I’ve pulled together a short marketing piece in time for the NBEA conference and ABS EAST. (see attached). Readers who follow this blog should recognize most of the pages. As my goal is to get the word out, please feel free to forward on to interested parties.
Category: Blog
Basics -Calendar Spreads
Calendar spreads may be a useful tool for trading or interpreting the boundaries of HPA (Home Price Appreciation) expectations.
Using the 10-city CUS contract as an example I’ve put together an illustration (some are not real prices) of a series of calendar spreads. The graph at the top show bids and offers in the CUS set of contracts,the middle table shows spread quotes, and the bottom table shows the implied HPA from each quote.
The first (and most important disclosure for trading in spreads) is that I’ve used numbers that represent the pay-up from one contract to the month. So for example the Nov ’11/Nov ’12 offer is shown as 680. In reality one would propose this trade by offering to sell Nov ’11 while buying Nov ’12 at MINUS 680. That is since the first contract is the one that drives the spread, and you want to sell the front contract (Nov ’11) lower than the back (or buy the back higher than the front) the spread would be quoted at a discount. Note that I’ve shown the quotes as how much one would pay up (bid) for the forward contract (e.g. +100 in Nov ’11//Nov ’12) versus how much one would sell (Ask) the forward contract over the nearer one (e.g. 680 on the Nov ’11/Nov ’12 spread). BE CAREFUL.
With that caveat, one can note that the implied bids reflect HPAs of less than 1% while the implied HPAs on offers range from 3.6 to 5.2%. Longer term (e.g. Nov ’11/ Nov ’14) offers tend to trade at lower HPAs (reflecting some element of potential for up and down price movements over longer periods) while back contract single year offered spreads tend to trade at higher HPAs (reflecting uncertainty and the possibility of the return of a more robust housing market).
It may be possible to use spread trades to take a more conservative view on the timing and magnitude of future home price moves. Note that the Nov ’11/ Nov ’14 spread quote is 1260 (1700-440) while the riskier outright Nov ’14 market (not shown) has a bid/asked spread of 2740 (18000-15360).
Finally, one last set of caveats. The contracts are not trades where one can hold the spread trade through to the maturity of the longer contract – like a TROR swap. (Total Rate of Return). It may be obvious to some, but once the front side of the trade expires, one would be left with an outright position in the longer contract. Therefore it’s important to consider not where home prices will go over the life of the back contract, but where expectations for home prices on the back contract will be, as the front contract moves toward expiration.
Basics -Hedging
I was recently reminded that for all of the discussions here on pricing, that I hadn’t shown a basic “how to hedge” example using the CME Home Price futures. Let me rectify that with the following illustration.
The graph to the right illustrates the P&L impact of a change in the value of a hypothetical $1mm house (for ease of math) over a holding period through the expiration of the Nov 2014 contract. The steepest line indicates what everyone learned during the 2000 decade -that you can make (or lose) a tremendous amount of money as home prices move up or down.
A key assumption in all three lines is that the price of this house moves in tandem with the Case-Shiller (CS) price index for the region. For the unhedged version, that means that if the house was bought when the current index is 172.76, that the value of the house will change (on a percentage basis) as the index changes. So, should the index be 200 in Nov 2014 (for a 15.8% total increase or a 3.5% HPA over the 4.25 years) then the $1mm home price would rise in value by the same percent change.
Is this realistic? Will all home prices move in tandem with the index? Of course not, and it’s important to appreciate and incorporate that into one’s analysis. The CS index doesn’t include houses with significant alterations or upgrades. There may be sub-regions that due to new business, new highways, new schools, may see price changes that are different from the overall index. Finally, no one can hedge against someone paying 5% too much for a house, or selling one at a distressed level.
That said, I go back to the website cover page and ask, has the price of your home moved generallyalong the index price path for your region. In most cases, the answer is that index price changes, that is changes to overall home prices in the region, account for the bulk of individual home price moves. Could your home’s “value” go down by 5% while the index goes up by 5% -YES, and if you’re going to hedge you have to know that (or worse) is a possible result. However, the Case Shiller index attempts to measure average home prices, and if average home prices rise or fall 10%, that would mean most homes moved in the same direction.
