CME market prices climbed after Tuesday's update to the Case Shiller #'s. As noted in the table below, mid-market levels rose by 3-8 points (with the exception of CHI which was about unchanged).
My sense is that there are two main, personal reasons for this bounce back.
1) I tend to price futures expiring beyond six months, conservatively versus expectations as (until recently) most of the inquiries I get have been from sellers. That is, if I were to buy contracts, there'd be little hope of unloading them in the near future. However, as the Feb 2023 contract gets closer to expiration, convergence begins to have a stronger pull. That is, the index covers home price sales from Oct-Dec, and as it's already near month-end October, not only has 1/3rd of the calendar for settlement values elapsed, but the sales for November are likely in progress. Net, I've changed my approach (for Feb 2023) from worrying about the unknown, to pricing more toward expectations of where the contract will likely settle. Index values will fall, but projecting the next 8 weeks of home price activity becomes more important than juggling inquiries (for the Feb '23 contracts).
2) The second component is a combination of a renewed appreciation for the decline in volume in home sales heading into the winter, combined with the decline in volume that arises from recent rise to higher mortgage rates.
Recall that the Case Shiller index is a average of the last three months of indices weighted by the activity for each month. Case Shiller publishes "pair counts" but not transaction dollar volume, so I can't see monthly weights, but have been using pair count as a proxy.^1 (This assumes that the mix of home sales is about the same from month to month, which qualifies any observations. )
The reported three month rolling pair-counts have been collapsing (see graph) to the lowest levels in years, which implies that the monthly pair-counts have been falling even faster.^2
The following table shows an example of the how impact of declining volume "might" (air quotes) play out. Note that (for this month) my projection of a lower August pair-count underweighted the August value relative to June and July, dragging the reported index value up slightly. (I'd estimate about one point from what a more equally weighted set of monthly indices would generate).
However, going into the winter, the heavier June and then July weights will roll off, leading to some slight downward pressure on year-end index values (again only about 1-2 points).
There are other implications of lower home price turnover that I am happy to discuss, and/or will blog about in the near future.
Feel free to contact me if you have any questions related to this blog, would care to discuss any of the ideas raised, and/or if you have any trading ideas that you'd like to share.
Thanks,
John
^1 Other entities with access to raw home price sales data should be better positioned to estimated indices.
^2 Given that each month's index is comprised of three weights and three monthly index values, it's impossible to mathematically back into monthly values of pair-count and index, but I try to estimate reasonable estimates.