home tags Search Feed
Myths, Markets & Manipulators: The Real Deal on the Homebuilders

Written by Reggie Middleton

Tuesday, 06 November 2007

This is an update of my analysis of Ryland and the US residential real estate market. If you recall, we have updated my land value forecasts to be more realistic and aligned with the aggressive fall in all residential housing classes (not just previously owned, detached single family housing as is found in the Case-Shiller index). The original posts were "What does Reggie Middleton and Ryland's upper management have in common? They are both selling shares faster than no doc loans get approved!" and "Okay, I have just recharged the batteries in my crystal ball: Back tested Home Price Trends - Historical and Forecasted". To get the full picture of my research on Ryland, you should read the two previous posts, for there is pertinent info that I did not include in this piece to avoid redundancy.

For those that are new to the blog, I have covered the effects of the real estate bust in detail, and its effects on banking, building and the macro scene. Please choose a category that interests you and browse through the archives. If you are unfamiliar with the homebuilding industry (which I feel is THE barometer for residential housing in the US), then see "Thoughts on the US Publicly Traded Homebuilders" as a primer and the informative "CFO" series as a more advanced course (you can start with "A fly on the wall - straight talk from the homebuilder CFO"). The "Bubbles, Banks and Builders" series is a more analytical take on individual companies and their macro situations.

Now, let me start off by clearing the air so there is no misunderstanding. I am short the real estate industry, in general. That is builders, insurers, financiers and thrifts, ancillary services, etc. Now, you know where I am coming from. I invested in residential real estate and studied the trends and macro/microeconomics of the asset class fastidiously. That does not make me an expert, but it does make me an eager student. There are many who offer advice and/or invest in this asset class and its derivatives, not as a student, but purely as a speculator or equity/debt analyst. Well, that is all well and good, except for the serious mistakes that can ensue from not knowing residential real estate trends and valuation. Let me be clear - We are in the BEGINNING throes of a very, very serious real asset and land recession (read land recession pt II as well) - one of the most serious in recent history, if not ever.

----- EXTENDED BODY:

It has already started manifesting itself in residential real estate with fury caused by the "The Great Global Macro Experiment" and is producing prodigious amounts of REOs and foreclosures, but will move forward to commercial and retail real assets. It has already manifested itself in the debt markets, spread geographically on a global basis - but this is just the start. It has also manifested itself in the insurance markets as well (reference MBI, ABK, RDN, MTG - down by up to 50% in mere weeks), but this is just the start. Be sure to read the links that are dispersed throughout this missive - they should be educational and definitely worth your time.

Why state what should be the obvious to many? Because, many don't get it. I hear advice such as, "Is now the time to dabble in the home builders?" or the home builders have hit a bottom. These people just don't seem to understand, or at least see things from my perspective. The homebuilders, the lenders and the insurers are the industries at ground zero of one of the worst burst bubbles that we have seen in some time. The home building industry is a staid cyclical industry that garnered practically no sell side analyst coverage until we had a residential real estate bubble. Now, it appears as if many actually believe the industry to be a growth industry, in lieu of a cyclical industry and are advising accordingly. What we have experienced over the last 8 years was a bubble, not a normal business cycle. We will probably never see performance numbers like that again in this lifetime. It's gone, outta here, vamoosed, vanished... Let me state it again different terms. The real estate boom has nearly doubled gross margins for the builders. Net margins are now negative in the bust. For those builders who survive the bust there will be a prolonged lull in building then a return to the mean, which will be a staid, cyclical 12% to 16% GROSS margin and low single digit net margins - no more multi-billion+ dollar quarters. It appears that many investors fail to study their history and believe that the housing bubble performance of the builders makes them a growth industry. As you will easily see, they are not. They are a average yielding slow growth cyclical industry - very similar to the utilities, albeit a bit more risky. Even if the homebuilders were to somehow make all of their current problems vanish into thin air (nigh impossible, but let's just imagine), they would still take 2 to 10 years to return to single digit net margins, not the 30+% returns that we have seen during the boom.

Now, after reading the many posts in this blog on the topic, you may realize that I don't think the builders problems are going to just vanish into thin air. The homebuilders have a big, big problem. Those who may be attempting to bottom fish at the beginning of a severe real estate bust either don't understand RE cycles or are totally underestimating the severity of this current cycle. Graphing the trends makes it obvious, so let's take a look at some pretty pictures that illustrate the historical trends of Ryland and the US residential real estate market (click any graph to enlarge).

