If interest rates keep rising, no time soon.
The last 8 weeks have been tough on the homebuilders. Just when it seemed things were hitting on all cylinders and the cycle was ramping, storm clouds arose on the horizon.
It appears the bond market is anticipating a once-in-a-lifetime upswing in the Interest Rate Cycle. Over $80BL was withdrawn from bond funds in June, 2013, the largest single-month withdrawal in history. Investors fear that the recent upswing could be the harbinger of a permanent rise in borrowing costs.
It is interesting to note that last year's July 23-25, 2012 low in the 10-year Treasury bond matched its previous lows of 1945. At 1.4%, it was only the second time since 1790 that there had been a print that low.
And until May 3rd of this year (2013), it remained at 1.63%. But the 10-year yield has now nearly doubled (96%) from its all-time lows, and the last phase of this rise created a 30% drop in the homebuilder shares. It's starting to look like panic-selling.
Is the selling warranted?
Most fundamental analysts say No. There have been numerous well-written articles exploring the bull case for homebuilders. John Tobey, Duru Ahanotu, and the nom de plume Money Investor (Seeking Alpha authors) have done an excellent job presenting the builder's case with charts, solid research, and a good knowledge of real estate.
There is also anecdotal evidence from Credit Suisse that the real estate market is red-hot, especially in DC, California, the Northwest, Texas, and parts of Florida and Arizona. Agents say that the recent rise in rates has had no effect on their real estate business, and that mortgage rates remain exceedingly low by every historic measure. Median-home prices are rising and supply is falling. So all systems look like a go.
Then what's the problem? Again, it's the rising interest rates. They are an inseparable part of the homebuilder equation in a way that is unique to this business. You can never take your eye off them. I recently wrote an article Rising Rates Squeeze Home Builders And Buyers (July 11, 2013), because I had an increasingly uncomfortable feeling about my homebuilder bullishness but could not fathom why. If the fundamentals were only improving, why was I so bummed?
Then I came across this small article on the housing-tracker website, Growth in Mortgage Purchasing Power (April 3, 2013) and the pieces began to fall into place.
"Growth in the capacity to borrow has replaced income growth over the last 30 years. Of course this is possible because interest rates have been falling continuously since the early 1980s making it feasible to borrow more and more with less income.
If we wanted to increase purchasing power in an environment where interest rates were not in decline, we'd need to see a substantial increase in real median income. For instance, say 30-year mortgage rates were at 6.5% instead of their recent level of roughly 3.5%. All else being equal, to get the same purchasing power as a 3.5% mortgage rate with a mortgage rate of 6.5% would require a 41% increase in real income! Clearly (and by design in recent years), record low mortgage rates are a huge stimulus for home prices.
What happens when the 30+ year secular decline in interest rates ends? Even if rates stay low, the stimulus of declining rates on asset prices (homes in particular) will disappear. While incomes will likely increase over the next few years, we've already seen how much they would need to increase to match the purchasing power effects of falling interest rates. And if interest rates rise even modestly, purchasing power will be significantly curtailed."
In early 2011, around the time of my original purchase of Pulte Homes (NYSE:PHM) at $6.75, I studied Pulte's price appreciation to see what happened as it emerged from the previous housing crisis in 1990.
The years 1989-1992 were the last big meltdown in our residential housing cycle, and because Pulte is such a nationally-diversified builder, it is a good window on the U.S. homebuilding industry. As Pulte goes, so goes the homebuilders. And because the company has been around for so long, there are ample charts to plot its course through the years.
I begin my sequence with its initial meteoric rise in late 1990.
- From October, 1990 to October, 1993 the TNX (10-year Treasury bond yield) dropped 360 BP (basis points) from 8.8% to 5.2%. In the same time frame, Pulte shares rose from 65 cents a share to $4.35.
- In October, 1993, the TNX turned on a dime (sound familiar?) and rose 70 BP in 4 weeks to 5.85%. In response, Pulte dropped 75 cents (-17%).
- The TNX stayed in a sideways holding pattern for the next 3 months (through Feb.1, 1994), when Pulte again shot up, this time to $4.70 on February 1, 1994, a full 620% rise from its October 1990 lows. (Of note, these percentages are almost identical to Pulte's recent 600% rise from October 2011 to May, 2013.)
- From October, 1993 to October, 1994 the TNX increased 280bp from 5.2 to 8.0 (54%) while Pulte dropped (-52%) from $4.70 to 2.28.
- During 1995 and through 1996, the TNX dropped 250 BP from 8% to 5.5% (-31%) and Pulte rallied 80% from 2.28 to $4.18.
- In 1996, the TNX rose from 5.5 to 7.2 (31%) and Pulte dropped 31% from $4.18 to $2.90.
