Reality, Or Lack Thereof

Something is indeed rotten in the state of our perceived reality, and the cracks are getting harder and harder to ignore. In a recent post, I shared the oddities I had found by reviewing historical results for median listing price, median new home price, median existing home price and Redfin median price side-by-side. There were patterns and repeating numbers that did not make a lot of logical sense. As I discussed in that post it is hard for even me to contemplate that there is such a thing as a “coordinated” housing cartel, but after recent results from National Association of Realtors (NAR) I had to question yet again whether I was a useful idiot for placing (any) faith in the numbers as reported by the largest lobbying firm by spend in DC. I’ve shared with you before that NAR had to restate historical prices after CoreLogic (who uses public record) called them out post the last housing bubble. The builders (or Census) this time around have already revised down their median price peak from 2022 by -7.35% or $30K. July’s reported median price is actually -18.72% below the previously reported peak and -12.27% below the restated peak.

Those reported prices do not include of course the kitchen sink the builders are throwing in to get any bites at all. Chew on that for a moment. Is anyone talking about this?

As previously discussed, I’m a numbers gal. Despite being called Ms. Wrong in a 2nd grade math class, I love math and can calculate things in my head that make the kiddo really mad. By the by, she is also great at math and will surpass me soon, but she really hates it when I figure out the answer before her. I also love looking for patterns. As many of you know I spend a lot of time each month diving into these figures. Although I’m a techie when it comes to certain things, I am not when it comes to reviewing results. I don’t let AI summarize or pre-populate my models. I look at all the detail. So, you can imagine my surprise when I saw this from NAR’s single-family schedule:

I posted the table on xTwitter to see if anyone had any thoughts. One follower replied that it looks like July 2025 is an estimate, and I would agree. Except they never use the word estimate in their press releases. Because I’m a freak, I just checked this schedule one more time on the EHS site to be sure I’m not crazy. One of the reasons I dug in this week was because I did not understand how NAR’s seasonal adjustment made any sense this month. I am not the only one who was surprised. My forecast, which is a mathematical calculation using historical correlation to Redfin results (who supposedly uses MLS…more on that later), came in right on the money of NAR’s non-seasonally adjusted results. NAR claiming seasonally adjusted sales were inching up and improving due to better affordability seemed bizarre based on how long it takes to close a sale. Recent, slight improvements in mortgage rates would not show up materially yet.

Remember when the character in the Big Short introduces his quant? Yun has been around for a while and there are folks that track, like they do Jim Cramer, just how wrong he has been over the years. Was Yun saying what NAR hopes to happen? If what these folks were shilling wasn’t so dangerous, you would feel sorry for them, trapped behind the glass, dancing their little dance.

As if this nugget was not enough, I told you folks in my previous post about these cowinkidinks that I would be watching more carefully the initial results because the revisions can really murky the waters when you try to find these patterns historically…except for when they get lazy. Remember when I found that Realtor (who provides the data to FRED) revised that median listing price in the Spring to ensure that we would not see a YOY decline in June, the typical month prices peak? To be clear the listing price is theoretically the median of all list prices (not sold) in the MLS systems to which Realtor has access. So, if you list your home this month for $375K, no matter what you sell it for, that is the list price in August. The FRED site did not include a note or any explanation about the revised data. Guess what I found this month? For those that had better things to do and missed it, NAR reported on August 21, 2025 that the existing median home price increased a paltry .24% YOY in July. I have to admit being a little surprised they went so low on the YOY comparison because it seemed like they were capitulating faster than I anticipated after going whole hog in June with that ridiculous reported price (which they revised down). Guess how much the median listing price increased YOY for June (published on the FRED site on July 31, 2025)? .24%. It’s almost as if someone used the YOY increase on the listing price in June to estimate YOY increase for the existing home price for July YOY results.

Is your head spinning? Having worked in Finance I know a little something about forecasting and assumptions. If you don’t have real-time data, you often use prior month or prior year results or assumptions. Are we looking at estimates and bad ones at that? A few months ago, I noticed that Redfin reports more sales than both the Census and NAR combined. For example, Redfin reported 471,958 home sales in July. NAR reported 388,000 and the Census reported 56,000 for a total of 440,000 home sales. Now the Census is reporting contracts and not actual sales, and there are often cancellations that do not get tracked or reported by the Census. Industry analysts put the historical average around 12%. It typically takes 30-60 days to close on a home, so let’s use May or June new home sales to calculate. Luckily, new home sales were the same for both months (57K), so our unadjusted total for combined sales for July is 445,000 using either May or June. If we assume a cancellation rate of 12% for new home sales, then the May or June adjusted sales figure is 50,160. Our combined total is now 438,160 or 33,798 fewer sales than what Redfin reported in July.

When I first did this analysis some months ago it made perfect sense to me because Redfin used to report sales from county records, and I assumed they were capturing private sales that don’t get logged in MLS. Like Case Shiller they use county records for their home price index. However, when you look at their current methodology, they state they are using MLS and do not mention county records for these metrics (that I can find). So, I emailed them and asked how they were getting more sales than NAR or Census…were they using recorded sales to augment the MLS data? Back when I was developing an underwriting tool for a private lender, that was my methodology for these metrics. Any data provider can buy this information from a MLS aggregator. If this was the methodology, I was kind of excited that we might be able to size the private market. No response.

I apologize if this is confusing or even boring, but if nothing else, I hope this analysis shows you that if we think we are watching what is actually happening when these entities report versus what the industry wants to show us, then we are sorely mistaken. I’ve told you how the builders get to tell us in the Census survey what the median price is and are never fact-checked. I believe in essence NAR is doing the same. When I posted the table above on xTwitter, many, many realtors commented that yes, NAR was fudging or “deriving” the numbers, and they didn’t match their local MLS. I’ve had another follower show me historical numbers in their local MLS and how they get revised over time so that new “all-time highs” can be announced each week. There are over 500 MLS across the country with their local rules and vested interests. Can you imagine just how garbage this data is even IF it is actually getting rolled up?

Wow, I just got really tired.

If you have been paying attention, which takes a lot of effort and time that most don’t have, we are learning how many stories we believed about our history and recent past are just not true. To effectively pay attention it means reading something other than your preferred newspaper or listening to independent journalists who are committed to finding the truth even if most folks have moved on. Each day brings yet another dizzying array of headlines. Ignore them. If you are interested in a subject, read books, or find an independent journalist you trust because you’ve fact-checked them.

If you are curious about your housing market, then use the things I’ve taught you, including your own eyes to check it out yourself. Don’t just go to that open house…drive the entire neighborhood on the way and stop in a local coffee shop to ask pointed questions. Before you leave for that open house, make sure you have checked New Home Source for new build communities nearby and check Realtor for both listings for rent and for sale. We know that even these tools are limited so we have to triangulate the data and use our eyes to confirm or debunk.

For today’s post, I will share my deep dive into July results for both new and existing home sales and prices. It’s actually fascinating to watch the themes I discussed play out. For instance, I cannot tell you how many people told me cities in CA would never see price declines. In NAR’s results both the South and West YOY median price came in negative for single-family homes.

Melody, you just told us we can’t rely on NAR’s data! Well, I believe that there is only so long this cartel, inclusive of your local realtor and loan officer, can prevail before what we are seeing in front of our very own eyes forces their hand. I believe we have reached that point.

But, but what about rate cuts? And what about Buffet’s recent bet on the builders?

Let’s begin…