From MarketWatch:
Home prices continued ascent in September despite soaring mortgage rates: Case-Shiller
Home prices showed no signs of slowing in September despite the record-high mortgage rates that rendered housing unaffordable for many Americans, according to the latest S&P CoreLogic Case-Shiller Indices report.
Nationwide, home prices rose by 0.3% in September and now stand 3.9% above its year-ago level. The 10-city composite gained 4.8% and the 20-city composite increased 3.9% – the indices measure home prices in major metros across the country. Both indices posted a 0.7% month-over-month increase in September.
Home prices now stand 6.6% above where they started the year despite rising mortgage rates. The September Case-Shiller tracks July, August and September when mortgage rates climbed steadily from 6.8% at the beginning of July to 7.3% by the end of September. At the same time, housing inventory has remained low. Existing home sales dropped 2% in September to a 13-year low, according to research from Realtor.com.
“Speeding up of annual home price growth reflects much of the pent-up demand that exists in the housing market amid very low inventories,” CoreLogic Chief Economist Dr. Selma Hepp said in a statement. “Nevertheless, home prices are feeling the weight of high mortgage rates, which will slow the rate of price growth in the coming months. Still, despite the dramatic increase in the cost of homeownership, home prices have risen 6.4% this year – meaningfully beyond expectations given the rise in borrowing costs.”
…
Detroit, San Diego and New York led the way for the fastest-growing cities in the U.S. Detroit reported the highest year-over-year growth, with an annual increase of 6.7%. San Diego and New York followed with gains of 6.5% and 6.3%, respectively.
September’s worst-performing cities were Las Vegas, Phoenix and Portland. Las Vegas saw the most significant year-over-year decline, with prices dipping 1.9%. Phoenix and Portland followed with a decrease in growth of 1.2% and 0.7%, respectively.
Foist
Second!
Government lenders Fannie Mae and Freddie Mac to offer home loans over $1 million in MORE US counties as house prices reach record highs
Oh baby! Come to Papa!
Eddie….not just any million dollar homes.. It’s truly different here.
Don’t buy a house in Burlington or Mercer County loan limit there is only $766,550.00 but just across the county line it’s $1,149,825.00
Zoomable map below. Most of Flyover country is the lower conforming loan limit.
https://www.fhfa.gov/DataTools/Tools/Pages/Conforming-Loan-Limit-Map.aspx
Yell at schiff, what are you yelling at me for?
leftwing says:
November 29, 2023 at 10:19 pm
Since you seemed to miss the the two points how about starting by getting the data right then asking if it actually matters in the way the author implies before caring how and if it continues…
Libturd says:
November 29, 2023 at 11:50 pm
Look at Japan’s debt to GDP numbnuts.
Juice: Shane Mc Gowan passed away.
Detroit, San Diego and New York led the way for the fastest-growing cities in the U.S.
Detroit –> opportunity. Whole blocks were razed, the riff raff was pushed out. Now is the time to rebuild and assemble a productive enclave of hard-working folks. Michigan may even turn red consistently. San Diego has 80 degree days from here to eternity. I think the muppets have already missed the opportunity on that locale. And then there’s the NYC sphere; a freight train of ever-increasing prices forever. What’s that phrase about the playa and the game? We joke about the million dollar cape in Saddle Brook but it’s now closer than you think. Punkin Noggin was right a few years ago; buy now or be priced out forever. What until the FED drops the hammer and says, “We’re done.” A blip into the “6” range is going to result in needing to make reservations to stand on line at open houses. I wouldn’t be surprised if there’s a bouncer at the door with a velvet rope picking out the prettiest muppets. “But… but… I never got my wristband!”
Don’t buy a house in Burlington or Mercer County loan limit there is only $766,550.00 but just across the county line it’s $1,149,825.00
Might be the only thing I agree with. Mercer is definitely an odd county. West Windsor, Princeton, Hopewell, Pennington, all towns people try to get into for their prized school systems. But unlike Morris County & Basking Ridge Area, all the homes are complete trash. Old and falling apart. No gas, no sewer. People north of Mercer County at least kept their homes in order. I think the entire Mercer Co. area was populated by grandma and grandpa and they let their homes deteriorate prior to leaving it to their children to flip.
