|Shares Out. (in M):||116||P/E||9.7||8.4|
|Market Cap (in $M):||3,926||P/FCF||0||0|
|Net Debt (in $M):||0||EBIT||0||0|
I am attracted to the homebuilders, and particularly to CalAtlantic. Our thesis is that a shortage of developed building lots will lead to a shortage of new homes, to higher prices for new homes, and consequently to a particularly sharp increase in the profits and cash flows of the large homebuilders.
A common scenario on Wall Street is that (1) housing completions (single-family plus multi-family) will increase at a 9-12% CAGR from about 1,075,000 units this year to a more level of 1,500,000 over the next 3-4 years and (2) the large public homebuilders will grow somewhat faster than 9-12% per year as they gain some market share from small public builders and from private builders. If this scenario proves correct, I believe that the shares of the homebuilders will appreciate at a good rate over the next several years.
However, I am more excited about a second scenario, which I believe is likely to occur. Some background is useful. In the past, most homebuilders continually re-invested the bulk of their cash flows into purchasing additional land that they would build on in the future. Thus, the homebuilders generated relatively little cash to return to their shareholders – and, for this reason, the homebuilders have sold at low PE ratios. One homebuilder has been the exception to this strategy. NVR typically purchases most of its lots close to the time when it builds homes on the lots. Because of this “asset light” strategy, NVR has generated large free cash flows, which the company has heavily used to repurchase shares. Between 2013 and 2015, NVR’s diluted share count declined at a 6.5% CAGR. Apparently, as a result of its asset light strategy and its aggressive repurchases of its shares, during the 2013-2015 period, NVR’s shares traded at an average PE ratio of 16.8 X.
Seemingly, other homebuilders have taken notice of NVR’s success and are adopting asset lighter strategies. I say “lighter” rather than “light” because NVR builds houses in parts of the U.S. where it is able to purchase land that already has been developed by independent land developers. Most of the other large homebuilders operate in parts of the country where it is more difficult to purchase developed lots or to purchase land via options. None-the-less, by emphasizing free cash flows over growth, by slowing the purchases of land, and by accepting lower gross margins on optioned land, the large homebuilders seem to be adopting the strategy of reducing their relative investments in land and of returning more cash to shareholders. For example:
* On its September 20th Q3 FY 2016 earnings call, Lennar reiterated that the company has adopted a “soft pivot land strategy over the past years, away from land-heavy acquisition”. Management added that it is “now targeting high-quality A-location land acquisition with a shorter two to three-year average life” and that it has “moderated” its growth target to the 7% to 10% range. Between the end of Q3 2015 and the end of Q3 2016, the number of lots under Lennar’s control declined by 6.3% from 169,801 to 159,122.
* D.R. Horton’s management told me that the company expects to hold its investment in land relatively flat for the next few years, favoring options on land as opposed outright purchases. For its fiscal year that ended on September 30, Horton reduced did hold its total investment in land flat – and it reduced its owned lots by 4% while increasing the number of lots under option from 56,000 to 92,000.
* On its July call for Q2 earnings, Pulte stated that “after four years during which our investment in new land grew significantly, we plan to slow the rate of growth in our new investment in 2017 and beyond”. And, under pressure from an activist investor (Elliot), management has decide to repurchase $500 million of its shares in the second half of 2016 and an additional $1 billion in 2017. Clearly, these large repurchases will restrict the company’s ability to make large purchases of land.
* Then, on its Q3 call, Pulte’s CEO stated: “based on our trailing 12 months of deliveries, we own approximately 5.2 years of lots, which is down from 7.7 years in 2011. Over time, I would like to see this number continue to move lower and ultimately get below 4 years, even approaching 3 years of ownership.” After increasing sharply in recent years, the number of lots under Pulte’s control (owned plus optioned) actually declined in Q3: from 152,253 to 145,252.
* And, on CalAtlantic’s Q3 earnings call, Larry Nicholson (CEO) said that his company would like to reduce its land positon to a four year supply. Over the past year, the number of lots under CalAtlantic’s control has declined by 7.3% from 75,515 to 69, 964.
On balance, the above four large homebuilders only increased the number of lots under their control by an average of only 1.6 % or the past two years at a time when the number of homes they sold increased at close to a 14% CAGR. Thus, their inventories of land, when measured in years of supply, declined materially over the past two years – and this decline is continuing.
At the same time that most of the larger homebuilders are purposely adopting asset lighter land strategies, many of the medium and smaller public homebuilders have leveraged balance sheets that restrict their ability to purchase land – and many of the small private homebuilders are finding it difficult to obtain the capital to purchase and develop land.
