post. There is no need to debate the marginal differences in gains and losses various actuarial assumptions might suggest, relative to the new fees. The onslaught of mischaracterizations wasn't about actuarial estimates. It is just one more drop of poison in the well against reasonable mortgage policy, and I felt I should address it.
"There is no evidence" is the last refuge of a destroyed position. It comes right before (or with) indignation that someone **dare** challenge you.
Using indignation as a tell for when you've made a mistake is a good technique for making fewer mistakes, if that's something you're trying to do.
"""I’m pretty sure what they’ve done here is cherry pick the low credit score that had the largest fee decrease. Then, they reported the total fee of a higher credit score. So, a low down payment 620 score has a fee that went from about 6.75% to 5% (when mortgage insurance is included). And, also, the fee for a 740 score went from 0.25% to 1%. (plus a 0.25% mortgage insurance fee). Why didn’t they just say that fees for 740 scores went up 0.75%? It would still get their partisan point across. It would still be weird, because it would be describing mortgages with two different down payments. And it would hide the fact that the 620 score still has a fee that is more than 3% higher than the 740 score. But, at least it wouldn’t be mixing levels with changes.'""
How about for accuracy's sake we use those numbers and report that the 620 fees dropped by 26% of interest charges, and the 740 score went up 400% when mortgage insurance is included for both?
This article is quite the gaslight.
Are you a writer for Fox News?
"Mortgage fees up 400%." is spectacular, and if you show them that and they don't hire you, that's on them.
“Are you a writer for Fox News?”
Reads to me like a guy who understands math. Nice work avoiding showing where his logic is wrong.
His logic is impeccable. His numbers are correct. And I recommended that Fox News hire him.
I’m very confident his brand of logic would never generate the clickbait that the mass media engine of chaos requires for fuel.
Nope, just someone who can do math. My point was twofold. First, if Fox News wanted to be over-the-top in their assessments, the real numbers are pretty shocking. They didn't even go that way.
Second, your implication about the overall point being incorrect is flat-out false, just based on the numbers you shared in here. It looks like mortgage rates did go up for people with better credit ratings, and down (sometimes significantly) for people with worse credit ratings.
I guess you could retreat to your motte and just point out that you have no idea what the rationale behind the change was, but in that case, why write the article?
I mean, the numbers speak for themselves on that graph. Note how the new 620 line decreases and the much more-creditworthy 720 line goes up significantly. Isn't that exactly what people are saying the problem is?
Look at the before and after for the 21% down.
780: Roughly unchanged at 0.5%
720: Roughly doubled from .7% to 1.3%
640: Decreased from 3% down to 2.2%
620: Decreased from 3% down to 2.8%
For 19% down it's even more stark
780: 0.4% (weird, lower than with a bigger down payment) to 0.5%
720: 0.6% doubles to 1.3%
640: 4.5% drops to 3.8%
620 5.2% drops to 4.8%
The craziest one is for no money down. Highly creditworthy borrowers came up a little bit (although a big percentage of interest payments); 640 dropped by over 1%, and 620 dropped by almost 2%.
Alternatively you could just respond by going ad hominem and bashing people who read your article and do math, something something Fox News! Rawr!
That's cool too.
Well, there are 2 levels here. First is what the appropriate steepness of these fees should be. Whatever the answer to that question would be, it doesn't justify the rhetoric in these articles. But, it turns out these fees didn't exist at all before 2008.
The only reason I wrote the article is because I was seeing and getting messages from several sources that clearly thought that they would now pay lower fees if they had worse credit and lower down payments. I wouldn't have written an article just to debate the slope of the fees. I wrote it because these articles clearly left a wrong impression for a lot of people, and it is a wrong impression that is relevant to my broader project here.
If you think current underwriting standards and mortgage fees favor or subsidize less qualified borrowers, I doubt that there is much for us to discuss.
But your 400% thing was brilliant, and it would have served Fox News' purposes much better than the numbers they used.
Do you have the data on total cost of a new mortgage for these bands, including fees? Do you actually know whether the all-in price of mortgages have gone up for more-creditworthy borrowers and down for less-, or the opposite?
That'd be an article that actually could make the point you're trying to make, if the data supports it.
Here's an article about how these fees are in addition to and unrelated to normal risk-based pricing.
https://nhc.org/much-ado-about-nothing-whos-paying-more-for-mortgages/
The leading paragraph there starts by saying two true things, and then finishes by attempting the same conclusion you attempted here, which isn't supported by the evidence.
"The truth is no one with the same credit score will pay more for making a larger downpayment, and no one with the same downpayment will pay more for having a better credit score. Those with good credit scores will not be subsidizing those with worse credit scores."
The first two points are total straw men. Nobody that I've seen, including the articles quoted, have said that people with better credit scores will be paying more than people with worse credit scores, they've been saying that the rate changes make it so that people with better credit scores' payments are increasing, and it's subsidizing people with worse credit scores whose payments are decreasing.
The final point, "those with good credit scores will not be subsidizing those with bad credit scores" is something that you conclusively demonstrated isn't the case, and the way it's presented make it seem like it follows naturally from the first two points, which it definitely doesn't.
Later on we have "[Director Thompson] went on to explain that the GSEs’ Congressional charters “specifically include references to supporting low- and moderate-income families by earning returns on mortgages for these borrowers that may be less than the returns earned on other products.... your editorial fails to recognize that modestly higher fees will enhance Fannie and Freddie’s safety and soundness, which will benefit taxpayers.” Allow me to translate for the people who are mislead by the article's implication that this supports their point.
What Director Thompson said is that she has been directed to subsidize poorer credit scores, but that because she's charging higher credit scores more the net safety will be enhanced.
These two articles, yours and the one you linked, are both full-on gaslighting attempts, trying to tell people "this thing you're upset about isn't happening, and anyways, it's good that it's happening."
Shameful.
I am a retired mortgage loan officer. My recollection is that loan-level price adjustments were first introduced in 2008. In my opinion, they have no relation to actual loan losses and should be abolished.
Thanks for the insight!
I guess it's even worse than I feared.