- In late 1996 the TNX dropped from 6.95% to 6.05% (-13%), and Pulte rallied 28% from $2.90 to $3.80.
- In April, 1997 the TNX dropped from 6.95% to 5.4% (-22%) and Pulte rose from $3.40 to $8.50 (150%). It was at this point that the fundamentals of the housing recovery kicked-in and home values and sales increased rapidly.
- Beginning with the President Clinton-inspired stock-market crash in October, 1998 the TNX rose 260 BP from 4.2% to 6.8% (61%) through March, 2000; and Pulte rallied off its low of $4.80 in the October, 1998 crash to $7.52 by January 1, 1999. But then it dropped to $3.75 (-50%) amidst the relentless rise in interest rates at that time. And this was despite the rapidly improving fundamentals of the housing market.
- From March, 2000 to March, 2001 the TNX dropped 205 BP from 6.8% to 4.75% (-30%), and Pulte rallied from $3.75 to $12 (220%) on the back of the accelerating housing recovery.
- During 2001 and into April of 2002, interest rates fluctuated wildly, from 5.5% to 4.25% (down 125 BP) and then back up again to 5.4%. Pulte responded by rising from $8 to $12 (+50%), then dropping to $9.50, only to retest the highs again at $11.75, before falling 45% to $6.50 in October and closing out the year at $11 (+70%).
- Interestingly, 2002 is only the second time that Pulte's shares actually rose alongside the TNX yield. Valuations had become so constricted during the bear markets of 1998 and 2001 that the stock rallied in an act of price-reversion.
- From April, 2002 to June, 2003, the TNX fell 240 BP from 5.5% to 3.1% and Pulte soared to $17.
- From June, 2003 to July, 2003, the TNX suddenly rose 150 BP from 3.1% to 4.6% (+50%) inside of a single month, and Pulte fell from $17 to 14 just as quickly.
- But when interest rates went sideways - and then down 90 BP - from 4.6% to 3.7% in March, 2004, Pulte soared to $28 (100%).
- It was in mid-2004 that Pulte's share price began to disconnect from the TNX for the first time. For example, in October, 2004 the TNX had fallen 90 BP from 4.9% to 4% and yet Pulte fell too, from $31 to $23.
- In the steady-rate environment that existed during 2004-2005 - when the TNX fluctuated between 4.6% and 3.9% - Pulte doubled from $23 to $46 by July, 2005, and that marked the height of the housing mania.
- From July, 2005 to July, 2006 the TNX rose 135 BP from 3.9% to 5.25% and Pulte dropped 44% from $46 to $26.
- But when the TNX fell 80 BP from 5.3% to 4.5%, Pulte rallied back up from $26 to $34 (+30%). When the TNX rose from 4.5% to 5.3% in July, 2007, Pulte fell from $34 to $25; and when the TNX dropped 200 BP into October, 2008, the fall in interest rates did nothing for Pulte's stock, which had already fallen to $7.50, a 70% sell-off from the 2007 highs.
- During 2009 and through July, 2010 there were wild swings in Pulte's price from $12 to $8 to $12 to $8 to $13 to $9 to $13 to $8 (a total of $30 in price swings); and all that volatility in just an 18 month time-span.
- On the last swing down, the TNX fell 160 BP from 4.0% to 2.4% and Pulte fell right alongside it.
- Finally, by October, 2011, Pulte had fallen another 58% from $8.50 to $3.50 while the TNX had also fallen 195 BP (53%) from 3.7% to 1.75%. This was during the commencement of Quantitative Easing, as the Federal Reserve experimented with what (if anything) could stimulate the then-moribund housing market.
- From October, 2011 to May, 2013, the TNX remained in roughly an 80 BP trading-band between 1.6% and 2.4%. During that same time period, Pulte rose 600% to $24.50.
- Finally, in May of this year, the TNX rose 80BP (from 1.6% to 2.4%), and Pulte quickly fell 25% to $18.
The Recent Rate-rise in Microcosm
The TNX began its rise on Friday, May 3, following a positive employment report. At first, the employment news seemed very good for the builders. More employment meant more new home buyers at record-low interest rates. So Pulte rallied 20% on that news, up to its multi-year May 15th high of $24.50. But interest rates also crept up 32 BP from 1.63 to 1.95% on this news.
To backdate this drama a little bit, let's go back to the last leg down of the TNX from 2.1% on March 11th to 1.63% on May 3rd, a 50 BP (-24%) drop in interest rates.
Interestingly enough, Pulte's shares did not rise as the TNX fell, but rather, went down too - as much as $4.20 (19%) - until they landed at $17.50 on April 22nd. From there, Pulte quickly rallied $7 (40%) to its multi-year high.
