Release Date: September 24, 2009

July 2009 By Quinn W. Eddins and Matthew R. Zogby

Key Characteristics • The gain in the RPX 25-MSA Composite between the end of March, when it hit its lowest point of the year, and the end of July was larger than would be expected given historical price trends. Between March 30 and July 23, 2009, the Composite increased by $13.24, or roughly 7.2%, from $185.16 to $198.40. The average gain over the same period during the last 10 years is $9.72 or 5.4%. • After controlling for seasonal price patterns using the Census Bureau’s X-12 seasonal adjustment program, we found that the RPX Composite increased $5.82, or 3.1%, from the end of March to July 23. Had the increase in the RPX Composite been the result of seasonal factors alone, one would expect the seasonally adjusted index to remain flat over that period. The fact that it increased indicates that the strength in home prices exceeds what one would expect given seasonal factors alone. • Notwithstanding strong spring price growth, the Composite remained 29% below its peak in June 2007. • In January, monthly sales volumes hit their lowest point in Radar Logic’s historical data, and since then monthly volumes have remained low relative to years past. Nevertheless, volumes have been increasing rapidly. The 25-MSA sales volume increased by 49,000 sales, or 87%, between January and June. In percentage terms, this was the largest June-over-January gain in the last ten years. In terms of the number of sales, the growth in 2009 was comparable to the gain over the same period in 2006 and 2007, the peak years of the housing bubble. • If history repeats itself, housing demand will start to decline in the autumn as seasonal factors influencing market demand, such as cold weather and the beginning of the school year, become more salient. However, several other factors will also influence demand in the near term: widespread negative equity, conservative mortgage underwriting, the expiration or extension of government incentives, the unleashing of pent-up demand by low home prices and changes in the unemployment rate. Each of these is discussed below and will be followed in Radar Logic research in coming months. • The supply of homes is large and is likely to increase. According to the National Association of Realtors, the inventory of existing homes was four million units in July, and that figure includes only a fraction of the large and growing inventory of foreclosed homes. Yet there is evidence that a significant percentage of bank-owned properties are being held off the market. By releasing their REO inventory onto the housing market slowly, banks are effectively reducing the impact of the distressed inventory on home prices. If the current trend continues and the demand for housing holds up, home prices could continue to recover despite the record number of foreclosures.

www.radarlogic.com

Recent Trends in Home Prices   The 25-MSA RPX Composite is still roughly 29% below its peak, but it has been rising quickly since hitting its lowest level of the year, $185.16, on March 30. By July 23 it had increased $13.24, or roughly 7.2%, to $198.40. To put this growth in context we compared it to growth over the same period in prior years. As shown in Exhibit 1, the growth in 2009 was larger than the ten-year average for this time of year in both dollar value and percentage terms. It was also larger than the average gain during the early years of the housing boom (2000 to 2002) and the gains during the peak years of the housing boom (2006 and 2007). Exhibit 1 25‐MSA Composite: Change in Price Between March 30 and July 23 % Change (left axis)

$ Change (right axis)

10.0%

$25.00

8.0%

$20.00

6.0%

$15.00

4.0%

$10.00

2.0%

$5.00

0.0%

$0.00

‐2.0%

‐$5.00

‐4.0%

‐$10.00 ‐$15.00

‐6.0% 2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

mean  mean  mean  mean  (2000‐ (2000‐ (2003‐ (2006‐ 2009) 2002) 2005) 2008)

In July the rate of growth slowed. The Composite increased 0.6% in the 30 days ending July 23, after increasing more than 2% over each of the prior two months. Rather than being a sign of flagging demand, deceleration in the rate of growth reflects the expected seasonal pattern.1 In fact, when we compare July’s month-over-month gain to the historical pattern, as shown in Exhibit 2, we find that the growth in 2009 was greater than the average gain over the same period during the last ten years, both in percentage terms and in terms of actual dollar value. The $1.19 gain was comparable to growth over the same period during the boom years of 2000, 2001 and 2004 and a marked departure from the $3.00-plus declines observed during 2007 and 2008.

Since 2000, the Composite price has shown seasonal strength from January to June, remained relatively flat from June to August, and displayed relative weakness from August to January.

1

www.radarlogic.com

2

Exhibit 2 25‐MSA Composite: Change in Price Between June 23 and July 23 % Change (left axis)

$ Change (right axis)

2.5%

$5.00

2.0%

$4.00

1.5%

$3.00

1.0%

$2.00

0.5%

$1.00

0.0%

$0.00

‐0.5%

‐$1.00

‐1.0%

‐$2.00

‐1.5%

‐$3.00

‐2.0%

‐$4.00 ‐$5.00

‐2.5% 2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

mean  mean  mean  mean  (2000‐ (2000‐ (2003‐ (2006‐ 2009) 2002) 2005) 2008)

To estimate how much of the price gains are attributable to factors other than seasonality, we controlled for seasonal variation in the Composite using the Census Bureau’s X-12-ARIMA seasonal adjustment program.2 The seasonally adjusted RPX Composite increased $5.82, or 3.1%, from the end of March to the end of July (see Exhibit 3). This increase was larger than the average increase from 2000 to 2009, both in dollar value and percentage terms. It was also larger, in dollar value, than the average growth during the early years of the boom, 2000-2002. Exhibit 3 25‐MSA Composite: Change in Seasonally Adjusted Price Between March and July % Change (left axis)

