Robert Shiller commented on CNBC yesterday morning that there are futures for the Case Shiller index (thank you, Bob) and that the futures imply a 2-3% rise in prices over 2013-2016.
Here’s a table showing what he meant. (There are a lot of numbers, and two parallel bits of analysis going on simultaneously, so let me walk through the tables).
The top half of the table shows the mid-market for the November cycle (X12-X16 = Nov 2012 to Nov 2016) for each of the 11 contracts. So, for example, the 152.1 mid for CUSX12 is the mid of the 151.0/153.2 bid/ask spread on the CUSX12 contract (not shown).
On either side of all the mid-market prices the bids and offers for the one-year calendar spreads are shown. So, the minus 4.0 under the X12/X13 column is the spread that someone is willing to buy CUSX12 below CUSX13. The offer, of minus 2.8 is the other side of the spread market.
On the bottom table, the middle number (the one in green) shows the percentage difference -of the mid-market prices - between one contract and the next. So the 1.91% is just the mid of Nov 2013/ mid of Nov 2012- 1, or 155.0/152.1-1 or 1.91%. In concept, this might be viewed as the implied HPA (Home Price Appreciation) using the mid-points of the outright markets.
Note that the numbers in green (for X12/X13 to X15/X16) range from about 1 to 4%. The strongest markets (e.g. SFR) have the highest implied HPA, while the weakest (e.g. LAV) have the lowest. Note thought that all percentages are positive, consistent with rising prices year over year. (I used year to year both to eliminate seasonal factors and so that percentage changes would also be annual changes.)
On either side of the green numbers in the lower table are what might be implied HPAs given the bid, or offered side of the calendar spreads. So, for example, the minus 4.0 bid on the CUSX12/CUSX13 calendar spread is a 2.6% difference (-4.0/152.1).
Calendar spread bids are typically wider than mid- to mid-market quotes (although not always) and offers tend to be tighter. Since these are the levels where one can put on a calendar spread (that is entering two sides of the trade simultaneously) these are a different indication of the range of debate over implied HPAs.
Thus, while the CUSX12/CUSX13 mid-market to midmarket qutoes have a minus 2.9 price difference (implying 1.91% HPA), absent locked markets, there’s no way to trade mid-market to mid-market, so the minus 4.0 (2.6%)/ minus 2.8 (1.8%) calendar spread shows where the actual calendar spread can be executed.
Typically calendar spread markets are deeper than outright markets as the risks may be different (and smaller).
Net, Professor Shiller was “spot on” in noting that the futures prices -whichever the two ways you look at them (calendar spreads or mid-market to mid-market) show 2-3% price increases per year, over the next few years.
Trading (or just observing) calendar spreads provides a different way of expressing a view on forward home prices. So, instead of taking an outright view that the Case Shiller index will rise or fall from today’s levels, one can also trade spreads based on changes in the market’s expectation of the pace of a rise or fall. Thus, if one thinks that prices will rise more slowly than market expectations (or even fall) one could buy the calendar spread, while if one is more bullish than the market, selling the calendar spread might be more appropriate.
I caution spread traders that while calendar spreads have two legs at inception, one has to have a plan for when the front leg expires. Some have argued that even after the front leg expires (and one is long that leg at the settlement price), one still has a postion that resembles an HPA “bet”.
There’s a lot of numbers here, and some basic algebra below the surface driving results. Please feel free to contact me (firstname.lastname@example.org) if you have any questions on this blog, or any other aspect of housing derivatives.