Going back to the illustration above, the two flatter lines indicate the net P&L of hedging a portion of the house’s value with a sale of CME contracts for the NYM region on the Nov 2014 contract. With the same qualifiers as above, it illustrates the notion that while you can’t sell a portion of your house, you can hedge as much of their house’s value as you care. As with all hedging strategies, the more of the underlying value that is hedged, the flatter the net P&L.
So, as the table to the right shows, if the index goes to 200, and the house rises in value by the same percent to $1,157,675 the gain of $157,675 on the house will be offset by the loss on the contracts. If one had sold 5 contracts that would be a $33,750 loss for a net gain of $123,925. (Recall that each contract has a value of $250 times the index price divided by 100. As such 5 NYM contracts at 17300 have a notional value of $216,250. Twenty contracts would have a notional value of $865,000).
The key conclusion is that hedging has the potential to dampen losses if one is willing to forgo some of the upside.
A few important points:
1) I used this example (NYM) because the Nov 2014 contract was bid (173) at almost the same level as the most recently released index (172.76) For other regions, and during other times, the forward price might be a premium or discount to today’s spot CS index.
2) The benefit of using the Nov ’14 contract is that it hedges the value of the house for four years. Today it is the longest dated contract. As time passed new contracts for Nov 2015 and beyond will be rolled out and a homeowner/hedger can determine whether they want to extend their hedging time-frame.
3) If the owner holds the house and the hedges through Nov 2014, there is nothing that they need to do to unwind the hedges. The closing value for the Nov 2014 contract will be the Case-Shiller index value announced that month. Note, though that if the hedge matures without a new one being put in place, that the homeowner will not have any hedge protection after that date.
4) If the owner sells the house before that period, they can independently decide what to do about the hedges. If they are moving within the region, they may consider keeping the hedges on. If they’re downsizing, moving to a rental, or leaving the region, they may want to “lift” (buy back) the short positions. Prices for all contracts (including this Nov 2014 example) may vary during the life of the contract and there’s no guarantee where prices might be, or even that (in this example) the Nov 2014 price will not be at a premium or discount to the then current Case-Shiller index.
5) Just as hedging can’t protect one from overpaying 5% for a house, the same applies on the initial futures execution (and unwind if lifted before maturity). The Nov 2014 contracts tend to have wide bid/asked spreads, with limited depth (# of contracts that can be traded at at price) and I would strongly encourage anyone using those contracts to work with limit (as opposed to market) orders, and to consider starting with offers inside the current bid/asked spreads.
With all of these caveats I do believe that this may be an interesting product for those considering hedging. Most people buy houses to stay in them for a while, yet the results of the last few years show the downside of unhedged positions. Using these futures as a hedge MAY offset the risk of large region-specific price declines over longer time frames.
As with all of my blogs and comments, but particularly with this one, this should not be considered investment advice or an offer to buy or sell contracts. Traders should consider their own unique situations and discuss the ramifications to any trade with their own tax, investment and accounting experts. While a number of risks were outlined, it is not an all-inclusive list and events and price moves could occur that might cause a trader to lose a significant amount of money. Only those who fully understand the issues associated with hedging should consider hedging strategies. Anyone trading CME Home Price futures should understand the nuances of the Case-Shiller index calculation. Margin calls might require cash outlays before the end of the contract that may require the hedger to post more collateral. The results shown here are hypothetical, may not be applicable to your situation and are used for illustration only. These are my views and they should not be inferred to represent the views of the CME, S&P, FISERV or anyone associated with the Case-Shiller index.
2010: Projections vs Market Prices
With the expiration of the August contract (the June Case-Shiller data) we are now half way through 2010 and can begin to turn attention the to Feb 2011 contracts. (Most of the time the front two contracts and the Nov cycle are where 90+% of the live quotes are.)
