May 22, 2021 will mark the 15th anniversary of the first trades in the CME S&P Case Shiller home price index futures.[1] To commemorate this event, I've pulled together a broad overview of activity and key trading themes during the life of the contract [2]. My intention is to dive into greater detail (on upcoming blogs) on six key phases of market activity, and to review lessons learned. As such, this blog will touch on the 15-year history only at a very high level.
But first, thanks are due to many parties for keeping this contract running, to include: Bob Shiller and Chip Case for the idea, CME for the hosting the contract (despite the years of low volume) and the team of people there who have supported my market-making efforts, S&P (through various corporate iterations) for being the index manager, Fiserv and Core Logic for their roles as calculation agents, and various early market makers who have been generous in sharing how the contracts worked in the first few years.
The graph below shows the Case Shiller 10-city index (in red) and closing prices for the last year of each of 16 selected contract expirations.[3] The graph highlights a key benefit in having a futures market in that contract prices (in being forward-looking) have tended to lead movements in the Case Shiller indices, which are measured using a lagged moving average, and therefore backward-looking. For example, contract prices were far below index values for much of 2007-2008, were generally flat from 2009-2011 (when there was limited market-making support) [4] and then futures prices moved above spot for 2013-2018 (and for many regions above the highest prices seen in 2007).
In the last two years, prices have been more volatile, expressing both concerns about the potential impact of Covid[5], as well as reflecting recent market enthusiasm (frothiness?) while highlighting the need for a home price hedge. As Bob Shiller alluded to in a recent email, supply may be the solution to rising home prices, but will builders do so if they can't hedge forward production?
In all cases the contracts and indices converged as time to expiration shortened.
At inception, the longest contract was for one year, but by late 2009, contracts were available extending out to five-years. [6]
While for many years the longest contracts were November expirations (referencing September indices), in Jan. 2020, the CME switched to having February contracts (referencing year-end values) as the longest expirations. I believe that this alignment of benchmark contracts with year-end index values should bring more attention to the CME futures as forecasters make year-end calls.
Options were available for the first three years, and traded in large volumes, were re-introduced in 2015, but then closed in Jan. 2020. [7]
Over the past five years, I've typically posted two-sided GTC live quotes on the first four expirations for each of the ten regional contracts (plus 1-2 selected longer-dated contracts), as well a full set of quotes on all eleven 10-city index contracts. [8] (Here's a table of quotes from May 19, as well as a graph showing the change in contract prices since Dec 2020.
While several hundred futures traded each month in the first two years, annual volume (across all regions) has averaged ~175 contracts since 2010. (Open interest has also dropped below 100 lots in 2014 and is today at record lows).
With low volume, some of the contract users I've done trades with have been retail speculators, and builders (or flippers) looking to hedge. [9]
In addition to direct users, I’m aware that the CME contract prices have been used by several entities (as a form of public good) and have been cited by academics in policy analysis. For example, HUD cited contract prices to show progress in housing recovery during 2009-2010. Users looked to contract prices to see if changes to state and local taxes (SALT) impacted forward values. I have participated with shared-equity appreciation programs managers who have compared forward prices vs. possible returns to users. Finally, mortgage servicers have looked to quotes to provide estimates of movements in home prices (that might impact refi volume).
In addition to this US market, the CME contracts have served as a basis for discussion of home price hedging in other countries. Euronext launched a contract on Paris home prices. Further, I've been involved in discussions related to hedging home prices in Canada, Hong Kong and New Zealand. (Given the volatility in Canadian home prices, I see that as the best opportunity for either OTC agreements, or an exchange).
Finally, no recap would be complete without addressing the elephant in the room- why is volume so low. Over the years, I've received feedback that I fit into five general areas. I'll try to address in a future, more detailed blog, but in sum:
1) There remains a lack of awareness that the contracts exist,
2) The lack of a deep market with large open interest has dissuaded institutional investors from participating.[10] (The classic chicken and egg syndrome).
