Homebuilders are cyclical businesses, but do you filter them as such?



Homebuilders are cyclical businesses. As such, they should be selected based on their performance during the housing starts cycle. But housing starts cycles differ in duration and amplitude from peak to trough.

One way to select these companies is to estimate the expected ROE and margin of safety. This takes into account the probabilities and values ​​of the metrics on the pessimistic and optimistic cycles.

Get the full series in PDF

Get the complete 10-part series on Charlie Munger in PDF. Save it to your desktop, read it on your tablet, or email it to co-workers.

Q2 2022 Hedge Fund Letters, Talks & More

Hedge funds reconsider their tech holdings after heavy losses

AssessThroughout 2020 and 2021, some of the biggest names in the hedge fund world have reaped huge profits as shares of tech stocks, which have featured heavily in their portfolios, have surged. However, this year the trend has reversed. Stocks that were once among the hottest in the market have fallen out of favor, Read More

At the same time, there is no long-term growth in average annual housing starts. As such, the margin of safety should be based on the value of earnings power over the cycle.

Looking at the top 5 home builders based on the above, I have identified Pulte Group Inc as the best.

Housing starts

US housing starts in July 2022 would be 8.1% lower than the July 2021 rate. The debate then is whether this confirms the current downward trend in housing starts or if it is just volatility. In other words, the upward trend in housing starts that began in 2010/11 is still intact.

You can imagine the impact of investing in home builders. If the uptrend is still intact, you would say that most homebuilders are currently cheap. But if you think this is the start of the downtrend, you may want to think twice about investing in these stocks.

I think the debate misses the point. It doesn’t matter where we are in the housing starts cycle.

This is a cyclical industry and the fortunes of homebuilders are tied to the housing starts cycle. Look at the correlation from 2007 to 2021 between housing starts and revenues for the top 5 homebuilders, as shown in Table 1. The correlations are well over 85%.


Table 1: Correlation

Note: The first 5 were based on the Diving under construction 2021 top 10 article. #5 was Taylor Morrison where I couldn’t get the data going back to 2007. As such, I replaced it with MTH which was ranked #6.

The key to analyzing cyclical companies is not to worry about their position in the cycle, but to look at their performance throughout the cycle.

“…the biggest problem we face in valuing (cyclical) companies…is that earnings and cash flow reported in the most recent year are a function of where we are in the cycle, and extrapolating these numbers into the future can lead to serious errors -valuations…trying to forecast the next cycle is not only futile but dangerous and far better to normalize earnings and cash flows throughout the cycle.

But there are 2 characteristics of US housing starts that should be considered when evaluating home builders. Referring to Chart 1:

  • There doesn’t seem to be any long-term growth.
  • The current cycle low is exceptionally lower compared to previous lows

Housing starts cycle

Chart 1: Housing starts cycle

Source: Tradingeconomics.com

Model without growth

When you compare the pattern in Chart 1 with those in Chart 2, you can see that the pattern for housing starts is more similar to the one shown in blue in Chart 2. This is the cyclical pattern with no growth. The cyclical pattern with long-term growth is shown in green in Chart 2.

Types of Cyclic Patterns

Chart 2: Types of Cyclical Patterns

Source: Author

The implications of the non-growth characteristic are then:

  • The annual average for long-term housing starts is approximately 1.5 million units. In rating homebuilders, we should look at their normalized revenues based on 1.5 million units.
  • If there doesn’t seem to be any growth, it makes more sense to focus on purchasing power value (EPV).

Different depth and cycle time

The duration and range (from peak to trough) of each housing starts cycle are not the same. As shown in Chart 1:

  • The cycle from 2007 to 2021 lasts 14 years and peak to trough is around 1.8 million units.
  • The cycle from 1972 to 1977 lasts 5 years with a peak to trough of approximately 1.6 million units.

The performance of a business over a cycle would vary depending on the duration and values ​​from peak to trough.

  • During the downtrend portion of a cycle, there could be asset depreciation. A longer duration could aggravate the situation.
  • Gross profit margins and SGA margins would likely be different during the downtrend versus the uptrend.

The challenge then is how to normalize performance over the cycle if there is no consistent cycle pattern? One way is to calculate the expected scenario based on the probability and performance of a pessimistic scenario and an optimistic scenario.

