Builders, Lenders: Next In The Falling Knife Saga

TIPS OF AN IMMENSE ICEBERG
James Grant, editor of Grant's Interest Rate Observer, explained the root of the problem in a recent interview. He believes the troubled part of the mortgage market is immense, far larger than most analysts are letting on, and it's a story with wider implications than just the borrowers themselves.

Hundreds of billions of mortgage-backed securities have been bundled in collateralized debt obligations (CDOs) over the last two years in what Grant described as "wild and woolly, wholly reckless" abandon in mortgage underwriting -- so much so that it "measurably heightens recession risk."

But only when one cuts to the heart of what he calls the micro issue -- the leviathan CDO derivatives market -- does the scope of the problem become clear. It also shows how the wall separating prime and subprime loans that many believed inviolable has been surreptitiously breached.

"Hundreds of billions of dollars of these %91science projects' of collateralized debt obligations and the like are sold in any year -- immense amounts," Grant recounted in a discussion with Bloomberg News in March. "What are they? You take a bunch of loans and then fashion an asset-backed security from those loans. Then take those loans and repackage them into a collateralized debt obligation, a CDO. There can be a million individual loans in a CDO.

"Now here is the magic and mystery of a CDO," Grant continued. "You can take 70% of a pile of triple-B minus [meaning marginal credit] and the rating agencies will bless it.

"They will say that %91we have studied the correlations and trends and we pronounce 70% of this pile of junk to be triple-A.' Triple-A!"

WHAT GOES UP...
How do they do it? "They have their models -- very elaborate and no doubt mathematically literate and internally logical models that tell them that in a given environment, these loans will perform up to triple-A standard," Grant continued. "My contention is that the last couple of years have been so anomalous from the point of view of underwriting practices [not to mention the biggest bubble ever in US real estate prices, according to Robert Shiller -- see Figure 1] that all those models are going to be wrong."

FIGURE 1: Chart showing quarterly changes to home prices from 1987 through the fourth quarter of 2006 provided by Shiller and Case, arguably the best overall indicator of housing prices across the nation, and a textbook example of a parabolic blowoff from 2001 to 2005. National home prices doubled between 1999 and 2006. Data: Chicago Mercantile Exchange.

Grant estimates that 40% of the mortgages issued last year were either subprime or "Alt-A" (one step above subprime) mortgages and therefore risky. He has a point. Subprime loans jumped from $100 billion in 2000 to $800 billion in 2006, thanks in large part to the willingness of investors to buy mortgage-backed CDOs, especially after they'd been blessed as predominantly triple-A (or prime) by rating agencies. It meant that there was lots of cheap money around and real estate prices soared. But the rocket ride couldn't last.

According to a March 2007 report by the Center of Responsible Lending, 91% of subprime loans made nationwide since 1998 were used to refinance existing mortgages affecting 2.4 million Americans. Approximately 50% of those funds were used to finance consumer purchases and have hence driven overall economic growth. The Center estimates that one million of these borrowers could lose their homes -- that's roughly one out of every seven homes sold each year in the US.

According to the Mortgage Bankers Association, lenders started foreclosure proceedings against more than one in 200 homeowners in 2006. Florida, the nation's real estate hotspot until recently, now leads the nation in foreclosures, with California running a close second. Doesn't sound like the kind of problem that has a quick and painless resolution, but what does it mean for stocks?

CHART-CHALLENGED CHEERLEADERS
In mid-November, the rally of the VectorVest Builders (residential/commercial) Index, a composite of 24 companies that build homes across the US, was still in a rally that had begun in mid-July 2006 (see Figure 2, blue trendline, far right). It was this chart pattern that caused a number of analysts to agree with perennial industry market cheerleader David Lereah of the National Association of Realtors that the worst was over (for about the sixth time since the housing melt had begun). But as we can see from Figure 2, while the index was heading higher, the underlying fundamentals and technicals looked anything but strong.

