HPA – HomePriceFutures.com

Should forward HPAs converge?

I’ve compiled the following graph to pose the question, and prompt debate: should forward HPAs (i.e. annualized home price appreciation rates) converge across regions?

The graph stitches together year-on-year (YOY) price differences both between the CME Case Shiller home price index futures contracts and historical Case Shiller indices, as well as between futures contracts. For example, mid-market values of the May 2018 contract are compared to index values released in May 2017 (and the same for Aug ’18, Nov ’18 and Feb ’19 futures prices against historical Case Shiller indices), and then mid-market values between pairs of futures contracts are compared (e.g. the mid-market value of the May ’19 contract vs. the May ’18 contract, Aug ’19 vs. Aug ’18 contract, etc.).

While recent momentum, local wealth creation (e.g. in SFR (San Francisco) on tech stocks), and possibly inventory shortage, have pushed expected YOY gains in some regions to higher levels than average (i.e. SFR and LAV (Las Vegas)), by Nov ’19 all YOY gains converge to just above 2.0%. (Note that the other eight regional contracts include: BOS (Boston), DEN, (Denver), CHI (Chicago), LAX (Los Angeles), MIA (Miami), NYM (New York) SDG (San Diego) and WDC (Washington DC). The prices on the futures contracts expiring in 2019-2020 are consistent with slowdowns in HPA, but still positive gains, at about the level of income growth. It may also be consistent with a notion I’ve had for years, that forward prices are biased to be lower than expectations, as there are (at least based on current inquires) an abundance of hedgers.

However, most of the price inputs are mine (so I can’t offer a third-party analysis of what “the market is thinking”), and YOY gains across the regions -despite different momentum and despite differences in population gains -converge to about the same low rate.

Does it seem plausible that YOY gains for 2019, at least as priced in by longer-maturity CME contracts, should have SFR/LAV (the strongest contracts for the next year) and CHI/WDC (the weakest two based on one-year forwards) converge, and to converge to only 2% HPA?

I suspect not, and if there are readers looking to debate the point, I’d be open to trading some calendar spread pairs. (Recall that in trading calendar spreads one can simultaneously buy (or sell) one regional contract, while selling (or buying) a longer contract at a pre-determined dollar spread.) I would think that buying one regional calendar spread, while selling another region’s spread (covering the same time periods), might be a way of expressing views on relative HPA gains.

Please feel free to contact me (johnhdolan@homepricefutures.com) if you’d like to discuss this blog, or any aspect of hedging home price indices.

Thanks,

John

Using InterCity Spreads (IC) for relative HPA

In yesterday’s blog I touched on the notion that one might be able to use InterCity (IC) Spreads to express a view on the relative HPA performance of one region to another. While some regional pairs might make sense to debate (e.g. BOS v NYM, or LAX v SDG) a good place to start, and to illustrate the concepts in greater detail, is in contrasting one region versus the CUS 10-city index.

The example I’ve chosen is HCI vs. LAX. LAX is both the second largest Case Shiller region, but also (useful for this exercise) one where implied HPA is visibly lower than the 10-city index.

The table to the left shows all the necessary parts for converting CME futures prices into implied HPA. For example, the HCIG16 mid of 195.7 is 4.2% above the spot level, while the LAXG16 mid is only 2.7% above. As such, one might infer that the CME prices are consistent with the LAX index lagging the 10-city index over the next year. (I’ll let the economists and researchers offer reasons why this might/might not/ make sense.)

But what do you do if you disagree with the 1.5% premium of HCI vs. LAX (and you don’t want to take outright risk by just buying the LAXG16 contract)? If you believe that it’s too wide, then selling the HCIG16 outright (at 195.0 or 3.8% above spot levels) while buying LAXG16 outright (at 234.8.0 or 3.6% above spot levels) results in trades consistent with only a 0.2% implied difference. (A similar exercise results if you think the 1.5% is too wide and you try to go the other way. You’ll get 2.7%)

This is where IC spreads may be helpful. As with other spreads, IC orders are presented as the first contract (HCI) minus the second (LAX). Someone with a more bullish view on LAX (vs. HCI) might look to sell the HCI/LAX G16 spread. To do so at the outright levels described above would result in “paying up” 39.8 points. However in this example, the spread can be sold by hitting the -39.0 bid (Be careful with signs) or, as I’ve have already done, offer the spread at -36.0.

Net, the debate about the relative performance of the LAX index versus HCI has been narrowed to a 3.o point price range. The red and blue lines on the graph to the right show the narrower ranges. The blue line is consistent with the -39.0 bid, while the red line is consistent with the -36.0 offer. In both cases HCI prices would result in out-performance vs. LAX, but now with the -39.0/-36.0 market the debate has been narrowed to a range of 0.6% to 2.1% (vs. the 0.2% to 2.7% range from trading the contracts at the outright levels).

Net, the candle graphs (from yesterday) may be useful for illustrating implied HPA across regions, and intercity spreads give traders a way to observe and to express relative value over narrower ranges.

Please feel free to get in touch if you have any questions with this approach, or if you have any ideas for IC trading in G16 or other expirations. johnhdolan@homepricefutures.com

A market for 2015 HPA projections?

With yesterday’s release of the Case Shiller index numbers for Dec. 2014, several research firms have posted their projections for home price appreciation (HPA) for 2015. What can you do with these forecasts whether you believe them or think that they are too bullish/bearish?

I’d submit that the CME Case Shiller Feb ’16 (G16) contracts gives analysts a tool for both seeing what “the market” thinks as well as a platform for traders to express a view on 2015 HPA.

The diagram to the right shows bids, offers and mid-market values for the 11 CME housing contracts for Feb ’16, converted into percent changes versus the year-end Case Shiller #’s released yesterday. Recall that the Feb ’16 contract settles on the CS #’s released that month, which will be the index as of Dec.2015. As such, one might use the Feb ’16 contract numbers (bids, offers and mid-markets) for implied 2015 HPA.

The HCI contract (the one referencing the CUS 10-city index) is quoted at levels that translate into 3.8-4.6% increase in the CUS index between now and contract expiration. This seems in line with some research reports. (Recall that different indices- e.g. Zillow, Core Logic, FHFA reference different data, so projections may differ.) All ten regional markets are also quoted so users may infer forward HPA for specific regions (or trade).

As I noted in yesterday recap. bid/ask spreads in the Feb ’16 contract widened out as many quotes had been a function of Feb ’15/Feb ’16 calendar spreads. While bid/ask spreads are tighter today, they still have a way to go. I expect to see much tighter G16 markets over the next few weeks.

Nevertheless one can already see trends (e.g. quotes on SFR and MIA are consistent with those regions performing better than average (defined here as the CUS 10-city index) while LAX and WDC prices are consistent with lower projected HPA.

On can trade outright G16 markets, or trade one region on a spread versus the other (or the CUS 10-city index). I’m trying to develop inter-city spread trades and would be very open to trade proposals for any regional G16 contract vs. the HCI contract.

Feel free to contact me at johnhdolan@homepricefutures.com if you have any questions or if you’d like to discuss a trade.