Year-on-Year “Forecasts” vs. Reality: Two observations

I get questions from time to time as to how good are the CME Case Shiller futures at forecasting eventual index levels.  Let me offer two comments.

  1. The “forecasts” are what traders think prices might be at some point in time in the future.  Therefore, as the world changes, it shouldn’t be a surprise, or a failure of the contracts, if the end result varies from quoted markets some months or years before.  YOY changesFor example, forward oil markets were priced a few years ago at $90.  When prices fell to $50 it didn’t imply that the contracts were wrong, or flawed, but that sentiment had changed.  The same holds true for home prices.  Changes in prices reflect changes in sentiment, not the accuracy of the contract.  Note that the graph of year-end historical forward Case Shiller indices has been rising year over year since Dec. 2011.  The Dec 2011 prices weren’t wrong -it’s just that sentiment has gotten even more bullish every year as actual Case Shiller indices have improved.
  2. That said, over the short term, futures and index levels have to converge, as the contracts cash-settle just like the S&P 500.  So while longer-maturity prices might wander, one might expect shorter forecasts to be more accurate and unbiased.  In fact if the general level of expectations has been risingYOY (as noted above) then one might expect one-year forward contracts to have under-priced eventual index levels.  Surprisingly (to me) this doesn’t seem to have been the case (until very recently).  The table to the right shows the index level and one-year forward futures contract in the month prior contracts have expired.  For example, on Feb 28th, 2014, the spot CUS index was 180.07 and the mid-market contract level one-year forward was 191.0.  Such forward prices would be consistent with a 6.07% YOY gain.  However, the index released in Feb.2015 was 187.74, which was only a YOY gain of 4.26%.  The one-year forward contract had overstated the actual HPA by 181 bps.  This overstatement seems to have been the case in the last three quarters.  However this trend might be coming to an end as the mid-market price on the Q15 contract (196.5) is higher than the mid-market on the same contract one year ago (194.7).  We’ll see on the 25th if the trend breaks.

So, if longer-term the contract prices have seen upward price moves (once evidence of stronger home prices is announced with every index release), but in the short-term, contract prices have been higher than eventual index releases, what’s going on?  How can one part of the curve understate expectations, while another overstates?

My sense is that the answer is based on trading nuances and the roles of hedgers and the market maker.  On short-dated contracts I’ve been pushing up bids in a rising market to find sellers.  With no evidence of turn in home prices (to date) there have been few sellers for one-year forward.  By contrast, there is a combination of natural hedgers sitting on the offered side of longer-dated contracts, while buyers have to justify higher forward prices not with one year of 4% HPA, but 3-4-5 years of such implied gains.

If (historically) front contracts were overstated, while longer-dated contracts resulted in upward surprises, then it might have made sense for hedgers to have shorted one-year forward contracts (where BTW there was much more liquidity) and then roll the forwards 1-2 years forward at time as expiration approached, and for those more bullish, to buy longer dated contracts for even short-term outlooks.

Now this may have been a more profitable strategy over the last few years, but that’s no guarantee that it will work going forward.  Index levels have been coming in closer to one-year forwards over the two quarters (this expiration plus May) and the longer-dated YOY upside “surprises” (e.g. Dec 2013 vs. Dec 2011) have become much more dampened.

Still, I think that this is interesting food for thought. It might lead some traders to consider buy and roll, or sell and liquidate early strategies, rather than matching their targeted holding period with the contract of the same expiration.    There might be more rolling of contracts (via calendar spreads) and/or traders moving to more liquid markets to execute a strategy with a different time frame.

Please feel free to contact me (johnhdolan@homepricefutures.com) if you have any questions on this blog or any other aspect of hedging home prices.