FIGURE 2: An updated chart showing changes in the group of 24 builders that make up the VectorVest Builders (residential/commercial) Index. Arrows 1 and 2 signify trendline breaks in early 2006 and earnings growth or GRT (red line), earnings per share (black line), and growth to PE (teal line) have since deteriorated even as the index rallied from July 2006 through February 2007, until the latest rally trendline was breached March 5 (third red arrow). Chart by www.VectorVest.com

As we see from one of the strongest builders in the Builders Index, Lennar Homes (LEN) in Figure 3, the chart looks anything but bullish. In the beginning of April 2007, the stock broke down through the head & shoulders top neckline (see red arrow, far right) in a pattern that had begun forming in mid-2003. But only after the technicals had broken down did earnings growth turn negative (mid-November 2006).

FIGURE 3: Lennar Homes, one of the strongest builders in the group, was giving some very bearish technical and fundamental signals in mid-March. Note the head & shoulders top pattern looks just about to be confirmed, with price sitting on the neckline (March 30), and if (when) confirmed by a neckline breach, puts a target price on the stock, an equal distance below the neckline as the head rose above it would be $67.81 (head) - $42 (neckline) = $25.81, making the target $42 - $25.81 = $16.20. This bearish scenario is further supported by the breakdown in red earnings growth (GRT) line to -7%, as well as a drop in relative value (not shown) from 1.52 at the head in July 2005 to just 0.74 by the end of March 2007. Relative value broke below 1 in October 2006 for the first time since April 2000.

Earnings growth rate had declined to zero by the end of November 2006 and has since declined to -2%. Growth to PE (GPE) (teal line) turned negative in mid-December and has dropped farther into negative territory since. After peaking at $10.89, earnings steadily dropped, hitting a low of $3.73 per share in early January before rebounding to $4.68 in early February, but they started dropping again. After peaking in June 2006, relative value (RV) began a slow but steady decline, turning negative in October, and as of mid-March sat at 0.89.

Industry fundamentals continued to decline and the bear rally trendline for the Builders Index in Figure 2 was breached on March 5, 2007 (red arrow, far right). Of the 26 builders that make up the index, zero were rated a buy, 22 were rated sell, and four were rated hold by the end of March. This was a big drop from just a month before.

FISHING IN A PERFECT STORM...
Following its peak in the summer of 2005, builders began to decline, but many in the industry were hoping that the latest rally was the reversal they'd been hoping for. But what impact would the emerging problems in the mortgage sector have? Industry spin doctors and bulls were again working overtime to get out the message that the worst was over. Did their argument have merit? Those who knew how to read both charts and fundamentals wouldn't be fooled.

Not only do both indicate that the downtrend is closer to the beginning than the end, economic fundamentals further supported this view. As evidenced by the latest GDP and productivity figures, economic growth was contracting and in a slowing economy, jobs are the next casualty. How could this help spur home sales?

While it was not good news for bulls, how would the bears know when to pounce? Other than the VectorVest Builder Index and other charts discussed here, the bears can track the Philadelphia Housing Sector Index (HGX) to follow homebuilders. For lenders, the Dow Jones US Mortgage Finance Index (DJUSMF) provides a good heads-up.

CATCHING THE BIG ONE
In any bear market, there are significant rallies to confuse investors and traders. For this reason, it is always important to keep the big picture in mind. Between 2000 and 2005 the VectorVest Builders Index rose more than 830%, and star performers like Hovnanian Enterprises (HOV) increased more than 2,200%. This parabolic rise makes the NASDAQ bubble rise between 1995 and 2000 of 680% pale by comparison. It eventually gave back 85% as it dropped back to 1996 levels. By March 30, 2007, the Builders Index dropped 37% from its July 2005 peak and HOV was off 65%. While that may seem like a serious correction, the Builders Index was still up more than 500% and HOV still up more than 680% above 2000 levels. Given continued economic and market headwinds, they have a lot farther to fall.

FIGURE 4: A picture is worth a thousand words. Here we see Dominion Homes (DHOM), one of the weakest builders in the VectorVest Building (Res/Com) Index, that collapsed 88% from a high of $37.60 in April 2004 to $4.47 by March 30, 2007. Earnings growth dropped into the basement in late 2004 and was sitting at -9% in March 2007.