UPDATED THU, NOV 30 2023AT 9:47 EST
Dow jumps more than 180 points, approaches new 2023 high after more cooling inflation data: Live updates
Are you turning Democrat? Seems like Biden is making you happy!
Fast Eddie says:
November 30, 2023 at 8:29 am
Government lenders Fannie Mae and Freddie Mac to offer home loans over $1 million in MORE US counties as house prices reach record highs
Oh baby! Come to Papa!
Juice Box says:
Detroit –> opportunity. Whole blocks were razed, the riff raff was pushed out. Now is the time to rebuild and assemble a productive enclave of hard-working folks.
They are coming as you asked. Thousands of them across the border trying to make their way north.
Maybe give them a 1M mortgage and some free college tuition and they will turn Detroit into the next Silicon Valley.
Are you turning Democrat? Seems like Biden is making you happy!
If they start sending me checks, I’ll push slo joe’s wheelchair personally.
3b – Heard last night from a family member on Facebook.
https://www.youtube.com/watch?v=j9jbdgZidu8
Grim – AI Prompts question? Did you build in a tokenization service that encrypts or sanitizes the data for PII etc?
User Query —> Constructed Prompt–> Token Service—>LLM—->Token Service—>Answer
Also similar question but for multilingual support?
790, and I know which one I got wrong.
leftwing says:
November 29, 2023 at 9:41 pm
*770* oops.
690
When in doubt, pick C out.
Don’t ask about my verbal score. Please!
For preprocessing user-provided query to remove accidental/inadvertent inclusion of PII, check out Microsoft’s Presidio or Azure PII redaction (if on MS). If on AWS, there are options using Comprehend. If you are in the huggingface ecosystem, there are some interesting variations on spacy/distilibert for PII (https://huggingface.co/beki/en_spacy_pii_distilibert).
Some of these are way fancier than simplistic regex pattern matching or old-school table matching (for name identification). They are basically using simplistic LLM/AI for entity recognition, taking into account the context in which the PII might be being used.
I’m not sure there is yet the LLM equivalent of the OWASP Top 10/SANS 25 for prompt sanitization, but it very much feels like we’re in the equivalent of late 90s web development, when text fields on websites presented a realistic vector for hacking (ala sql injection, etc).
I love this forum when heavy duty X’s and O’s are discussed. As was then and is now, I was the troublemaker in the back disrupting the class on occasion. However, I must have did enough to survive because I was in Group 1 from Grades 1 through 8 and managed to graduate from a prep high school. That said, after reading Grim’s message above, my only reaction was: “What’s the Frequency, Kenneth!”
On the multilingual, the quality of the translations are going to be highly model dependent – mixed bag – from good to worthless. We’ve only ever played around with passing the translation request into the prompt programmatically to match the end-user locale (‘provide the answer in Spanish’).
While some models do an OK job at matching the input language to the output language (if prompted in Spanish, they respond in Spanish), this is not at all a guarantee. We’ve seen inconsistent output where output gets generated in English occasionally, especially if using mixed-language in the prompts (mixing in an English word in an otherwise Spanish prompt). Explicitly prompting for the output language yields better results, but even that is not 100%.
You need to let go of any expectations of deterministic behavior in this ecosystem. Run your code 10 times, you’ll very often get 10 different results. Large scale Q&A testing of these systems is going to be a horrorshow. This is not an ecosystem where you can automate test cases easily.
Juice: Fairytale of NY is a classic. Rainy Night in Soho is another classics. Shane was a mess, but an incredible song writer.
Grim – Langchain Framework for our API, LLM Azure for now. Goal is to build it for multi LLM support but as you already know no common prompt dialect yet so switching LLMs will require work etc. I see support for lots of what we have discussed Presidio looks promising.
I am only focused on ring fencing it and limiting damage, as there is no stopping the AI train…
As far as “You need to let go of any expectations”. My only expectation is change, won’t be in the same place in a year for sure. Hyrum’s Law etc.
chi, nice, lol
720 verbal, 680 math. Back in the 80s when if you were lucky you got one practice test, and one real test and that was it. No SAT test prep industry in FL back in 1986. Surprisingly a 1400 back then was good enough to get 99th percentile and an invitation to join Mensa.