$ Change (right axis)

8.0%

$20.00

6.0%

$15.00

4.0%

$10.00

2.0%

$5.00

0.0%

$0.00

‐2.0%

‐$5.00

‐4.0%

‐$10.00

‐6.0%

‐$15.00

‐8.0%

‐$20.00

‐10.0%

‐$25.00 2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

mean  mean  mean  mean  (2000‐ (2000‐ (2003‐ (2006‐ 2009) 2002) 2005) 2008)

The period for which historical Composite prices are available includes several very strong years in which prices increased through the autumn, and therefore the seasonal adjustment model likely overestimates seasonal strength in the spring and underestimates seasonal weakness in the autumn. As a result, the increase in the seasonally adjusted Composite from March to July may underestimate the contribution of non-seasonal factors to the increase in the unadjusted index.  

2

www.radarlogic.com

3

The cities where prices have increased the most from their respective troughs are Boston (+38%), Denver (+21%), San Francisco (+19%), San Jose (+16%), Columbus (+16%) and Washington, DC (+14%). Boston, Columbus and Denver have historically been characterized by large seasonal shifts in housing demand, which no doubt contributed to their recent price growth. Nevertheless, the magnitude of the price growth in these cities in spring 2009 outstrips that of historical seasonal price changes. San Francisco, San Jose and Washington, DC, do not typically display much seasonality in pricing, so price gains there can largely be attributed to non-seasonal factors. These cities are characterized by high home prices and their RPX values have each fallen in excess of 30% from their respective peaks over the last few years. The increase in affordability may have unleashed pent-up demand for housing from people who were priced out of the market during the boom. The cities that have increased the least since their respective troughs earlier in the year are Seattle (+2%), New York (+4%), Detroit (+5%), Sacramento (+5%) and Phoenix (+5%).

Recent Trends in Sales Volumes As reported by Radar Logic, the National Association of Realtors (NAR) and others, sales volumes have grown considerably since the beginning of the year. In January 2009, the total monthly sales volume3 in the 25 MSAs tracked by Radar Logic was roughly 56,000 sales. By June the monthly sales volume had increased 87% to roughly 105,000 sales. As shown in Exhibit 4, this was the largest June-over-January increase in percentage terms since 2000, even larger than the January-to-June percentage increases at the height of the housing boom. Exhibit 4 25-MSA Composite : Change in Monthl y Sa les Volume betwe en Janua ry and June % Change (left axis)

# Change (right axis ) 100,000

100% 90%

90,000

80%

80,000

70%

70,000

60%

60,000

50%

50,000

40%

40,000

30%

30,000

20%

20,000

10%

10,000 0

0% 2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

mean (20002009)

mean (20002002)

mean mean (2003- (20072006) 2009)

For the purposes of this analysis, sales volumes include all arm’s-length purchases of residential property whose records Radar Logic received within 63 days of the transaction date, regardless of whether the records included the requisite data elements for calculating Radar Logic Daily Prices. To enable an apples-to-apples comparison of recent and historical monthly sales volumes, we inferred whether or not we would have received the records for historical transactions within 63 days of the transaction date based on the date our data provider first processed the records.  3

www.radarlogic.com

4

This large gain in percentage terms came off the lowest monthly volume in Radar Logic’s historical data, and monthly volumes remain low relative to the same periods in past years (see Exhibit 5 below). Nevertheless, when we compare the absolute increase in monthly sales from January to June 2009 to the changes over the same period in years past, we find that the 49,000-sale increase in 2009 was comparable to the increase at the peak of the housing bubble in 2006 and 2007. As shown in Exhibit 4, the absolute increase in monthly sales in 2009 was larger than it was in 2001 and 2003 (early boom years), 2007 (the peak of the boom) and 2008 (the housing bust) but smaller than it was in 2004 and 2005, the years with the greatest volume levels. To get a sense of how much of the January-to-June increase in sales volumes can be attributed to non-seasonal factors, we applied the X-12-ARIMA seasonal adjustment program to monthly sales volumes from January 2000 to July 2009 to control for the typical seasonal variation over that period. As can be seen in Exhibit 5, seasonally-adjusted monthly sales volumes increased from January to June, 2009, suggesting that the recent strength in home sales exceeds the strength one would expect to see given the seasonal pattern alone. When one controls for seasonal factors the recent uptick in sales activity is quite large relative to sale growth over the same period in the past nine years. Exhibit 6 depicts the June-over-January change in seasonallyadjusted 25-MSA sales volumes from 2000 to 2009 in both percentage terms and raw sales. Whereas the growth in non-seasonally-adjusted sales volumes shown in Exhibit 4 was below the ten-year average in 2009, growth in seasonally-adjusted sales volumes was above the ten-year average in 2009. Only 2002 and 2004 posted greater growth in seasonally adjusted sales. Exhibit 5 25‐MSA Composite: Monthly Sales Volume Non‐Seasonally Adjusted

Seasonally Adjusted

250,000

200,000

150,000

100,000

50,000

www.radarlogic.com

5

Jan‐09

Jul‐08

Jan‐08

Jul‐07

Jan‐07

Jul‐06

Jan‐06

Jul‐05

Jan‐05

Jul‐04

Jan‐04

Jul‐03

Jan‐03

Jul‐02

Jan‐02

Jul‐01

Jan‐01

Jul‐00

Jan‐00

0

Exhibit 6  25‐MSA Composite: Change in Seasonally Adjusted Monthly Sales Volume between January and June