  • The conservative view is to take the values ​​over the cycle from 2007 to 2021.
  • The optimistic view is to peg the normalized values ​​to the year when housing starts were at the long-term average of 1.5 million units. These are the 2019 and 2020 performances.

Standard screens

I used two metrics to screen companies – normalized ROE and normalized margin of safety.

For the conservative scenario, I normalized the measures based on the average annual values ​​from 2007 to 2021. For the optimistic scenario, the normalized measures were based on the average of the values ​​from 2019 and 2020.

To illustrate the results of these screens, I compared the performance of the top 5 home builders, as shown in Tables 3 and 4.

Conservative screen

Table 2: conservative screen

Notes: Safety margins were obtained by comparing market prices as of September 2, 2022 with the respective EPV.

Optimistic screen

Table 3: optimistic screen

As can be seen, there is no margin of safety based on the conservative scenario. Based on the optimistic scenario, DHI, PHM and MTH passed the test in terms of positive ROE and margins of safety.

These are extreme views. It is possible to assign probabilities to each scenario and then calculate the expected ROE and safety margins.

One way to estimate probabilities is to look at historical performance according to Chart 1. Of the 8 cycles, there has only been one where the bottom hit 0.5 million units. On such a basis, I estimated the probability of the conservative scenario at 0.125 (ie 1/8) and that of the optimistic scenario at 0.875. Based on this, I got the expected ROE and the expected margin of safety.

I plotted the results as shown in graph 3. In the graph

  • The vertical axis represents ROE
  • The horizontal axis represents the margin of safety. I defined the Margin of Safety = Price / EPV.

Screen based on the expected values ​​on the cycle

Graph 3: Screen based on the expected values ​​on the cycle

The screen results suggest that you should focus your fundamental analysis on Pulte.

If you based the analysis on LTM values, you have all 5 companies with good margins of safety and ROE as shown in Table 4. This is the danger warned by Damodaran.

LTM screen

Table 4: LTM screen


My safety margins depend on how I got the EPV. I evaluated home builders based on the one-step Free Cash Flow to the Firm model where:

Firm value = FCFF / WACC

EPV = Firm Value + Non-Operating Assets – Minority Interests – Debt

FCFF = Free Cash Flow to the Firm = EBIT X (1 – t)

EBIT = Gross profit – SGA

Gross Profit = Revenue X Gross Profit Margin

SGA = revenue X SGA margin

The respective gross profit margins and SGA margins are as shown in Table 5.

Assumptions on GP margins and SGA margins

Table 5: Assumptions for GP margins and SGA margins

t = nominal tax rate assumed to be 21%

Turnover = Respective long-term turnover. This is assumed to be the 2021 Revenue X (1.5 / 1.7) where 1.5 is the long term average for housing starts and 1.7 is the housing starts for 2021.

WACC = derived according to the Damodaran approach. I assumed a risk-free rate of 2.52% based on the average 10-year treasury bill rates from 2007 to 2021. Deleveraged beta = 1.59. Equity risk premium = 4.24%

Non-operating assets, minority interests and debt are assumed to be 2021 values.


Home builders are cyclical businesses. When analyzing them, the advice is that you should look at performance over the cycle. The challenge is that the duration and magnitude of cycles from peaks to troughs are not consistent.

To work around this problem, I estimated the expected ROE and safety margin. The expected values ​​were based on the probabilities and values ​​of the pessimistic and optimistic measures over the cycle.

Looking at the top 5 home builders on such a basis, I identified Pulte Group as the one for further analysis. I hasten to add that this is not an investment thesis. There is still further analysis – financial situation, potential deficiencies and management. Still, the results suggest that it is a good candidate for further analysis.

Do you really want to master value investing

Editor’s Note: The article is by HC Eu who blog at Investing for Value. He is a self-taught value investor and has been investing in Bursa Malaysia and SGX companies for over 15 years. He recently published a value investing book “Do You Really Want to Master Value Investing” on Amazon. This will be available for free download on September 13, 2022, Pacific Date Time.

About Myra R.

Check Also

Yen jumps after authorities intervene for first time since 1998

Join now for FREE unlimited access to Reuters.com Register TOKYO/LONDON, Sept 22 (Reuters) – The …