Homebuilding stocks have been on a wild ride in 2020. They boomed from the start of the year, rising 19.1% (according to my index of 11 publicly-traded builders) to February 21, handily beating the 3.0% gain on the S&P 500 and the flattish 0.6% return on the Russell 2000. Then in just four weeks, with the onset of economy-busting measures taken to combat COVID-19, the sector plunged 60.6%, much worse than 30.9% drop in the S&P and 39.6% drop in the Russell. Since the March 21 lows, however, the homebuilders have come roaring back, surging 117.4% to June 5, compared with the gains of 38.6% in the S&P and 48.7% in the Russell.
The extreme volatility in homebuilding stocks reflects both the fears of the negative impact of the COVID-19 lockdown of the economy and the hope that the housing market will return to normal in relatively short order. Since the onset of the lockdowns in mid- to late-March, the homebuilders have reported 20Q1 results and/or their managements have participated in broker industry conferences where they have described how their business has changed. In these communications, a surprisingly consistent narrative has emerged.
For nearly all of these builders, sales activity began to fall off sharply beginning in mid-March. In most states, builders were forced to close their model home centers, sales offices and design centers or at least see potential customers by appointment only. As a result, traffic to their communities fell nearly to zero. All of the builders were able to switch quickly to virtual meeting formats to maintain communications with potential buyers and customers with signed contracts who were at various stages of the closing process. Traffic to their websites surged, as house- or apartment-bound potential customers spent their now ample free time looking at their housing options. At the same time, cancellations of existing contracts soared, as buyers feared that their jobs might be at risk or that house prices would collapse. Most builders reported that sales from mid-March to mid-April net of cancellations fell 50% or more compared with prior year levels.
Since mid-April, however, net sales orders have improved significantly. Cancellations are still higher than historical averages for most builders, but buyers have come back into the market, mostly online, to sign sales contracts. Most builders (and other real estate sources) say that there has been a surge in interest from apartment renters who are fearful of COVID-19. Others say that with the approach of the phased lifting of lockdown restrictions, potential buyers are becoming more confident in their outlook and less fearful that they might lose their job or have to take a cut in pay.
All of the builders say that the sales rebound that began in mid-April has continued through the month of May. Although contract signings were down 50% or more during the early stages of the lockdown, they have come back so strongly that many builders are running at paces even with or even above the year ago period. According to figures compiled by John Burns Real Estate Consulting LLC, new home sales rose 21% in May from a year earlier.
At first glance, it is surprising that sales would bounce back so strongly even while the nation’s unemployment remains near historically high levels. Steve Hilton, the CEO of Meritage Homes, recently said that of the 36 million people now unemployed (as of mid-March), most had incomes below $40,000 per year. These are typically not buyers of Meritage’s homes (and probably not the buyers of most other builders’ homes). He estimates that only about four million of the total 36 million unemployed had incomes sufficiently high to qualify as potential Meritage buyers.
Even so, the builders will be watching employment trends closely going forward. Of primary concern is the risk that ongoing weakness in the economy could lead to more layoffs of higher paid, white (and blue) collar workers who are prime homebuying candidates.
Despite the quick rebound in sales, builders are still cautious about sales trends going into the fall and into 2021. While most builders believe that the long-term outlook for their businesses is still solid, it remains to be seen whether the recent spike in sales is just temporary and that the recovery may be more gradual.
To be sure, the quick rebound in sales may be due to temporary factors, including a catch-up from people who had planned to buy in March or April, those who previously had cancelled contracts but are now more confident in the outlook, those wanting a quick move out of an apartment complex to escape COVID-19 and those jumping back in to take advantage of the recent decline in mortgage rates. As this noise abates, it is possible that the base level of demand could be lower than what builders have been seeing recently and so the recovery could face a slower “U” or “W” track, rather than a “V”.
Several factors support this more cautious outlook:
First, there was a divergence between single-family housing starts and new home sales in April that is noteworthy. As noted above, builders said that sales began to rebound in mid-April. The Commerce Dept. reported that April new home sales were up 0.6% to 623,000 units (on a seasonally-adjusted annual basis); while single-family housing starts declined 25.4% to 650,000 units (also on a seasonally-adjusted annual basis).
Although the data is still subject to revision, the decline in single-family starts has been much more severe than the decline in new home sales. If this holds up in the final tally, it is surprising that builders would hit the brakes so hard on starts even though sales held relatively steady. This could be a signal that the builders do not yet have great confidence in the sustainability of sales and so are taking a more cautious view in their direct market exposure.
This caution is also evident in the NAHB/Wells Fargo Housing Market Index, a measure of homebuilder sentiment. The HMI rose slightly to 37 in May after its historic plunge to 30 in April. It remains far below its recent historical range of 60-70. (An HMI reading of 50 is neutral with equal numbers of builders as good and bad.)
All three components of the HMI rose in May, but remained far below 5-year averages. The HMI’s reading on current sales conditions was 42 vs. the five-year average of 70. Expectations for sales six months out were rated 46 vs. the average of 70; and current traffic was clocked at 21 vs. the average of 47. Despite the apparently stronger readings from the big builders, sentiment across the entire industry remains negative.
The notion of only modest improvement can also be seen in other indicators of the housing market and the broader economy. Fannie Mae’s Home Purchase Sentiment Index, a demand-side measure of sentiment among potential homebuyers, increased slightly to 67.5 in May from a near all-time low of 62 in April. Year-over-year, the Index was down 24.5 points. Buyers have a slightly more optimistic view of buying conditions, but remain concerned about the prospects for subdued income growth and job losses.