% Change (left axis)

# Change (right axis)

20%

30,000

15% 20,000 10% 10,000 5% 0%

0

‐5% ‐10,000 ‐10% ‐20,000 ‐15% ‐20%

‐30,000 2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

mean  mean  mean  mean  (2000‐ (2000‐ (2003‐ (2007‐ 2009) 2002) 2006) 2009)

The growth in sales volumes slowed in July, but the month-over-month sales growth in the 30 days leading up to July 23 was exceptional when compared to the change in sales over the same period in past years. From June 23 to July 23, the total sales volume in the 25 RPX MSAs increased roughly 6%, slowing from a 12% average monthly increase over the prior five months. However, in most years, sales volumes experience a mid-summer lull from June to July. Of the last ten years, only two others, 2004 and 2005, posted month-over-month gains in sales volumes over this period. As noted above, 2004 and 2005 were the peak growth years of the housing boom in terms of sales volume. The spring growth season is lasting longer this year than it has in years past, which is a further indication that housing demand remains robust, at least for the moment.

Outlook for Housing Demand While absolute volumes remain low relative to earlier in the decade, the healthy growth in sales volumes relative to years past indicates that demand for housing has gained strength in recent months. However, as we enter the fall it is difficult to say whether housing demand will continue to grow or start to retrench. There are a number of factors to consider: Seasonality As we have already mentioned, seasonal factors such as the beginning of the school year, colder temperatures and shorter days will put downward pressure on housing demand through early 2010.

www.radarlogic.com

6

Negative Equity In a recent analysis, Citigroup estimated that 26% of mortgages, or 13 million loans, are currently “underwater”.4 Negative equity has a chilling effect on housing demand because it makes it difficult for homeowners with mortgages to sell their existing homes and buy new ones. If home prices start to fall again the number of underwater mortgages could increase drastically, exacerbating the downward pressure on housing demand. Conservative Underwriting After years of lax underwriting, lenders have now shifted to the other extreme. Only borrowers with the best credit scores can qualify for low advertised interest rates and lenders are requiring down payments of 20% to 30% or more. Since the vast majority of homebuyers use a mortgage to purchase a home, these tight lending standards are a significant constraint on demand. Inflation Inflation is of particular concern moving into 2010. Government spending has created an unprecedented increase in the money supply, which leaves a high likelihood of inflationary pressure in the years ahead. Current demand is no doubt being partially fueled by the favorable interest rates that exist in the market today. If the Fed is forced to move rates higher at a faster rate than is expected in order to combat inflation then some of the demand for housing may be suppressed. Government Incentives Government incentives, particularly the $8,000 first-time homebuyer tax credit and the temporary increase in conforming loan limits, are almost certainly adding to the current demand for housing. Both of these benefits are set to expire at the end of this year. The expiration of one or both would almost certainly depress demand in the short-term, though the impact would vary widely from MSA to MSA. For instance, the expiration of the expanded conforming loan limits would have a greater impact in Los Angeles, California than it would in Cleveland, Ohio. Granularity in the housing market generates a need to analyze each MSA independently in order to understand the how a change in a particular variable will affect a specific MSA or region. Low Home Prices The decline in home prices since 2007 has probably unleashed pent-up demand from would-be home buyers who were priced out of the market during the housing boom, particularly in the metropolitan areas with higher home prices. High Unemployment Rates Although unemployment has improved recently, it is still the most salient factor affecting the housing market. If unemployment were to spike it could further restrict consumer spending power and further restrict the demand for housing.

Outlook for Housing Supply The supply of homes is large and likely to increase in coming months. According to the National Association of Realtors (NAR), the supply of existing home sales in July was around 4 million units, and while that figure was lower than it was a year ago, it remained well above the historical level of 2 to 3 million units. The NAR Hayre, Lakhbir, and Robert Young. “Another Look at Home Prices: Optimism May Be Premature,” Topics in Mortgage Credit Markets Quantitative Analysis (10 Aug. 2009): 1-12. Citigroup Global Markets Inc. 4

www.radarlogic.com

7

numbers likely underestimate the total supply of homes as the count of realtor listed homes does not include much of the large and growing inventory of distressed homes. According to the recent analysis by Citigroup, there were approximately 3 million distressed units for sale in July. The same report indicates that 4 to 5 million loans are in various stages of serious (60 days plus) delinquency. While not all delinquent loans end up as foreclosed properties for sale - many are cured or modified - the inventory of foreclosed homes will certainly increase in coming months. Of course, the inventory of distressed homes will only put downward pressure on prices to the extent that they are offered for sale, and there is evidence that a significant percentage of bank-owned properties are being held off the market. RealtyTrac’s sampling of Q4 2008 data showed that as much as 70% of foreclosed homes owned by banks have not been listed for sale, though other industry estimates put that figure closer to 50%.5 By releasing their REO inventory onto the housing market slowly, banks are effectively reducing the impact of the distressed inventory on home prices. If the current trend continues and the demand for housing holds up, home prices could continue to recover despite the record number of foreclosures. Of course, the majority of foreclosed homes will have to be sold eventually, and when they are they will be sold at a discount to other homes. Therefore, the large overhang of distressed inventory is likely to prolong the recovery process even if it does not interrupt it.