1992 was the last significant down period for Ryland. It shows as a near imperceptible blip in the land value graph (primarily CA and the Mid-Atlantic states), yet dropped Ryland's margins by 50% and lasted about 5 years. Notice how small the blip is in the graph - seriously, the size of the blip is important - for if you compare it the latest boom in 2000 - 2006 you can see the severity of the current situation.

Looking at this graph, one can see how minor blips in margins cause leveraged and extreme fluctuations in Ryland's share price. Looking at what we just went through in the last 7 years, starting around the half way mark of 1999 - BIG difference! Ryland's share price trend spiked towards the sky with the ever so slight but increasing trend in margins, and now that margins are negative the share price trend is dropping like a meteor from the sky.

These charts give us a magnified view of the correlation between gross margins, net margins, housing price trends and Ryland's Z (bankruptcy) score. More pessimistic (and realistic) than my previous analysis, Ryland is well entrenched in the 72% probability range of going bankrupt within 8 quarters, with about 8 quarters of negative net margins - yet they are buying back company stock as management is selling their stock. Amazing!!! At least somebody is reading my blog, though. Moody's is considering downgrading Ryland to junk status.

Revenue Timeline

Ryland has been in business for a long time, and has endured real estate booms and busts before - but not to the extent of what we are currently going through. They, like many other builders, banks, and monoline insurers drank from the cup of easy credit and euphoric price increases too heavily and for too long. Reference the following chart to see a comparison of the last few troughs compared to this one.

Looking at historical and projected revenues, you can see what I mean by a reversion to the mean (no pun intended). Ryland's revenues (if they survive) will stop tanking at just about the point where the boom began. Thus, all of that artificial profit and margin will be nearly erased, as it always is in a boom/bust scenario - and business will continue along at the inflation adjusted norm - again, that's if Ryland is still around to see this occurrence.

I have focused this historical analysis on Ryland, but it is applicable to nearly all of the public builders. Compare the graphs above to this one of the major builders and you can easily see where I am coming from. As a matter of fact, after scrubbing the books with a more realistic view of housing values and apply the average price to book valuation of the major builders to Ryland, one finds that Ryland is trading at about a 100% premium to its peers. Does this company really deserve such a valuation? Obviously, management doesn't think so, at least for thier personal holdings, and neither do I.

The Ryland Group Inc.

RELATIVE VALUATION

Company	Bloomberg ticker	P/B
		2006	2007E	2008E	2009E

Lennar Corporation	LEN US Equity	0.7	0.8	0.8	0.7
D.R. Horton, Inc	DHI US Equity	0.7	0.8	0.7	0.7
Pulte Homes	PHM US Equity	0.6	0.7	0.7	0.6
KB Home	KBH US Equity	0.8	0.8	0.9	0.8
M.D.C. Holdings, Inc.	MDC US Equity	0.9	1.0	1.0	1.0
RYL Group	RYL US Equity

Industry Average		0.7	0.8	0.8	0.7
Source: Bloomberg

Relative P/B Valuation
Ryland Group	2008
Book value per share	16.72
Industry P/B	0.82
Target Stock Price (US$)	13.66

Ryland closed at $26.28 on Nov. 5th, down from nearly $30 the week before. Just do the math...

So where does Ryland stand now?

Not very well in my opinion. Then again, I am selling the stock short, while Ryland management is just selling the stock. Let's take a quick look at my researched opinion on the company.

Land Sales - In an attempt to shore up cash, the builders (and Ryland as well) have been trying to sell land as well as houses. The problem is they have paid so much for the land with so much debt that they cannot practically offer it at at price that anyone with a brain is willing to pay. The key word here is practically, not profitably. In a distressed sale, it all goes for pennies on the dollar and the true devaluation game begins (Tousa's pre-packaged bankruptcy affair). Thus far, Ryland has $8.3 million of land sales (Q307) vs. $37.4 million of land sales the year before (78% drop). The gains on land sales last year was 19.52% while the gains on land sales this year was 3.54% (an 82% drop in profitability). Since they probably used the FIFO method of accounting for the gains, it is safe to say that this will be the last quarter for some time that they can report a positive number for land sales gains due to the fact that the real overpriced property hasn't moved yet at least from an accounting perspective.