Our strong conclusion is that there is a high probability that the number of developed lots will be insufficient to meet the demand for new single-family houses, with the result that there will be shortages of new homes on the market. The shortages, in turn, likely will lead to pricing power and therefore to higher margins and earnings than presently projected by Wall Street. Furthermore, the higher earnings, combined with the restricted land purchases, will lead to large free cash flows that can be used to further strengthen the companies’ balance sheets, to pay materially larger dividends, and to repurchase shares.
Already, there are shortages of lots and houses in some parts of the country. The following is a quote from the minutes of the September 20, 2016 joint meeting of the Federal Open Market Committee and the Governors of the Federal Reserve Bank: “The sluggishness in the housing sector appeared to have continued into the third quarter. A couple of participants pointed to limited availability of lots and a shortage of skilled labor as restraining residential construction activity in their Districts.”
While we believe that all of the large homebuilders’ shares are attractive, CalAtlantic currently is our favorite. CalAtlantic is the 2015 merger of equals of Standard Pacific Corp and the Ryland Group. MatlinPatterson Global Advisors (a private equity firm) had purchased control of Standard Pacific in 2008 when the company needed an infusion of cash. Currently, MatlinPatterson owns about 36% of CalAtlantic’s outstanding shares. Our logic is that the MatlinPatterson fund holding the shares will liquidate during the next several years, incentivizing MatlinPatterson to achieve a high stock price for CalAtlantic before then. Thus, CalAtlantic’s management has pressures to achieve intermediate term performance.
CalAtlantic’s management told me that they are adopting a relatively asset light model and that they are interested in repurchasing shares. In the first nine months of 2016, the company repurchased about 4.3 million shares, or about 3% of the number of diluted shares outstanding on January 1. I would not be surprised if the company steps up its repurchases.
We estimate that CalAtlantic will earn about $3.50 per share in 2016, $4.00 per share in 2017, and $5.00-5.50 in 2018. Reasons for the large EPS increases include: (1) mid single-digit annual increases in the company’s community count, (2) general increases in the demand for houses, (3) price increases modestly in excess of cost increases, (4) leverage over fixed costs, (5) synergies stemming from the 2015 merger, and (6) share repurchases.
On September 30, CalAtlantic’s tangible book value per share was $26 (assumes conversion of the 1.635% debentures convertible at $31.40 and the 1.25% debentures convertible at $40.22). Based on the above earnings estimates and allowing for some slight dilution from the repurchase of shares, we estimate that CalAtlantic’s tangible book value will increase to around $33 in mid-2018 (the company currently does not pay a dividend). To the extent that a company’s book value mitigates risk, CalAtlantic’s risk-to-reward ratio seems compelling.
I note that some investors are concerned that increasing interest rates will dampen the demand for new houses. However, people need to live somewhere – and, of importance, during the 20-year period 1980-1999, thirty-year mortgage rates averaged about 10.4% and yet housing completions averaged 1.43 million per year a time when the population of the U.S. averaged about 250 million. Adjusted to the current U.S. population of 320 million, the 1.43 million average annual completions during the 1980-1999 period are equivalent to 1.83 million today. Thus, over the intermediate term, we are not concerned that increasing interest rates will prevent housing starts from returning to a normal level of 1.5 million.
A shortage of developed building lots will lead to a shortage of new homes, to higher prices for new homes, and consequently to a particularly sharp increase in the profits and cash flows of the large homebuilders.
|Entry||11/23/2016 11:24 AM|
Thanks for the interesting write-up. We have been spending some time on the homebuilding space recently and were attracted to TOL and CAA initially because of near book valuations and the share repurchase programs. I see that you had previously written up TOL. Any updated thoughts on that name would be much appreciated. Current results seem to be very good but I worry that we are in the latter innings of the high end recovery. Exposure to some of the urban markets like NYC is also a concern but may be more than priced in here. Thank you in advance.
|Subject||Risk to the option model of other builders|
|Entry||11/23/2016 01:08 PM|
I think you are overstating the ease of executing an options model like NVR. The homebuilders broadly have for years said they admire the asset-lite option model of NVR and want to move in that direction, yet it remains a very small part of their portfolio. It only works if you have a dominant market position in a given territory and therefore have scale/lower costs and can pay more for the land (after all, you will be paying more if you option it). If you have no cost advantage then you won't be competitive. If you look at NVR's G&A, it is 50% lower as a % of sales than the other builders.
Not saying you are wrong in your assessment they will be banking less land and perhaps that leads to lot shortages but I don't see them meaningfully moving toward options today anymore than they have in the past.