Regional Outlooks   As has been discussed above and in previous reports, home prices and home sales have stabilized and are currently displaying a healthy rate of growth in many metro areas. However, an analysis of individual MSAs shows that, while the overall market is recovering, the recovery is heterogeneous and the potential for price appreciation and depreciation in coming months varies widely from one MSA to the next. MSAs in the Western region experienced some of the fastest price growth during the boom and some of the fastest price declines during the bust. It is therefore not surprising that MSAs in the Western region, Los Angeles in particular, have shown some of the first and strongest signs of recovery. Transaction volumes in Los Angeles, Exhibit 7 Los Angeles: Cumulative Sales Volume From January 1 to July 1 120,000

100,000

80,000

60,000

40,000

20,000

0 2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

See “Bank-owned homes surge, communities stung,” The Michigan Messenger (4 Mar. 2009) and “Inventory of Houses Falls in July,” The Wall Street Journal (5 Aug. 2009). 5

www.radarlogic.com

8

which bottomed in 2008, are now increasing in both percentage and absolute terms. Cumulative transactions over the first six months of the year were higher in 2009 than in 2008, and slightly below 2007 levels. They are currently at about 66% of volumes during in 2004 and 2005, the fastest growing period of the boom. The Las Vegas MSA has yet to show any signs of appreciation potential. Currently the Las Vegas RPX is down 55.2% from its peak in May 2006. While this may indicate that home values in Las Vegas have the potential to appreciate in the long term, price growth in the Las Vegas MSA will remain unlikely until motivated sales (i.e., sales of homes out of foreclosure) decrease as a percentage of total sales. As of July, 54% percent of all transactions in Las Vegas were motivated sales. The New York MSA is suffering from different ailments than other MSAs. The current uncertainty in the financial sector, the inaccessibility of jumbo loans and substantial unsold development create structural weakness in the New York housing market. Moreover, home prices in New York have fallen only 22% from their peak, less than most MSAs tracked by Radar Logic, and they likely have further to fall before they return to equilibrium. As a result, the lack of affordability may dampen housing demand, at least in the short term. Declining transaction levels are another signal that prices in New York have yet to bottom, especially when compared to many other MSAs where transaction counts are increasing. Exhibit 8 New York: Cumulative Sales Volume From January 1 to July 1 90,000 80,000 70,000 60,000 50,000 40,000 30,000 20,000 10,000 0 2000

www.radarlogic.com

2001

2002

2003

2004

9

2005

2006

2007

2008

2009

Mix Shift: Motivated v. Other Sales A shift in the mix of home sales away from motivated sales and toward other (i.e., not “motivated”) sales has contributed to the increase in home prices we have observed over the last few months. Since April,, the 25-MSA motivated transaction count6 has remained more-or-less constant at about 16,000 transactions, while the 25MSA other transaction count has increased considerably, from roughly 35,000 transactions to about 57,000 transactions (see Exhibit 9, and please note that the transaction bars for motivated and other transactions are stacked). Over the same period, the 25-MSA composite price for motivated sales has been 36% to 39% less than the 25-MSA composite price for all other sales. Due to this substantial premium in other prices relative to motivated prices, the increase in other transactions as a percentage of total transactions has put upward pressure on the 25-MSA RPX Composite, which reflects pricing in both motivated sales and other sales. As a result, the RPX Composite has increased and started to converge with the composite price for other sales. Exhibit 9 Composite Motivated TC

Other TC

Motivated Price

Other Price

MSA Price 100,000

$300

90,000 $250 80,000

70,000

60,000

$150

50,000

40,000

Transaction Count

Price Per Square Foot

$200

$100 30,000

20,000 $50 10,000

$0

0

Transaction Date

As mentioned above, banks have only put a fraction of their inventories of foreclosed homes on the market. Perhaps institutional constraints have prevented them from selling enough of their REO inventory to meet the seasonal uptick in demand. If so, there could be a surplus of demand for homes at motivated prices, which could insulate motivated transactions from a seasonal decline in the autumn and winter. If this is true, other transactions should decrease by more than motivated transactions and the mix of sales should shift back toward

Transaction counts reflect sales used in the calculation of 28-day RPX Daily Prices. As such they reflect transactions over a 28-day period ending on the given transaction date. Transaction counts may not reflect sales volumes in the market. 6

www.radarlogic.com

10

motivated sales. To the extent that foreclosed homes sell at a discount to other homes, this mix shift could put downward pressure on MSA-wide prices. However, aggregations of motivated and other prices do not tell the whole story. In many cities, such as Los Angeles (see Exhibit 10), sellers have cut their prices significantly to compete with financial institutions selling homes out of foreclosure. When one looks at the distribution of sales across the range of prices per square foot, one finds that “other” and motivated sales mostly occur at the same price levels. In 15 of the 25 MSAs tracked by Radar Logic, over half of all homes sold in July were sold within one standard deviation of the mean motivated price. In these cities, a shift in the mix of sales toward motivated or other sales may not have a significant impact on pricing. Exhibit 10

Los Angeles transactions for the 28-day period ending July 23, 2008 Mean Motivated Price ($199.69) +/- One Standard Deviation ($97.90) 600

Motivated

Other

57% of all transactions had a price per square foot within one standard deviation of the mean motivated price.