During that brief window of time from April 22nd to May 8th, officers and execs across the homebuilding sector began selling their shares aggressively. Five Pulte insiders sold stock between $21 and $22. As a matter of fact, every time the stock had risen above $19 during the Spring of 2013, insiders sold boatloads of stock, the largest sale being the CEO's 500,000 shares on March 5, 2013. His options had vested on that day, and he sold $10ML of stock in a single morning.
Pulte insiders were not alone in their Spring selling. The CEO and CFO of Meritage (NYSE:MTH) sold shares in the same time frames and to the same degree. The Meritage CEO sold 100,000 shares just a whisker below Meritage's 7-year highs. Similarly with Toll (NYSE:TOL) brothers' officers. On May 1-3, the CEO of Lennar (NYSE:LEN) sold a 100,000 shares right at the top.
In recent conference calls, these CEOs have stated that rising interest rates have not affected their businesses on the ground YET (and that's obviously true from the earnings reports we're reading).
But these guys were selling their shares en masse this Springtime; and especially during that small window of time between late April and early May when rates began to seriously rise from decade lows for the first time. Were these execs just lucky or prescient? Or did they have their eyes on a chart that looked something like this, or the TNX's long-term chart?
Like it or not, those CEOs, CFOs and officers sold at the top. They sold boatloads of shares for millions of dollars this spring, coincidental with the last hurrah in homebuilder price rises on May 15, 2013.
On May 22nd, the TNX began an additional 27 BP move from 1.88% to 2.15% on May 28th. Pulte's stock had not yet been affected by the rise. It opened that day at 23.25. At this point, the TNX had risen 50 BP off its lows (nearly half its eventual move), and yet Pulte seemed immune to it.
But beginning on May 28th through June 10th, Pulte dropped $3.50; then it rose $1.50 through June 19th, the day preceding Mr. Bernanke's re-explanation of his tapering comments.
From June 17th through the 24th, the TNX rose another 50 BP to 2.65%. In those 5 days, Pulte fell $4 to $17.75. The interest-rate rise was now beginning to bite. And in just the last two trading days, Pulte has fallen 15%.
From the looks of things, it appears that investors are getting the message that interest rates might rise again in anticipation of next Friday's employment report, and panic-selling has set in.
We have quickly reached a point where all news - really, any kind of news - seems bad for the builders, and they are out there like sitting ducks. If employment rises, interest rates will rise, and the builders could get clobbered. If housing-starts or building permits soften, they will get hit again.
It's no secret that the builders are metering-out their best lands and developed lots in order to let demand - and demand only (not speculation) - determine their rate of construction.
Remember, these companies have just survived the largest housing debacle in 80 years, and they are not about to let the lessons learned from that experience get away from them. In a room once-crowded with thousands of small builders in 2006, there are only a handful of public builders left standing today.
So the builders are in no hurry to be in a hurry. Supply keeps falling; demand far-outstrips supply, and they like it like that. The metric that critics keep ignoring is the bottom-line. It's doing just fine because ASPs are rising and the builders are making more profit per home on better margins, even though starts and permits are a little softer. The "slowdown" (if there is one), has been engineered by both the builders and the interest rate rise.
But the real question to ask is "When does this P/E compression and book-value compression stop?
Based on the historical evidence of the last real estate cycle, the compression stops only when interest rates stop rising. I find NO example to support the wishful thinking that rising rates support the share prices of homebuilder stocks. Historically it's been exactly the opposite, and this time around seems no exception. Interest rates are intimately connected to the economy and employment expectations, and thus the builders are held captive to these two things. Employment brings buyers to the builders - but interest rates raise the cost of their homes - so it is an economic push-me pull-you.
Based on the twenty-year price action I have outlined, I don't think we'll be seeing Pulte at $24 or even $19 any time soon. Nor any of the other builders higher than 20% BELOW their May, 2013 highs. The timing and degree of insider-selling this Spring was a tip-off that the bloom is likely off the rose.
And if interest rates on the TNX rise another 100BP to 3.6%, the builders could lose an additional 25% to 30% capitalization from here. That would put Pulte at $13 or less.
The 20-year history of the homebuilder sector began with a bang off the bottom (600%+), which was followed by 4 subsequent years of wildly fluctuating rates and stock prices. I think the steady-eddy rise in the homebuilding sector of 2011-12 has quickly morphed from an investment into a volatile trading position, just like it did in October 1993.
Going forward from here, the rock-solid business fundamentals of the builders will likely take a back seat to macro-economic issues centered on interest rates. This will dismay some investors, but encourage others to trade the waves of action and reaction that are surely on the horizon.
As I write, the SPX 500 is as far above its 300 day moving average as it has ever been; and we are coming into one of the weakest times of the trading year - August and September. Be careful out there.
Disclosure: I have no positions in any stocks mentioned, but may initiate a short position in DHI, PHM over the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.