There were some kids in our AP classes that typically got better grades than me who I knew were devastated by their scores in the 1100s or 1200s.
And I didn’t even get any actual use out of the 1400 SAT in college applications because my GPA was weak. Preferred reading on my own to doing homework. My parents took zero interest in my education, and I was naïve. So I never actually ended up applying to any colleges anyway, I just showed up at the nearest state run college the week before classes were supposed to start, because otherwise I’d have to start delivering pizzas or something like the guys I played D&D with.
What I wouldn’t give for some prompt sanitization!
Also. When did “appreciate you” enter the vernacular? Perhaps it was brought to us by the team that popularized, “We Open!”
Lib – I am fresh out of re-education aka work training, so I’ll explain.
Work encourages (mandates) we tell people we appreciate them.
We even have an app for it where we say why publicly and give them points for rewards to cash in for gift cards, tickets, streaming services etc. I have nearly a years worth of points to give out…I don’t know wtf I am gonna do and the HR Gestapo will come for me if I don’t.
Juice. We have it too. It’s called Kudos. For some reason, my team get hundreds of Kudos and I have received like two. HR is unrelenting in their pressure for us to write them. I am so tempted to start giving them out for the most menial of tasks. Maybe even lob some backhanded kudos their way.
“Christopher punched in on time for three consecutive days. Kudos!”
“Let’s hear it for Tom, who remembered to courtesy flush while using the 12th floor men’s room. Whoop whoop!”
“Praise Mary. Thanks to her, the delicious aroma of her fish sandwich will be remembered every time someone uses the break lounge microwave for weeks to come.”
1260, but I was in the sixth grade. 700+ verbal, 5 something math, but I hadn’t been exposed to most of it yet.
Midwest we used the ACT. I got a 35, out of 36. My joke has always been I aced it, but I misspelled my name. I always tested well.
My grades, not so much. But in my defense I only went to school half of a day for the most part the last two years, and I was usually stoned. Took my AP requirements and bailed. We mutually decided it was best if I left early in the day and went to take college classes. Kept me out of more trouble.
I always signed up for a college class for every afternoon, turned in my schedule, then dropped the MWF class. Either worked or played golf instead those days.
Hughesrep,
You must have become a good golfer within a couple years.
Sounds like your school wasn’t challenging and engaging you adequately.
I just dumped 100 points on a new team member for clarify a CI/CD issue, only 900 more to go and I can avoid the reeducation camp aka class on recognition.
Peculiar mixed market ttoday.
Dow up 1.2% and Nasdaq down .5%, but it was closer to 1% a while ago. Sometimes one heavily weighted stock moves an index, but I’m not seeing it.
Perhaps it was brought to us by the team that popularized, “We Open!”
Everyone Must Eat!
I always signed up for a college class for every afternoon, turned in my schedule, then dropped the MWF class.
Damn, guy! You had a married, white female class? Were you like, exhausted by the end of the day?
Stu – must be something in the New Jersey air and water. Coworkers son had a seizure, scan showed a mass so they had to go in and remove. I offered to help in anyway, even if it is just walking the dog.
Let’s play R hypocrisy wheel of fortune. Spin the wheel, land on:
1. Dumpy says he is ‘very honored’ to have support of BLM activist
2. DeSantis ally and Florida GOP Chair Christian Ziegler, husband of Moms For Liberty cofounder Bridget Ziegler (Casey D’s best friend), is being accused of sexual battery by a woman who alleges that the three of them were involved in a longstanding consensual three-way sexual relationship and that they may have illegally taped sex acts??
3. US Oil production at record highs, stock market at record highs, unemployment at record lows, inflation down to 3%, GDP revised up..but R hypocrites paint BS ‘scorched world’ view that would make the TV show Silo jealous.
Juice,
If they need any REAL advice, let me know. Gator and I do a lot counseling for families entering the fray. It’s good to have a survivor at home.
Juice,
Hope it’s benign. I have been in hundreds of those procedures. Location and type makes all the difference.