Number of Sales

500

400

300

200

100

$0 $20 $40 $60 $80 $100 $120 $140 $160 $180 $200 $220 $240 $260 $280 $300 $320 $340 $360 $380 $400 $420 $440 $460 $480 $500 $520 $540 $560 $580 $600 $620 $640 $660 $680 $700 $720 $740 $760 $780

0

Price per Square Foot

www.radarlogic.com

11

Source: Radar Logic

Volatility in RPX Fixings Since June, pricing for the RPX Composite forwards has increased at a rapid pace. Contracts maturing in 2009, which traded around $160 in June, are now trading around $190. As can be seen in Exhibit 11, not only has 2009 contract pricing been driven up as the RPX 25-MSA Composite has increased, but the discount to spot has decreased from over 15% to around 4%. This would indicate that the institutions participating in the RPX market do not expect significant price declines before the contract settlement date at the end of December. Exhibit 11

2009 Forward Pricing vs. Spot Price 2009 Discount to spot

2009

$205.00 

Spot ‐16.00%

$200.00 

‐14.00%

$195.00  ‐12.00% $190.00  ‐10.00%

$185.00 

$180.00 

‐8.00%

$175.00 

‐6.00%

$170.00  ‐4.00% $165.00  ‐2.00%

$160.00 

$155.00 

0.00%

Source: Radar Logic

Volatility in the 2009 contracts has been increasing since the beginning of the year. Since January, the average daily price change has been 14 cents a day. Since June, the average daily price change has been 52 cents a day. These amounts annualized for trading days account for a 278% increase in average volatility, which provides ample trading opportunity. Exhibits 12 and 13 show similar trends of convergence to spot and increased volatility in the 2010 and 2011 RPX Composite forwards.

www.radarlogic.com

12

Exhibit 12

2010 Forward Pricing vs. Spot Price 2010 Discount to Spot

2010

Spot ‐30.00%

$205.00 

$195.00  ‐25.00% $185.00  ‐20.00% $175.00 

‐15.00%

$165.00 

$155.00  ‐10.00% $145.00  ‐5.00% $135.00 

$125.00 

0.00%

Source: Radar Logic

Exhibit 13

2011 Forward Pricing vs. Spot Price 2011 Discount to Spot $205.00 

2011

Spot ‐25.00%

$195.00  ‐20.00% $185.00 

$175.00 

‐15.00%

$165.00 

‐10.00%

$155.00 

$145.00  ‐5.00% $135.00 

$125.00 

0.00%

Source: Radar Logic

www.radarlogic.com

13

Exhibit 14 depicts the actual 2009 contract price movement and a trend line representing the low-to-current increase of the contract price. At the time this report was written, investors would have had an unrealized return of 36% if they had purchased the 2009 contract at the 52-week low of $142. Exhibit 14

2009 RPX Forward Contract Movement Since 52‐Week Low 2009 Contract

2009 Point‐to‐Point Growth Since low

$205.00 

$195.00 

$185.00 

$175.00 

$165.00 

$155.00 

$145.00 

$135.00 

Source: Radar Logic

www.radarlogic.com

14

RPX Forwards: Rich or Cheap? Historical RPX data show that the RPX Composite declines by an average of 2.5% each year between July and the end of December. However, the average decline may not be a useful predictor of the settlement price of the RPX this year due to the unusually large movements in home prices over the last decade. Exhibit 15 illustrates this by showing the entire price history of the RPX Composite. Orange markers represent July and vertical lines provide a frame of reference for the end of the year. As can be seen, some years show an increase from July to December and others show a decline. This variation is created by the boom and bust cycle that has played out over the last decade. As the peak of the housing boom, 2006 was the only year in the last ten that was characterized by neither large price gains nor rapid price declines from January to December. As such, 2006 was an exception to the boom or bust dynamics that characterized the rest of the decade and represented a return to a more stable seasonal pattern, including both modest strength in the spring and limited weakness in the fall. As the trough of the housing boom, 2009 resembles 2006 because it marks a shift back to normal growth after a period of rapid change. Therefore it is reasonable to expect that declines in the RPX Composite between July and December 2009 will resemble the declines over the same period in 2006, about -4%. If this is the case, then currently 2009 contracts are priced correctly and 2010 and 2011 contracts are at bargain prices. Exhibit 15 RPX Composite Price History $300.00 

$280.00 

$260.00 

$240.00 

$220.00 

$200.00 

$180.00 

$160.00 

$140.00 

$120.00  3‐Jan‐00

3‐Jan‐01

3‐Jan‐02

3‐Jan‐03

3‐Jan‐04

3‐Jan‐05

3‐Jan‐06

3‐Jan‐07

3‐Jan‐08

3‐Jan‐09

Source: Radar Logic

www.radarlogic.com

15

July 2009 S&P/Case-Shiller Composite Home Price Indices In the RPX Housing Market Report for June 2009, which we released August 20, we used our Daily Prices to predict the S&P/Case-Shiller 10-City and 20-City composite home price indices for June. We predicted the 10City index would be approximately 155, and the 20-City index would be about 143. The actual values, released July 28, were 153.2 and 141.86, respectively. This month, we expect the S&P/Case-Shiller composites to increase again relative to their June levels. The July 2009 10-City composite will be roughly 155 and the 20-City composite will be roughly 144.