The thing that makes the Feb ’11 contract interesting is that it settles on the December Case-Shiller data. As such expectations for year-on-year, or 2010 home price moves, can be tallied by a simple comparison of the Feb. 2010 Case-Shiller release and the CME contract values for the Feb. 2011 contract.
As the graph above shows, the Case-Shiller index for the 10-city index improved 1.8% through June, since the release, in February, of the December 2009 data. Prices for the bids (squares) and offers (diamonds) for the Nov ’10 and Feb ’11 contracts are shown, with lines interpolating prices between those points and the most recent index values.
As quotes in the Feb ’11 contract only begun to show after Aug ’10 rolled off, today’s bid-asked spread is wide at 8.4 points. Trading may remain challenging as we go through a period where the CS spot index has benefitted from the housing credit program, while sentiment has turned negative. (See 8/31 blog).
Nevertheless the bid-asked spread translates into expectations of 2010 home price performance of -2.9% to +2.4%. This lower bound is important as the recent MacroMarkets survey has economists calling for a decline of, on average, 2.08%. (Contact MacroMarkets to register for their survey results http://www.macromarkets.com/). As there tends to be a reasonable distribution in those esitmates I would imagine that many are expecting declines of more than -2.9% (as implied by the 15300 bid for the Feb ’11 contract).
Stay tuned. It’s a long way to February.
(While I am illustrating the CUS index and CME contracts, the same analysis can be applied to each of the 10 indices where CME contracts are traded. Look for city-specific discussions in future blogs.)
Basics: Impact of lagged, moving averages (Aug CS)
While the headlines for this month’s release of the Case-Shiller indices look positive, many who follow this market know that the index does not reflect today’s (8/31/2010) sentiment, but measures a 3-month window, that is lagged two months.
As the illustration indicates, today’s numbers measure the results of home prices from April through June. Note that today’s index is primarily the result of dropping the March numbers (that were part of the May series -released in July), while adding the June numbers. As such, one of the key parameters for month to month comparisons on the current index is the difference between the March and June single month numbers.
The formulas involved are not simple averages of the three months -hence, why I show index weights (w1, w2…) for each month. As such solving for each month’s precise value is impossible (six variables for each month’s equation), but one can strongly infer that a reason the Boston contract rose from 155.95 to 157.83 was a large (~5point) difference between March and June single month values. (Possibly a result of seasonals and the pending expiration of the tax credit program).
This feature of the CS index calculation allows for some averages to rise, even when sentiment has turned negative. As such, you get situations such as this month where the actual NYM and CHI indices were higher than where CME contracts were offered yesterday (when trading expired). (See table). As bid-asked spreads in the current month averaged 3.5 points across all contracts at expiration, this kind of trading opportunity will likely to continue to exist when sentiment changes are not aligned with index math.
(BTW-Hat’s off to Radar Logic for “calling” the CUS quote. They had predicted that the index would be 161 in a press statement pre-#’s).
So with the index trending one direction, and sentiment the other, where do we go from here?
First the good news. With the expiration of the August contract open interest declined by only 6 contracts. We next face a November expiration (which has historically had the biggest open interest) to be followed by the Feb. ’11 contract which people should focus on for year-on-year comparisons versus 2010 home price predictions. So, lots of people who should care.
Next the bad news (or good if you like volatility). With the divergence between index calculations and sentiment, I would expect bid-asked spreads for near contracts to be wider than they’ve recently been. As Professor Shiller noted on today’s S&P call, prices might revert to their historical long-term upward trend, or they could revert to the strong downward leg that was interrupted by the housing stimulus plan. Timing on the (well touted, upcoming) downward turn in home prices will be critical to trading Nov ’10 and Feb ’11 contracts.
“Be careful out there”
Boston Bubble website
One of the upsides of blogging, and focusing on a single topic, is that you get introduced to, or stumble across others, who share your enthusiasm for a topic. As such it was my good fortune to run into the folks who have put toghether www.bostonbubble.com. While there are many sites that bang the drum on bearish views, this site has -for 5 years – pulled together headlines and forum topics with some great analytical support. While some of the focus has been very neighborhood-specific, they have also used the Case-Shiller Boston index, and CME Boston contracts to ground discussions about history and expectations.