Similarly, the Conference Board’s Consumer Confidence Index nudged up to 86.6 in May from 85.7 in April, after hovering around 130 for the past two years. Consumers were slightly more pessimistic about present conditions, but also slightly more optimistic about future conditions. They are more optimistic that business conditions and the labor market will improve in the next six months, but also see little chance that their incomes will improve over that period.
Along with consumer confidence, employment is a key determinant of home buying demand. Last Friday, the Bureau of Labor Statistics reported that the economy added 2.5 million jobs in May, according to its establishments survey and 3.8 million jobs, according to its household survey. The consensus estimate had anticipated a decline of 7.0 million jobs, according to Barron’s (via FactSet), an estimate supported by continuing elevated (though declining) initial jobless claims of 10.9 million in May, including 1.6 million for the last week of May. With the surprise net increase in May jobs, the unemployment rate fell to 13.3% from 14.7% in April.
Included in the May job report was a description of the impact of COVID-19 on BLS’s data collection practices. The Bureau reported that despite instructions to its household survey interviewers to classify people who were absent from work due to coronavirus-related business closures as unemployed on temporary layoff, many of these people were classified as employed. Consequently, the number of employed persons, according to the household survey, was overstated.
The Bureau estimates that if the number of people classified as absent from work for other reasons were reclassified as unemployed, the unemployment rate would have been about three percentage points higher. The April jobs report contained the same qualification and indicated that a similar reclassification would have raised the official unemployment rate by five percentage points.
The BLS and the Census Bureau are investigating why this misclassification error continues to occur. However, in order to maintain consistency with its data collection policies, there will be no revisions to previously reported data. Thus, it is possible that the unemployment rate could go higher or be slow to improve in the months ahead. This episode highlights problems in data collection that may be affecting other economic metrics. Consequently, it is difficult to draw firm conclusions about economic trends from current data reports.
The recent rebound in net new contract signings, if sustained, should cushion the decline in revenues and earnings in the 2020 second quarter and beyond. Year-over-year comparisons will likely be unfavorable for the balance of the year, but less so with each passing quarter.
For the 2020 first quarter, the 11 builders that I follow reported an average increase in net orders of 12.0%, down from 26.7% in the 2019 fourth quarter. The first quarter figures included the plunge from mid-March associated with COVID-19. Up until that point, average net order increases were running significantly higher, perhaps around 20%.
Despite the rebound since mid-April, many builders are likely to report net orders for the second quarter that will be flat with the prior year, give or take maybe five percentage points. Orders for the third and fourth quarters are tougher to predict, as they depend upon where the base level of new home sales activity settles after the significant noise generated from the volatility in March and April.
Investors will also be paying attention to quarterly home deliveries (i.e. actual closings). Most builders say that they have slowed their construction rates to meet social distancing guidelines for their subcontractors. As a result, construction cycle times will be extended. Even if sales rebound in 20Q2 and beyond, it is likely therefore that most of the increase will be reflected in higher backlogs and it will take time to deliver these backlogs, until at least the social distancing guidelines are eased or eliminated.
With the longer construction cycles, deliveries for the second quarter will likely be up low single-digits at best and could be flat or down for many builders. In the two immediately preceding quarters, deliveries had been up in the low teens. The slower growth in deliveries will also be mirrored in a slower rate of revenue growth.
At the same time, operating costs as a percentage of revenues will likely be higher in 20Q2 and perhaps beyond. Lower than originally planned selling volumes will cause overhead costs to be spread over fewer units. Consequently, most builders have focused on reducing costs, including wage costs through layoffs, furloughs and cuts in executive compensation, as well as other overhead costs. Some of these layoffs will likely be permanent.
Along with the impact on their income statements, the builders have taken steps to shore up their balance sheets. All of the builders have delayed land purchases and slowed land development activity to reduce cash outflow. A few have drawn upon their bank revolving credit facilities in order to guarantee their liquidity. (However, Toll Brothers reported that it has paid down its revolver as it became clear that the capital markets were likely to remain accessible. Other builders may soon follow suit.)
So with a slower rate of revenue growth and likely higher operating costs, 2020 earnings for the homebuilders are likely to be meaningfully below where the market thought they be at the beginning of the year. As already noted, most builders have withdrawn their earnings guidance for 2020.
The current average consensus estimate (for 10 publicly-traded builders) is down 27% from six months ago, but up only about 3% from a month ago, according to my calculations of S&P Capital IQ estimates. Analysts have therefore taken note of the recent rebound in net orders, but have raised their forward estimates only a smidge so far. This conservative response appears to be consistent with the economic sentiment indicators and the adjustments that builders have made in their operations and finances.
Given the cautious approach by builders and analysts and considering that the dust has not yet settled on the reported economic and housing market data, it is therefore surprising that homebuilding stocks have run up as much as they have in recent weeks.
At current share price levels and with the decline in earnings outlook, homebuilding shares are trading at the historically elevated levels of about 13.2 times projected 2020 earnings and 12.2 times projected 2021 earnings (vs. recent historical averages for one-year forward estimates of about 11 times and 10 times for two-year forward estimates).
Although it can be argued that some premium may be justified given the recent cut in consensus estimates and therefore the potential that earnings could rebound in the second half of 2020 and into 2021, it is also possible that baseline earnings and near-term earnings growth could be lower than anticipated (and lower than recent historical trends). Consequently, I believe that investors should approach homebuilding shares selectively and with caution at current price levels.
June 10, 2020
Stephen P. Percoco
839 Dewitt Street
Linden, New Jersey 07036
© 2020 Lark Research. All rights reserved. Reproduction without permission is prohibited.