Developments in the Housing Market By now, the housing recovery is old news. For the past few months many macro indicators in addition to the RPX have posted positive figures and confirmed our observation that housing markets have stabilized. Of importance moving forward is the capability of the current economic environment to sustain a housing recovery in the wake of many unknowns. Mounting evidence is pointing to an upward trend in home sales. The National Association of Homebuilders posted a third straight gain for their Housing Market Index, which is a survey of builder opinions about present sales of new homes, sales of homes in the next six months and prospective new home buyers. The interest rate for a 30-year Freddie Mac mortgage, which averaged about 5.21% in August, currently averages 5.04%. Although still far from their 2006 levels, housing starts rose 1.5% in August and permits were up 2.7%, both of which are indications of future strength in the housing market. Also of importance to the future of housing is the jobless rate, which has been declining in September. The pending and existing home sales indexes produced by the National Association of Realtors have posted increases and NAR has reported that raw inventory figures are 10.6% lower than a year ago. At the current rate of sales, the existing home inventory at the end of July is down to a 9.4 months’ supply. While still high by historical standards, these figures show a considerable improvement over the last six months. Overall, the current state of the housing sector is definitely better than it was six months ago. However, a great deal of uncertainty still surrounds whether or not housing markets can sustain their current strength and how the market will react when we move closer to the traditional winter down-season. The following calendar contains upcoming economic and housing market data releases in September and October 2009. Housing Release Calendar Release

Release Date

Standard & Poor’s: S&P Case-Shiller Housing Price Indices for April 2009

September 29, 2009

U.S. Dept. of Commerce: Second Quarter 2009 Gross Domestic Product

September 30, 2009

National Association of Realtors (NAR): Pending Home Sales U.S. Dept. of Labor: Employment Situation Summary

October 1, 2009 October 2, 2009

U.S Census Bureau: Retail Sales

October 14, 2009

U.S. Dept. of Labor: Consumer Price Index

October 15, 2009

National Association of Home Builders (NAHB): Housing Market Index

October 19, 2009

U.S. Dept. of Labor: Producer Price Index

October 20, 2009

U.S Census Bureau: Housing Starts

October 20, 2009

National Association of Realtors (NAR): Existing Home Sales

October 23, 2009

www.radarlogic.com

16

Exhibit 16: 25 Metropolitan Statistical Areas (MSAs, Ranked by 1-Year % Change) July 2009 Rank

June 2009 Rank

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

9 5 7 6 3 1 2 8 4 10 15 11 17 13 14 19 20 12 16 18 22 23 21 24 25

PPSF

July 2009 vs. July 2008

July 2008 vs. July 2007

July 2009 vs. June 2009

July 2008 vs. June 2008

Boston, MA Denver, CO Columbus, OH Washington, DC Milwaukee, WI Charlotte, NC Philadelphia, PA St. Louis, MO1 Cleveland, OH Los Angeles, CA Atlanta, GA San Diego, CA San Jose, CA New York, NY Seattle, WA Jacksonville, FL San Francisco, CA Minneapolis, MN Sacramento, CA Tampa, FL Chicago, IL Miami, FL Detroit, MI Phoenix, AZ Las Vegas, NV

$212.84 $133.16 $95.88 $186.94 $116.50 $93.66 $144.51 $103.69 $80.74 $256.95 $85.32 $203.96 $336.11 $242.77 $180.93 $94.54 $278.19 $121.22 $126.16 $91.81 $147.21 $119.27 $72.66 $84.55 $83.65

-0.8% -2.4% -3.2% -4.4% -5.3% -5.4% -5.6% -6.7% -9.1% -9.4% -11.4% -11.6% -13.3% -13.5% -15.1% -15.5% -16.0% -16.8% -17.8% -18.2% -20.1% -20.6% -22.1% -24.2% -28.5%

-12.5% -10.0% -1.3% -17.4% 0.4% -1.9% -3.4% -5.0% -7.6% -27.9% -7.3% -26.2% -17.5% -8.2% -9.1% -10.7% -26.7% -8.5% -27.8% -18.1% -3.1% -23.2% -12.6% -29.9% -32.1%

5.5% 2.1% 1.5% 0.9% -2.2% -3.2% -0.9% -2.0% -3.2% 0.0% 3.5% 0.0% 3.4% 0.2% -1.8% 1.9% 1.6% -2.5% -0.4% 0.0% 1.6% -0.9% 0.2% -0.6% 0.2%

-4.2% -1.4% -1.5% -0.5% -1.9% -0.3% 1.5% -3.3% 0.6% -2.7% -2.0% -1.9% -2.5% -0.5% -2.1% -2.4% -2.5% 1.1% -0.9% -0.6% -0.5% -3.3% 3.1% -4.4% -4.2%

Manhattan Condo2

$961.59

-15.9%

1.9%

MSA

Source: 28-Day RPX value for each MSA as of 7/23/2009 1 Historical prices used to calculate changes in St. Louis include late-arriving data not included in published series 2 Manhattan Condo is a subset of the New York MSA