As evidence of their work I share here one of their graphs where they overlay CME contract prices at various points in time versus a combination of the Case-Shiller index and recent (this was in May) CME prices.
The graph is interesting because while it highlights the relatively slow changes in the underlying index, it show how expectations (I think the mid of bid and offer) might have changed more quickly during 2007-08.
(My expectation is that with forward contracts getting more attention, that price jumps will be smaller, and forward curve volatility will be lower.)
Check out their site, especially the forums. Good work Tim.
Basics – CUS 10 v CUS 20
A number of people interested in following the Case Shiller indices have focused on the 20-city index. While this is a broader index, the CME contracts are on the 10-city index. Therefore, if you’re going to trade the CUS contract, it’s important to know the differences between the two indices.
While the CUS-10 index has many of the larger cities, it is not a blanket list of the top 10. The list to the left shows that Detroit, Dallas and Atlanta are all excluded from the CUS-10, while smaller cities such as Denver and Las Vegas are included. Nevertheless, ten cities comprising the CUS-10 account for more than 71% of the CUS-20.
The CUS-10 has outperformed the CUS-20 since the indices were benchmarked at 100 in Jan. 2000, with the CUS-10 valued at 159.36 (as of the July release of the May indices) and the CUS-20 at 146.43. However most of that 11.7% outperformance came in the first four years.
Since Jan 2004 the CUS-10 index has declined 2.2%, while the CUS-20 has almost matched that with a decline of 3.5%. The oft-cited woes of Detroit and Cleveland, that are counted in the CUS-20, have been offset by the strength of cities in the Northwest (Portland and Seattle.)
Furthermore, since the CUS-10 accounts for such a large percent of the CUS-20, it shouldn’t be surprising that the two move together. Since less than 30% of the CUS-20 doesn’t overlap with the CUS-10, you’d need some pronounced movements in the remaining ten cities for the indices to diverge dramatically.
The other major difference (than size) between the two indices is kind of red state vs. blue state syndrome. That is, the CUS-10 has three cities from the Northeast, and three cities from California. The “other” ten cities in the CUS-20 index tend to be more Mid-America with exposures to the Southwest (Dallas, Phoenix), the Southeast (Charolette, Atlanta, and Tampa), and a broad definition of the Mid-West (Cleveland, Detroit and Minneapolis).
Any pronounced difference in performance between the CUS-10 and CUS-20 index would have to be based on these two features: size and location. Given some extreme risks(earthquake, terrorism, rising sea levels) one might find fewer risks in the CUS-20 index. On the other hand, as globalization continues, or foriegnors look to invest in US real estate, the larger cities might outperform. While the smaller cities might have more low-income borrowers, at least there’s lending going on to non-jumbo clients. Who’s going to lend $1mm on a starter home in Palo Alto?
Net, I’d be surprised to see big differences over the next few years.
Washington follows us!
One of my personal highlights of the recent Treasury meeting to address housing problems was to discover that HUD has been following the CME futures markets – and uses changes in the prices of forward contracts as indications of changes in sentiment on forward home prices.
This graph, that shows the changes in CME contract prices between Jan. 2009 and July 2010, is from the second page of a monthly publication titled “The Obama Administration’s Efforts To Stabilize The Housing Market and Help American Homeowners”, otherwise known as the monthly housing scorecard. (see link below for entire 8-page report.
Regardless of your politics, the package is a useful tool with 16 graphs and a tally of many housing metrics.
Beyond this report the HUD website is great springboard to press releases on programs designed to help borrowers. As many efforts by the Government to address distressed homeowners may have an impact of foreclosures, housing inventory, and therefore pressure on home prices, the programs are worth understanding.
Basics -Convergence
With the expiration of the August contracts approaching, I thought that it might be useful to illustrate how convergence in the CME Home Price futures contracts as taken place in the past.
