I received an inquiry the other day about hedging NYC condo risk. I wrote a somewhat extensive reply as:
- I figured that there might be others interested in hedging the same exposure,
- there already exists Case Shiller indices for the more geographically dispersed NYM index, as well as a NYC condo index, and
- the issues raised here might apply to other one-off regional inquiries whether they be local (e.g. Greenwich), another Case Shiller index not traded on the CME (e.g. Seattle), a different index (e.g. Zillow, Core Logic, FHFA or other), or international (I’ve had discussions on how to hedge home prices in Vancouver, Beijing and London).
On the first point, NYC is right in my back yard (so I have a sense for it), and many of those who understand hedging and derivatives, work there. It’s also a market with a massive notional amount, yet possibly with more fungibility than the New Haven to Jersey Shore to Putnam County NYM index. So, lots of possible hedging interest with possibly less basis risk for individual home-owners (versus the index).
Second, I noted that there is a NYC condo Case Shiller index (as well as condo indices for BOS, CHI, LAX, and SFR -so we could have the same conceptual discussion with those cities). (See “Additional Info” tab at S&P website: http://us.spindices.com/index-family/real-estate/sp-corelogic-case-shiller for data).
A quick analysis (see diagram below) suggests that the NYC condo index has been highly correlated with the NYM index. (I used YOY % changes to minimize seasonality issues). While the relationship has been strong, NYC condo index values have steadily outperformed the NYM region. Both started at 100 in Jan 2000, but the condo index is now 270.92, while the NYM index is 185.26. (I leave discussion of why urban areas have outperformed regional indices to other commentators.)
However, given the strong correlation, one possible approach to hedging NYC condo risk, may be just to short some amount of NYM futures.
- price discovery on forward prices, and
- counterparty risk, as there is no exchange-traded product that might typically help with both.
(Note that there are many NYC condo indices that one might use. I’ve chosen to illustrate my example using the Case Shiller index as the correlation with the NYM would make hedging the futures exposure possible.)
An example of the challenge of price discovery might be if (hypothetically) you might want to sell the CS NYM condo index at +6% , one-year forward, versus the most recent level (i.e. 270.92*1.06= 287.16) while a buyer might only be willing to pay a price consistent with a +4% increase (i.e. 270.92* 1.04 = 281.76). However, without an exchange, and/or a central place to meet, or a chat room to compare levels of interest:
- you might not find each other, or worse,
- you might not know that there’s a buyer at 289! (BTW- the same exercise would be true if the parties wanted to do an options trade.)
This is where I’m trying to help, via website blogs, and social media touting. Think hub-and-spoke models.
However the challenge of the second point (counterparty risk), is that even if the parties agree to a price, how can one party ensure that the other side makes good should the prices move sharply against him. I’ve tried to address this with an OTC (over the counter) trades where each party puts up some margin (e.g. 5-10% of notional amount) to a third party custodian, where that becomes the total maximum payout.
For example, let’s say that we agree to take opposite sides on a NYC condo trade at a price of 284. We might agree that the notional amount equals $100/ lot, or $28,400 notional per lot. So, if you wanted to put on a $500k hedge we might agree to use 17 lots (for $484,280 notional exposure). Each of us would put up either 5% ($24,140) or 10% ($48,280) of the notional amount to a third party. The payout one-year from now would be based on the difference between the CS index published one year from today, versus the agreed-upon price (e.g. 284).
As such, if the final index was 281 (the seller would be awarded (3 points (=284-281) *17 lots * $100/lot) and would get their initial payment plus $5,100, while the buyer would get back their initial payment less $5,100). (I might charge some fee per lot for orchestrating the trade). If 5% margins were used that payouts would be capped at 269.8 (-5% versus trade, where the seller would get all the collateral) and 298.2 (where the buyer would get all of the collateral. Better still, I’d probably suggest bounds of 270 to 300 so that trades could be assigned and somewhat fungible.
While -5%/+5% might seem like a narrow range for one year moves, the range bounds expectations, not the spot index. So if the trade price was +3% over the spot index, payouts would max out on moves of about -2% to +8% moves in the spot index over the next year. On-year moves of more than 5% on a forward price are unusual, but not impossible. One could work with wider margins, but down-payments on each side would be higher. That is, the lower the margin, the better the leverage.
I’ve already done one such trade –for a different region – so this is not hypothetical.
There would be other issues. Data vendors might demand licensing fees on the use of their index in a derivative trade. I’m aware of some fee schedules but would have to check on other data vendors. In one instance there is a large fee for the first trade (referencing an index) but then marginal fees are much lower. Hence it will likely make sense to reference indices where either fees are low, or expected trading volume is higher.
So, does anyone want to talk about hedging NYC condo exposure, condo exposure where there’s another CS index, or any other home price index around the world? If so, feel free to contact me (email@example.com)