0.6% = positive

2.9%

= neutral

= negative

Exhibit 17: Metro Areas Ranked by 2-Year and 5-Year Annualized Change Leading 5 Metro Areas (2-Year Annualized % Change)

Rank

Trailing 5 Metro Areas (2-Year Annualized % Change)

MSA

% Change

Rank

MSA

1

Columbus, OH

-2.2%

1

Las Vegas, NV

-30.3%

2

Milwaukee, WI

-2.5%

2

Phoenix, AZ

-27.1%

3

Charlotte, NC

-3.7%

3

Sacramento, CA

-22.9%

4

Philadelphia, PA

-4.5%

4

Miami, FL

-21.9%

5

St. Louis, MO

-5.9%

5

San Francisco, CA

-21.6%

Leading 5 Metro Areas (5-Year Annualized % Change)

% Change

Trailing 5 Metro Areas (5-Year Annualized % Change)

Rank

MSA

% Change

Rank

MSA

% Change

1

Seattle, WA

2.9%

1

Las Vegas, NV

-11.6%

2

Philadelphia, PA

2.6%

2

Detroit, MI

-9.4%

3

Milwaukee, WI

2.4%

3

Sacramento, CA

-9.3%

4

Charlotte, NC

1.2%

4

San Diego, CA

-8.9%

5

New York, NY

0.4%

5

San Francisco, CA

-4.7%

Source: 28-Day RPX™ analytics as of 7/23/2009

www.radarlogic.com

17

Exhibit 18: Transaction Counts1 MSA

July 2009 vs. July 2008

July 2008 vs. July 2007

July 2009 vs. June 2009

July 2008 vs. June 2008

Philadelphia, PA

82.5%

55.3%

24.0%

-5.1%

Chicago, IL

36.6%

-28.4%

10.9%

-0.7%

San Jose, CA

28.2%

-11.0%

-5.0%

-13.4%

Miami, FL

18.7%

-27.5%

-3.4%

-13.6%

Boston, MA

14.3%

-17.5%

35.0%

-8.8%

Seattle, WA

11.9%

-45.1%

-2.4%

-18.4% -7.9%

Phoenix, AZ

9.7%

7.9%

-5.5%

Los Angeles, CA

9.4%

16.4%

-3.1%

0.2%

San Diego, CA

8.6%

12.4%

-4.0%

-7.6%

Tampa, FL

7.4%

-19.0%

-0.2%

-10.9%

Jacksonville, FL

6.0%

-28.5%

8.5%

-11.2%

Sacramento, CA

4.2%

31.3%

-0.4%

-7.4%

Washington, DC

4.0%

19.0%

-8.9%

-6.1%

San Francisco, CA

3.8%

-0.4%

4.6%

-4.2%

Milwaukee, WI

2.2%

-21.2%

17.1%

-12.9%

New York, NY

1.9%

-8.3%

14.5%

7.7%

Cleveland, OH

1.3%

-28.6%

-5.9%

-15.0%

Atlanta, GA

0.9%

-41.0%

11.3%

-5.2%

Detroit, MI

0.8%

-21.0%

18.5%

-2.7%

Las Vegas, NV

-3.7%

33.9%

-7.2%

5.7%

Minneapolis, MN

-3.8%

-8.4%

0.9%

6.5%

Columbus, OH

-7.0%

-23.4%

-0.8%

-14.7%

St. Louis, MO2

-12.3%

-51.9%

-17.8%

-37.0%

Denver, CO

-13.9%

-11.0%

-0.3%

-3.0%

Charlotte, NC

-29.7%

-24.5%

-0.9%

-10.6%

Manhattan Condominium

-37.4%

-32.7%

18.8%

9.0%

Source: 28-Day RPX™ analytics as of 7/23/2009 1 Transaction counts represent the transactions included in the calculation of the RPX Daily Prices and may not reflect transaction volume in the market. 2 Historical transactions used to calculate changes in St. Louis include late-arriving data not included in published series

Exhibit 19: Transaction Counts: Motivated3 vs. Other Sales July 09 % Motivated Sales

July 08 % Motivated Sales

Composite

20.6%

20.5%

7.4%

6.4%

-8.5%

5.5%

Los Angeles

27.4%

34.0%

-11.9%

20.3%

-9.9%

-0.2%

Miami

15.7%

13.0%

43.1%

14.9%

-14.4%

-1.1%

4.9%

2.7%

87.8%

-0.5%

-3.7%

15.6%

35.9%

33.1%

19.0%

5.0%

-14.3%

0.3%

New York Phoenix

July 09 vs. July 08 T.C. Change (Motivated)

July 09 vs. July 08 T.C. Change (Other)

July 09 vs. June 09 T.C. Change (Motivated)

Source: 28-Day RPX™ analytics as of 7/23/2009 Radar Logic defines motivated sales as foreclosure auction sales and liquidity-driven sales by financial institutions and foreclosure service firms.

3

www.radarlogic.com

18

July 09 vs. June 09 T.C. Change (Other)

Exhibit 20: Tradable MSAs

Composite Motivated TC

Other TC

Motivated Price

Other Price

MSA Price 100,000

$300

The following graphs contain MSA price data and transaction counts available from Radar Logic.

