Transition to February expiration cycle

The changes to the Case Shiller home price index futures that the CME announced on Jan 27 have taken place, and pricing of the new expiration cycle has begun.

The table to the right show the bids and asks for the six February contracts as of last night's close, as well as (estimated) last values and the quotes as a percent of spot values. ^1, ^2 One can see that users are willing to post prices consistent with higher index values over each of the next five years. ^3

As has been the case in the past bid/ask spreads are tightest with the shortest contracts (i.e. the Feb '20 contract only has two weeks to expiration and has a 0.4 point bid/ask spread), and get progressively wider with longer expirations.^4 As yesterday was only my first day in quoting the new contracts, I'd expect longer-dated bid/ask spreads to contract over the next few weeks.

The graph in the middle shows the bids and asks versus the historical Case Shiller HCI (10-city) index (in black). Again, prices are higher in the future, but the slope of forward prices is flatter.

Finally, the graph at the bottom shows the year-over-year ("YOY") changes to the Case Shiller index (in black) versus YOY price differences between mid-market values of two February contracts (green squares).

I'd note four things related to this graph:

1)  The use of Feb vs Feb contracts helps to reduce the influence of seasonal factors.

2)  The use of Feb vs Feb contract comparisons relate to year-end index comparisons. Recall that the Feb Case Shiller contracts settle on the index from two months earlier -in this case the index that covers activity through Dec 31. My hope is that in switching to a focus on February contracts, that these futures can be used to observe, or participate in trading related to,  year-end vs. year-end comparisons. That is, a comparison of Feb '21 vs Feb '20 contracts may be useful for expressing a view on changes in the Case Shiller index for 2020.

3) The use of mid-market values (as the point of comparison) is an approach I've used for years as IMHO it reduces the impact of one trade on closes. For example, if someone bought the Feb '24 contract today at 245.8, the close on that contract would be 245.8, but the Feb '23 close would still be 237.8, and the Feb '25 contract close would still be 240.2. Further, changes to either bids or offers will impact mid-market values, while a lower offer (which I think has informational value) will not change the close. While this approach may not differ much from using closes on shorter contracts with tight bid/ask spreads, it often results in more frequent changes (than using closes) for infrequently traded, longer-dated contracts.

4) YOY differences on forward contracts are lower than many forecasts (albeit using different indices). As I've noted before, I suspect that clearing levels on longer-dated contracts are lower than expectations, given an imbalance between the many natural sellers (people looking to hedge) and the smaller number of natural longs. As such, discounted bids (vs. expectations) might pull down YOY percentage differences. ^5

I'm eager to facilitate activity in the Feb expiration cycles (particularly the Feb '21, '23 and '25 contracts) as trading tends to reduce bid/ask spreads, leading to greater interest from those who have not started trading. Further, tighter markets in key benchmark contracts (again '21, '23 and '25) can be used to help populate quotes on the ten regional contracts via Intercity ("IC") markets. I'll be tweaking IC quotes on these expirations today and hope to have a blog showing bid/ask spreads on regional contracts (and regional replicas of these illustrations) in the very near future.

Please feel free to contact me if you have any questions on this blog, or any issues related to home price index derivatives.

Thanks, John

^1 There are four other expirations (K20, Q20, X20 and X22) that can also be traded, but for this blog I'm highlighting the February cycle.

^2 Note that the CME has not updated Last values for the new contracts, so I've estimated "Last" using the CME closing price logic of "Last trade, or a higher bid, or lower offer".

^3 Note that many of the bids and offers for G22-G25 are a function of calendar spreads vs. the Feb '21 contract. In time, I'd expect some of these contracts to see outright bids and offers.

^4 Feb '20 (0.4), Feb '21 (1.2), Feb '22 (3.0), Feb '24 (6.8), and Feb '25 (9.2)

^5 Shorter-dated contract prices should converge toward expectations as contracts approach maturity, so i'd predict that bid discounts are higher on longer-dated contracts.