Cash Generation - Ryland has had a very hard time turning over inventory to generate cash. I illustrated this in detail last month, and it is now being publicized by Moody's. I modeled a big drop in cash this quarter, even though their latest press release states that they have generated positive cash for the quarter enabling them to pay down $36 million in debt. We will get the details when the 10Q is released. If the large shortfall doesn't occur this quarter, it will occur in the next quarter or two. The reason??? They have to start building more houses in order to attempt to monetize their raw land. As shown above, they aren't very successful at selling the land to generate cash, thus the only other way to generate anything (even losses) is to sell houses - which must be built in order to be sold. This takes money. Last quarter, according to their press release they generated $42.6 million of cash flow from operations which they used to pay down $36.9 million of debt (out of $1.5 billion - that's right, BILLION). That leaves $5.7 million of cash left over after debt service and paydown. That's a small amount of money for a multi-billion company. I also think it is misleading, due to management trying to rob Peter to pay Paul. To approach from a different perspective, RYL has generated $17,074 of cash per each closing they made, on average (and macro conditions are getting much worse, much faster). This is abysmal performance. Look at it from this point of view. The real estate broker makes $15,600 (gross) @ 6% on each $260,000 house sold - without the excessive debt, construction, time lag, and market risk. Now I am hopeful that this builder negotiates better commissions, but word on the street is that the builders are actually paying higher than average commissions to garner extra marketing interest. Some builders market exclusively in-house (requires additional upfront cash), some outsource, and some do both. Back to my point and extrapolating further, Ryland will need 53,974 closings to pay down their existing debt at their current pace or 24 quarters. I foresee them having 8 quarters to get their act together as a going concern before having to do the Tousa (no, it's not a new dance).

New orders (net) will not be positive until about 2nd quarter of 2008

Closings will not trend upward reliably until 2009.

Order cancellations will continue to trend upward until around 2009, with Texas leading the way to positive territory before the rest of the regions.

Average closing prices are geographically specific, but in general, don't look good until 209, again with the exception of Texas. California, Nevada and southern Florida are in for some pain, with Cally leading the way.

This is the formula that determines the cash in the builder's pipeline in terms of days of backlog to sell. The higher the number the better. Ryland will be in danger of running out of sales, for some time. This contributes heavily to a low Z score.

Mortgage origination will drop because of less houses being sold in addition to a tighter mortgage market. There is still risk that the builder could bet stuck with unsalable mortgages in its warehouse credit facility. Financing was a profit center for builders with a high margin, look for it to become a possible drag on earnings and source of unhedged risk.

In aggregate, revenues will not break positive territory until 2009, but will trend upward through negative territory in 2008. Unfortunately, negative revenue growth is bad, even though it may be trending towards the positive. The wait 'til '09 will bankrupt a few builders.

Margins look to be negative for some time, excluding Texas. This portends misery for equity investors of the builders. I've pasted the graph that illustrates the correlation between margins and share prices below the margin projection graph to give you an idea of what could happen even a couple of years out into a recovery.

It looks bad if history is any indication of the future...

In many regions, it even looks bad if you exclude impairments and writedowns...

Assumptions going into our Ryland model

Overall region-wise outlook	Among all the four regions of RYL, we expect the west to experience the most decline and Texas to experience the least decline. For both southeast and northern regions, we expect moderate declines in the new orders
	West: We expect a significant decline in RYL's western region, led primarily by the the ongoing slump in its Californian and Las Vegas markets. Both these regions have bore the brunt of the overall nation-wide housing slump owing to their similar economic and demographic dynamics as that of the US nation. We expect RYL's Western region to witness the severest decline in housing market among all its four regions of operations
	Southeast: We expect some decline in RYL's Southeast region, as all the areas of operations of RYL in southeast are witnessing a housing slump. However, we predict a relatively less downfall in RYL's SE region as compared to its Western region (with the exception of the Southern Florida area)
	Texas: We expect the Texas region to be the strongest among all of RYL's operative regions. This is primarily owing to the region's high lack of correlation to the national housing sector. Historically, Texas gained relatively less in times of US housing boost and hence should loose relatively lesser in times of slump. There is always the risk of the abusive use of exotic financing instruments that may exacerbate a downtrend, but it should still be relatively less than most of the country. Further, the state's strong job market and export economy provides it savings against the nationally falling US industry
	North: We expect RYL's north region to witness moderate decline since most of the states within those regions did not experience the boom as experienced by other states and hence, we believe the north region would witness a relatively lesser decline