90,000 $250 80,000

70,000

60,000

50,000

$150

40,000

Transaction Count

Price Per Square Foot

$200

$100 30,000

20,000 $50 10,000

0

$0

Transaction Date

Los Angeles Motivated TC

Other TC

Motivated Price

Miami Other Price

MSA Price

Motivated TC 7,000

$450

Other TC

Motivated Price

Other Price

MSA Price 6,000

$250

$400 6,000

5,000 $200

$350 5,000

$200

3,000

$150

3,000

$100

$150

Transaction Count

$250

Price Per Square Foot

4,000

4,000

Transaction Count

Price Per Square Foot

$300

2,000 2,000

$100

$50 1,000

1,000 $50

$0

0

0

$0

Transaction Date

Transaction Date

New York Motivated TC

Other TC

Motivated Price

Phoenix Other Price

$350

MSA Price

Motivated TC 8,000

Motivated Price

Other Price

MSA Price 7,000

$180

$160

7,000

$300

Other TC

6,000 $140

6,000 $250

5,000

$150 3,000

4,000

$100

$80

3,000

$60 $100

2,000 2,000

$50

1,000

$0

1,000 $20

$0

0

Transaction Date

0

Transaction Date

Source: 28-Day RPX™ analytics as of 7/23/2009

www.radarlogic.com

$40

19

Transaction Count

4,000

Price Per Square Foot

$200

Transaction Count

Price Per Square Foot

$120 5,000

Composite

Exhibit 21: RPX Forwards: Historical Fixings

RPX Composite Price

May 21

July 21

September 21

$300

Price fixings are established each trading day by a dealer poll and represent the midmarket expectation for the reference value to be published on the contract expiration date. Reference values represent the simple average of the 28-day RPX Daily Prices from the last five publication dates of each quarter (which correspond to transaction dates 63 days earlier). For the following charts, the RPX prices are plotted on a publication date basis. The names of the series indicate the dates in 2009 those price fixings were published.

$250

$200

$150

$100

$50

$0

Miami

Los Angeles RPX Los Angeles Price

May 21

July 21

RPX Miami Price

September 21

May 21

July 21

September 21

$250

$450 $400

$200

$350 $300

$150

$250 $200

$100

$150 $100

$50

$50 $0

$0

Phoenix

New York RPX New York Price

May 21

July 21

RPX Phoenix Price

September 21

May 21

July 21

September 21

$180

$350

$160

$300

$140

$250

$120

$200

$100 $80

$150

$60

$100 $40

$50

$20 $0

$0

Exhibit 22: Forward Contract Implied HPA (Cumulative) as of 9/21/2009 28-day RPX on 7/23/2009

Dec 09

Dec 10

Dec 11

Dec 12

Dec 13

25 MSA Composite

$198.40

-2.7%

-4.0%

-3.7%

-3.5%

-3.2%

Los Angeles, CA

$256.95

-21.8%

-21.8%

-21.8%

N/A

N/A

Miami, FL

$119.27

-37.1%

-37.1%

-37.1%

N/A

N/A

New York, NY

$242.77

-27.9%

-23.8%

-23.8%

N/A

N/A

Phoenix, AZ

$84.55

-23.1%

-23.1%

-23.1%

N/A

N/A

Source: Official 28-Day RPX fixings as of 9/21/2009

www.radarlogic.com

20

About Radar Logic Radar Logic Incorporated, a real estate data and analytics company, calculates and publishes the Radar Logic Daily™ Prices. The prices track housing values for major U.S. metropolitan areas and are the basis of the Residential Property Index™ (RPX™), a market that enables real estate to be traded as a liquid asset, via property derivatives marketed by major financial institutions. RPX allows real estate and financial professionals to manage opportunity and risk, invest in real estate values without owning physical assets and effectively analyze markets using a consistent metric: price per square foot. Data in the RPX Monthly Housing Market Report reflect the 28-day aggregated value of Radar Logic Daily Prices. The price per square foot metric used significantly reduces the influence of property sizes on overall housing price trends, which can skew results. The Daily Prices for each MSA are not adjusted for seasonal variations. In some cases, Daily Prices may vary based on reporting characteristics within individual MSAs. The RPX Monthly Housing Market Report provides insight and detailed analysis of Radar Logic’s 25 MSAs and the Manhattan Condo market. This study is based on the premise that there is not a national housing market; rather, each MSA, while having some economic influences in common, is influenced primarily by local conditions. The August 2009 RPX™ Monthly Housing Market Report will be released on October 22, 2009, at 12:01 AM EDT.

RPX Analytics & Research Radar Logic offers specialized analytic services which allow real estate and financial professionals to view current and historical price per square foot and transaction count trends for all markets and sub-markets we track. MSAs can be segmented by location (zip code and county), property type (single family, multi-family and condo), property size, date range, and sale price. The database is derived from our neutral, public source records. Our data provide a means for all entities associated with or affected by housing prices to maintain market data streams on a constant, neutral and daily updated basis. For additional insight on this report or for inquiries about research or analytic products, please contact: Radar Logic Incorporated 180 Varick Street, Suite 502 New York, NY 10014 212.965.0300 [email protected]

© 2009 Radar Logic Incorporated. All Rights Reserved. Data presented “AS IS”. Radar Logic does not make, and hereby expressly disclaims, any representation or warranty of any manner in connection with the information including, without limitation, with respect to its accuracy, completeness or fitness for any purpose. www.radarlogic.com

21