3) Retail users have expressed concerns about the basis risk between the value of an individual property, and the large geographic areas that are needed to support a repeat-sales index.
4) The contracts only reference ten large cities, and
5) Options have a more conservative risk/reward profile.
I've tried to address each of these points over time.
On the first one, I blog 2-3x/ month from www.homepricefutures.com, tweet from @HomePriceHedger, and have started posting to Facebook and LinkedIn. In past years I've spoken at universities and trade meetings, and in this age of Zoom, am happy to jump on a call with any reporter, or interested user. I wish that the press (and Wall Street research) would give what forward markets are "saying" 1% of the attention they give to historical analysis and opinions. If they think that the numbers are wrong, please weigh in with bids/offers or feedback. Net, please let others know of the existence of these contracts, and the concept of hedging using home price indices.
My belief is that the second point will be resolved when organizations appreciate that they would benefit in having a deep market where forward home price exposure risk can be transferred, but that it will require them to take baby steps (to include someone "working" their orders), to help build it. I'm eager to try and match larger interested sellers and buyers (starting with a focus on the 10-city contracts) and see the benefits to both 100-lot buyers and sellers of hedging exposures on an exchange. This is the public good argument I noted above, and a reason that I've kept at this for so many years.
For point four, I'd love to see the CME trade off the current 11 expirations for each regional contract (to 5-6) while doubling the number of cities referenced. (With Zillow and others buying individual properties -based on improved ability to analyze Big Data, I see no reason that an EBay-style of market with quotes on 100 cities, won't evolve over time. While this might also eventually address point 3, I would argue that today, hedgers have no publicly traded alternative to CME Case Shiller futures, if they are worried about large moves (versus expectations).
Further,, to address points 3, 4, and 5, I created Home Price Hedging Fund (HPHF) as a template to allow users to enter OTC agreements (to include options) on any of 50 cities. My goal would be to broker exposures but I am willing to take some small exposures to "prime the pump". While I believe that options trading may be a better fit for retail investors (i.e. the downside -the premium they pay -is known upfront), an exchange might be overwhelmed by too many combinations of strikes, expirations, puts, calls and combination trades. However, some very limited, standardized set of benchmark options might make sense (once it has been demonstrated that there is demand on both sides).
In sum, while trading has been low, with the roll-out of new mortgage programs (many of which involve a forward valuation of home prices), as well as concerns about a second bubble, IMHO the CME Case Shiller futures remains the best, public, granular way to express a view on changes in forward home price expectations.
Again, there will be future blogs going into the 15 year history of these contracts in more detail. Feel free to contact me if you have questions about this recap, have ideas for future blogs, or just want to learn more about the use of home price index derivatives in hedging strategies.
Thanks,
John
[1] I have been the market maker in these contracts for the last 10+ years. See www.homepricefutures.comfor blogs and other resources related to these contracts. To the extent that this recap has any personal observations those are mine, and not the views of the CME.
[2] Feedback on areas to cover would be much appreciated
[3]Only the last year of each contract is shown.
[4] I would caution any researcher from putting too much weight on closing prices from 2009-2011 as with so few trades, and lack of updates to bids and offers, closing prices tended to be stale.
[5] The decline in 2020 was exacerbated by a 10x increase in margin requirements from a key FCM.
[6]Today, the longest expiration is Feb 2026 (although I try to focus longer-dated interest (e.g. from 2024-2026) to the 2025 contracts. (The contracts might suffer from too many expirations diffusing interest).
[7]There was much less trading in option during 2015-2019 and the process for calculating ~1,000 possible combinations of strikes, expirations, regions, and puts and calls, was daunting.
[8]Most quotes are 1x1 (one lot bid vs one lot offered), but trades for larger amounts or at inside quoted levels are available.
[9] For both builders and home price flippers, options were their preferred hedging product, as they preferred having a finite value at risk.
[10]In my experience, risk managers won’t allow participation in a contract where a) the user has too high a percent of open interest, or b) there are concerns about the depth of the market.