Over the last two trading days, it appears that the CME Case Shiller home price index futures finally got around to participating in the stock market selloff. Since Friday, prices in intermediate expiration contracts (e.g. Feb '21-'23) dropped by over 10 points. a price move without precedence in the last eight years.
The attached graph shows the historical Case Shiller 10-city index in black^1, the closes on the eight CME contracts from Feb 21 in green ^2, the closes from Friday April 3rd in red, and the closing bids (blue x's), offers (red pluses), and closes (in purple) from Tuesday April 7th. Note that since closing prices are defined by last trade, a lower offer, or a better bid, in this infrequently traded, falling market, the last offer seems to define most closes.^3 As an example of the sell-off, the HCIG21 contract closed Tuesday (April 7th) quoted 213.0/217.0 (bid/ask), down from 234.2/234.6 on Jan. 31st. The drop brings the benchmark Feb '21 contract to being quoted at levels consistent with a negative 6-8% HPA for 2020, versus +1-2% implied HPA gains in January.^4
While these markets had been drifting lower through late March (as illustrated from the differences between the green and red lines), this week saw pronounced inversion in first three expirations (May '20, Aug '20 and Nov '20) (as illustrated in the drop in closes from the red line to the purple line). For example, last week selected Aug and Nov contracts traded at premiums to May. Such pricing might have been consistent with positive seasonal factors and a sense that home prices might be slow to react. Trading this week flipped Aug and Nov contracts to a discount to May, consistent with a view that home prices would fall in price this summer, as unemployment soars.
I've only tallied 12 trades in April, and all have been "bid hit" so as before, treat quotes (particularly longer-expiration quotes) cautiously.
I expect some to ask "what happened", or "why this week", particularly since the S&P futures were up 1,000 points at one point. I can't speak for others, but I may have gotten complacent, interpreting lack of selling ( in a very thinly traded contract) as lack of reasons to sell. I had no interest to offer contracts lower (in effect trading against myself), and had seen buyers of Aug/Nov at a premium to May just the week before. Further, I thought that increasing margins at IB was going to be the lead Case Shiller trading story. New eyes, and/or new research over the weekend by current players, have decidedly changed the tone of these markets- probably bringing shorter expiration prices more in line with the macro-economic themes driving other markets.
One of the challenges (opportunities?) in trading the Case Shiller futures is that the contracts provide numerous ways to play different themes. That is, there's not just one price for any region (much less across regions), but multiple expirations^5. Also, unlike stocks where one can be a long-term bull without a specific time frame, here users need to have a view as to when a price will be reached by a specific point in time.
Given the multiple target time frames, here's a few themes I'd tee up for such debate and your feedback:
* While the HCIG21 contract closed today at levels consistent with negative 6-8% HPA for 2020, closes on the regional contracts ranged from -3-4% HPA (BOS) to -10-15% HPA (LAV). Intercity (IC) contracts might be a good way to express relative value views. For example, someone thinking the BOS contract will outperform the 10-city, or that the LAV contract will fall more (or less) than 10% than the HCI contract, might consider using IC spreads to express relative value opinions, while taking less outright risk. ^6 I tend to maintain IC quotes on Feb '21 and Feb '23 contracts if someone is looking to express a relative value.^7
* This week's trading has shown that the front K20 contracts were not immune. Even though the May release will cover activity from January to March, and even though two month's activity (Jan-Feb) was "in the books" by March 1st, May contract bids got hit at reasonable discounts to spot values. This suggests that users see the stock market selloff as having a meaningful impact on home price closes during March. As such unlike past expiration cycles when front contracts traded in a tight range in the run-up to expiration, I expect there to be a good amount of activity in the May contracts, as this expiration approaches, as users digest mortgage closing information, and the potential impact on May index releases.
* While the May contract might have only one month of home closings that will impact index calculations, the Aug '20, Nov '20 and Feb '21 contracts will likely bear the full brunt of the COVID recession. The Aug and Nov contracts have historically had positive seasonal support, and this year, all home closes for these indices would be during the run-up to the Election. However, with this week's price moves, those factors now seem small relative to the potential impact of a recession.
*Finally, for those who want to express views on what a post-COVID home price market might look like, the longer-dated contracts (e.g. Feb '23-'25) might be far enough into the future to think about pricing in a "new normal" economy. Those with an interest in co-investment programs might consider clearing levels on the longest-dated contracts for color on those regions. (more in a future blog). (Again, there's been no trading, so treat quotes as an invitation for users to post me a trading axe that I can share with the community.) If users want to express views about the dip into X20/G21 and/or possible rebound from G23 to G25, calendar spreads might be an interesting approach, and will probably trade on narrower spreads than outright markets.
Net, there's plenty to debate, and numerous markets where both outright and hundreds of relative values themes can be expressed to answer the question "Where are home prices headed?" ^5
Please feel free to contact me, if you'd like to discuss this blog, if you have a trade proposal you'd like to share, or if you have questions on any aspect of hedging using home price index derivatives.
Thanks, John
^1 Tables and graphs for other regions available on request
^2 There are ten expirations per CME contract buy I'm only showing what I view to be the eight more likely to trade.
^3 This should further illustrate the hazards of relying on closes for historical analysis on such thinly traded contracts.
^4 I caution first-time readers that the expression "is consistent with" is not to say that these levels reflect expectations. The contracts represent prices where home price index risk can be exchanged. Contracts do converge with index levels over time, but that is most pronounced as expiration approaches. Risk-clearing levels seem to increase (i.e. clear at a deeper discount to expectations) when inquiries shift to more one-sided selling.
^5 There are 110 contracts (11 regions * 10 expirations). The number of contracts, and potential for dispersion of interest, is why I keep harping on the need for users to focus trading on one benchmark contract (Feb'21) and then letting many other contracts trade as derivatives of that via calendar and intercity spread trades.
^6 See my March 20th blog for a more detailed review of IC spread trading.
^7 I am willing to quote IC spreads on other expirations.