Gross margin estimates

West	The previously red hot markets of California, Las Vegas and Phoenix are now cooling off. During boom times, these regions witnessed the highest growth, as also reflected in their high gross margins historically.
	In line with the national housing growth, these regions are now facing the severest problems - in terms of highest price declines and sluggish home sales. We expect the trend to continue, especially in view of the industry leading growth in foreclosure rate in California and Nevada
	Notably, around 42.5% of the inventory impairment written off in 2Q2007 was in California, 31.5% in Las Vegas, NV, and 15.0% in Phoenix, AZ - a sum of US$113 million, 89.0% of the total inventory impaired in 2Q2007. This is primarily owing to the steepest price decline in these markets
	Though RYL has only 12% of its assets in California which should be a slight saving grace against the falling prices, we expect the region to be worst hit among RYL's other regions and estimate a dip of 1.5% in 2008, post which we assume slight recovery in the region's gross margin
	On the other hand, had RYL already had bought a substantial amount of land in California, it would have returned to normal gross margins relatively earlier. However, since only 12% of its assets are in California, we have not assumed slight normalcy in the region's gross margin before 2010
	Further, California ranked at the bottom in the NAHB's House Opportunity Index (HOI) released in August 2007 indicating lower housing affordability and thus lower prospective sales
	As per Los Angeles County Economic Development Corporation (LAEDC), housing permits in California, which declined 1.9% and 21.5% in 2005 & 2006, respectively, will decline a furhter 27.2% in 2007 and 4.5% in 2008 before it will recover with a 7.0% growth in 2009

Southeast	We expect all states/ cities of RYL's Southeastern region to witness significant decline in the near term, albeit we expect it to be moderate than in the West
	While Atlanta, GA has limited scope owing to the presence of a large number of small builders, both NC-SC are following the US national housing industry's trend, reflecting a downturn in the coming quarters
	Consequently, we have estimated a decline in RYL's Southeast operations till 2009 when it reaches 4.2% and grows to 5.2% in 2010, the margin in 2009 (4.2%) is still higher than that of West in 2009 (3.7%)
	While SC is witnessing an overall decline in the number of homes sold and their prices, the trend is uneven across cities. Despite the overall decline, only five of the 15 reporting regions showed significant declines in sales in 2Q2007. Though we expect South Carolina to recover early owing to the uneven housing trend in the whole state, we believe RYL would be affected by this decline for a relatively longer period and witness a substantial decline in new home sales in the near term, led primarily be huge declines in Charleston and Myrtle Beach areas
	Interestingly, as per Charleston Metro Chamber of Commerce's Developers Council, the wages of people in Charleston are not rising as fast as the house prices leading to reduced affordability and hence, lower demand for new houses

Texas	Texas has never witnessed significant growth, even during the boom times in US housing industry and consequently, the state should not witness a steep decline in times of downfall
	Further, among leading housing states in the US, Texas has relatively expensive land availability, limiting the state's ability to achieve high gross margins
	Consequently, while we expect the region's gross margins to decline slightly in 2007 and 2008 assuming an adverse effect of the nation-wide decline, we expect faster recovery in the area

North	We expect RYL's Northern region to witness some decline, however, it should be more moderate than that in West and Southeast
	The region had relatively more stability than other RYL regions indicating a relativly less decline in times of slump


Price expectations	For price building, we have used five key US home prices indices, blended together to create a proprietary, financially engineered indexing engine.
	For each region, we have identified key representative states/ cities, whose price indices are available, as follows:
	North: Chicago, IL and Washington, D.C. - since they are the bigger cities in that region and are large players among all the states/ cities
	Southeast: Tampa, FL - since the city has been most hit in the RYL's southeastern region
	Texas: Dallas and San Antanio - since their price trend has moved in line with Texas' state-wide trend
	West: Sacramento and Los Angeles, CA - since their price trend has moved in line with Californian state-wide trend; and Las Vegas, NV - since it has been hit badly in RYL's Western region
	Within sources, we have assigned weighs to the five indices to arrive at a weighted average predicted growth rate in average closing prices for each region.
	At present, we have assigned a higher weight to particular price indices since their trend is forward looking to arrive at homes' average closing prices. Further, as is noted in the Boom Bust Blog, the S&P index covers only single-family attached and detached housing prices and excludes condos, which are a part of RYL's housing portfolio

*Note -->*	The break-up of RYL revenues for single-family and condos is not available, hence, we have applied the price expectation for the company's overall housing units, although we have financially engineered condo price drops into our real estate pricing data.