Contrast this with an election nearing and politicians and government working overtime to keep the economy humming. The preelection year (as in 2007) has traditionally been the strongest year of the four-year election cycle. This means stocks could remain strong, which would be positive for lenders and builders. But while this may delay the inevitable correction, given the underlying fundamentals the longer these stocks rise, the more severe it will be.

When the rally is truly over, the bears can always short individual builders and lenders. But are there safer ways to play the downside?

FIGURE 5: The VectorVest Financial Mortgage Services Index consisting of 24 mortgage lenders from the largest Freddie Mac selling above $60 to Finova Group trading at about a dime. On March 14, 2007, the index fell to $13.56, the lowest price since October 1995. Earnings per share for the group fell from a high of $2.40 in November 2004 to $0.97 in March 20, 2007. Out of a total of 24, none were rated a buy, 10 were rated a hold, and 14 rated a sell by March 30, 2007.

The streetTracks Homebuilders (XHB), an exchange traded fund (ETF) launched in February 2006, is one way. While it also enjoyed the latest bear rally, as markets melted around the globe for a second time in a year on February 27, it also got hammered, dropping through trendline support at $36. It wasn't a good sign for the bulls.

Housing futures contracts traded on the Chicago Mercantile Exchange (CME) with contracts for each of 10 cities in the US, plus the CME Housing Composite Index, is another way to play the game. So far, these futures have attracted limited interest as evidenced by the low volume, but as the housing contagion continues to spread, they have the potential to get a lot more popular with builders, investors, and traders looking to hedge real estate (and mortgage) holdings.

FIGURE 6: Chart of one of the more recent lenders to get into trouble, Accredited Home Lenders (LEND) that had to fire-sale $2.7 billion in mortgages at a substantial discount to meet margin payments in early March. The stock, which dropped more than 90% from $60 in May 2006 to below $4.00 in March 2007, did what looks to be a dead-cat bounce in mid-March. Lax lending practices have been attacked by everyone from financial analysts to members of Congress.

While we have yet to find a subprime mortgage ETF that trades on US markets, we can trade some of the larger lenders like Countrywide Financial (CFC). Although I haven't analyzed them yet, real estate investment trusts (REITs) with a sprinkling of mortgages in their portfolio would be another option. There was even a New Century Financial REIT (NEWC), which was still trading in early April around $0.91. The Vanguard REIT ETF (VNQ) is another possibility, but from the looks of it, it must hold prime mortgages as it was in a strong uptrend as of the first week in April. Not a short candidate yet.

FIGURE 7: A mortgage industry bellwether, Countrywide Financial (CFC), a division of GMAC Finance, breached its long-term uptrend line on April 2, 2007, as fellow lender New Century Financial formally declared bankruptcy. Note that sales growth turned negative in early and earnings growth rate (GRT) (other red line in lower subgraph) has steadily declined since mid-2006. Expect more downdrafts for this and other lenders as the subprime storm widens. A perfect storm in the making?

Whatever issue you decide to trade, it's important to use all the tools at your disposal. Traders tend to focus on technicals but they are usually short term, and unless you are looking at longer-term weekly or monthly charts, you can miss the big picture. Fundamental investors, however, blunder into following corporate fundamentals without realizing that often, they only break down after stock prices have already fallen.

To get the best view of what is really going on, the most successful traders and investors that this author has had the benefit of knowing and trading with follow both.

Live links to related reading http://tradesystemguru.com/content/view/39/9/

- Mortgage crisis hits million-dollar homes (abbreviated link) http://tinyurl.com/27oq36
- Subprime shakeout could hurt CDOs (abbreviated link) http://tinyurl.com/2sjqer
- CME Housing Futures Overview http://www.cme.com/trading/prd/re/housing.html
- Fundamental definitions used in this article http://www.VectorVest.com/definitions
- Tables of the stocks in the Building and Finance Indexes http://www.VectorVest.com/builders-lenders
- Link to Bloomberg interview with Jim Grant (abbreviated link) http://tinyurl.com/2k5pbl