The graphs are of 1) the Nov 2007 CUS contract, 2) the Nov 2008 contract, and 3) the Case-Shiller CUS index over time. Closing prices on the CME contracts are noted by the lighter shaded lines. Actual trades are noted with darker squares.
A few observations:
1) (A simple one) The CME contracts trade daily, while the Case-Shiller index is only released monthly -hence, the infrequent, step-function in Case-Shiller numbers.
2) Note that trades tend to result in changes to prior closes. (See Basics blog on Closing for more details.) Note how in the Nov 2008 series that the closing price was unchanged at 183.8 from Feb 3-26 until a trade took place at 176, which then resulted in the subsequent lower close. (Again, getting back to the point that I’ve harped on that “Closes” may not necessarily represent current sentiment.)
3) Note how prices on both series converged toward the CS index that was released in the month the contract expired. As prices for the contract must be the same as the index at expiration any large variance from the expected CS index can be arbitraged. As the CS index is a lagging, moving-average index, it is extremely unlikely that the CS index will dramatically jump in any month. This means that longs and/or shorts have an ever clearer picture of the eventual value of their outstanding positions.
4) As such, an option for longs/ shorts is to hold positions to expiration. While this may dampen trading (bad for me as a market maker) it gives increased confidence to those thinking about entering new positions that they won’t be subject to wide bid-asked spreads if they hold the position to maturity. This increased confidence should give more people comfort that they can use these contracts as longer-term hedges (or insurance) against long/underweight positions (good for me as a market maker.)
5) Finally, note that early in 2008 the Nov 2008 contracts traded to pronounced discounts to the then spot Case-Shiller index. Did the contracts incorporate a consensus view that home prices had further to fall? Were sellers so desparate to find a hedge to other real-estate related exposurses that they oversold the contracts? Were the longs just being adequately compensated for providing a service (protection against a melt-down in home prices) that required a sufficient rate of return? All of these questions are the themes of academic studies that should be reviewed.
Book Review – The Housing Boom and Bust (Thomas Sowell)
One of my roles away from here is that of moderator for a book club at my local library. (Non-fiction, primarily foreign affairs and current events) That role has reintroduced me to a love of reading that proved elusive with a life (in the early 2000’s) of early business meetings, client presentations, and getting the kids off to bed. As it takes reading three books to find the one I can recommend to the group, and since “the crises” has been so much on every one’s radar, I’ve had the chance to read them all. (Collapse of Lehman, Collapse of Bear, Collapse of AIG, Collapse of …..).
The book that stands out from that crowd, that I would like to tout here, both for being well-written and due to its relevance to this site’s themes is –The Housing Boom and Bust(revised edition) by Thomas Sowell.
Mr. Sowell’s approach is to explain (in the newly updated, slim paperback version) how forces (incentives, mostly) built up over a period of years, were the reasons behind the inevitable real-estate crash. He tars politicians (both local and national, Democratic and Republican) with their use of policies that were designed to appease either in the short-run, or to limited constituencies.
A conservative by nature, he suggests that “Affordable Housing” programs represent an intrusion by the government into market practices. He argues that many of the affordability products that were created during the boom years, that were designed to increase home ownership, by having banks lend (and then securitize) to those borrowers, were only made to appease mis-guided government policies.
While the common wisdom is that we all learn from our mistakes, he offers a dire, but critical forecast that “…those who say that politicians never seem to learn overlook the fact that there is no reason for them to learn, when they pay no price for being wrong when they can simply blame others and continue on with policies that have been politically beneficial to themselves, however detrimental those policies may be for the country.” (No surprise to WSJ readers, Mr. Sowell cites Congressman Frank and Senator Dodd several times as their roles and views on housing policies have “evolved”).
This is the best book that I’ve read on the roots of the real-estate financial crises (although there are some other great ones on how money was made or lost). They say ” If you want to know where you are, figure out how you got here”. This book does a great job showing how small, well-meaning actions, took us down the path to where we are today.
Similarly, if you want to figure out the political angles on key issues related to home prices down the road (e.g. fiancial regulation, the role of the GSEs), and to home price futures contracts, then I strongly encourage you to read this book.