Volume estimates	For new orders' estimates, we have used three key US homes data - - No. of building permits, State of the Cities Data Systems (SOCDS) (available state-wise) - No. of housing starts, US Census Bureau (available US Census' region-wise) - New home sales, US Census Bureau (available US Census' region-wise)
	For each region, we have identified key representative states/ cities, whose price indices are available, as follows:
	North: Midwest region of the US Census and MD, DE, OH, IL, IN, MN, WA states of SOCDS
	Southeast: South region of the US Census and FL, NC, SC, GA states of SOCDS
	Texas: South region of the US Census and TX state of SOCDS
	West: West region of the US Census and CA, NV, AZ, CO states of SOCDS
	Within sources, we have assigned weights to the three databases to arrive at a weighted average predicted growth rate in RYL's gross new orders for each region.
	At present, we have assigned a higher weight to the SOCDS building permits data since it tracks the trend state-wise and further, it reflects the expected new construction activity, in turn indicating the new orders received

Cancellation rate	We have assumed the high cancellation rate for RYL till 2009 owing to low consumer confidence and disrupting mortgage conditions. We have assumed some recovery in the cancellation trend beginning 2010
	Region-wise, we have estimated cancellation rates based on our overall regional outlook on the US housing industry.
	Consequently, we have assumed the highest cancellation percentages in RYL's Western region, followed by Southeast and some moderated increases in the company's North and Texas regions. This is primarily owing to the steep decline in RYL's western region's prices and new orders, low consumer confidence, and the expected decline in the region which witnessed significant growth during boom times

Cost of sales per closed unit	We assume that RYL's cost of sales to grow at lower growth rates q-o-q, albeit marginally and beginning in 2009. As the inventory keeps on building in housing industry, we expect a slight decline in the land prices till 2009-end post which we assume slight recovery in land prices. This is a very conservative assumption and it is quite likely that land prices will slide further into recession beyond 2009. Further, we assume that the y-o-y rate of growth of wages and material cost will decline in going forward

SG&A	We assume higher SG&A expenses as percentage of RYL revenues in the near term, reflecting the rising cost of inventories being held and interest expenses. However, in line with the company's objective to incur SG&A expenses at the rate of 10% of its revenues, we have assumed some efficiency 2010 onwards

Financial Services	In addition to the increasing interest liabilities on warehouse lines of credit increase, a growing disability to re-sell their mortgages in the secondary market is posing a challenge for most homebuilders in the US who offer mortgage financing to its buyers. This should have an adverse effect on RYL's financial services segment as well as on its gross new orders
	Consequently, we have lowered the rate of growth in the number of mortgage origination units for RYL till 2009, post which we assume recovery to start off
	Further, we have assumed that RYL's re-sold mortgages as percentage of total mortgage value will decline steadily before recovering by end-2008
	Financial Services revenues: We have projected the Net gains on sales of mortgages and mortgage servicing rights, Title/escrow/insurance, and Net origination fees on per unit basis and have assumed a decline in the revenues per unit till 1H2008 post which the revenues should recover

Housing price change assumptions as of Oct. 23, 2007
	3Q2007	4Q2007	1Q2008	2Q2008	3Q2008	4Q2008
West	-5.30%	-4.90%	-3.90%	-4.10%	-3.80%	-3.10%
North	-4.10%	-3.80%	-2.80%	-2.10%	-1.80%	-0.90%
Texas	-0.80%	-0.20%	0.20%	0.30%	0.70%	1.30%
Southeast	-6.00%	-6.20%	-3.90%	-2.60%	-1.40%	-1.20%
Average	-4.10%	-3.80%	-2.60%	-2.10%	-1.60%	-1.00%

	1Q2009	2Q2009	3Q2009	4Q2009	2010	2011	2012
West	-2.70%	-2.30%	-2.30%	-2.00%	-2.50%	-3.10%	-2.20%
North	-0.80%	-0.70%	-0.30%	0.20%	0.40%	3.70%	4.10%
Texas	2.70%	5.40%	10.80%	21.50%	30.30%	24.20%	19.40%
Southeast	-0.40%	-0.50%	-0.60%	2.00%	-2.70%	-8.50%	-9.30%
Average	-0.30%	0.50%	1.90%	5.40%	6.40%	4.10%	3.00%