Guest Post – Homebuilder Pulte (PHM) Fundamental Analysis

Today, we cover – Homebuilder Pulte (PHM) Fundamental Analysis

Jason here with a quick note…

The following is a guest post from H.C. Eu over at the value investing blog Investing for Value.

Here is his third – of hopefully many – guest posts here on this site.  And if you love this analysis as I know you will make sure to check out his site at either the link above or those further below at the end of the analysis.

Also, make sure to show H.C. Eu some love in the comments below or once this hits our social media pages as well.

Now, I’ll let H.C. take it away with his article on How To Manage Permanent Loss of Capital.

I hope you enjoy it.

Always in your service,

Jason Rivera.

***

The market is not valuing Pulte as a cyclical stock

Pulte Group Inc (PHM or the Group) is a homebuilder. The homebuilding sector is a cyclical one. As such any fundamental analysis of PHM or the Group should be based on its performance over a cycle. Looking at performances over the past few years may not provide an accurate picture.

“Cyclical…companies share a common feature, insofar as their value is often more dependent on the movement of a macro variable (the commodity price or the growth in the underlying economy) than it is on firm specific characteristics…the biggest problem we face in valuing companies… is that the earnings and cash flows reported in the most recent year are a function of where we are in the cycle, and extrapolating those numbers into the future can result in serious mis-valuations.”  Damodaran

The challenge then becomes what to use as the values over the cycle. The current Housing Starts cycle started in 2005. Note that I defined a cycle as one peak to peak period. There is still debate about when the current cycle will peak.

The analysis and valuation of PHM then boils down to

• Forecasting the end of the current cycle so that we can determine the duration of the current cycle.
• Assessing PHM performance over this cycle.
• Valuing PHM assuming that the long-term value is based on the average values of the valuation parameters over this cycle.

I will illustrate one approach to handle the above in this post.

Investment thesis

The homebuilding sector is a cyclical one. Any long-term fundamental analysis of PHM should be via a cyclical lens. This is especially since the Group had not been able to grow its revenue from 2005 to 2021.

The US Housing Starts is currently in the uptrend part of the cycle. But it will eventually mean revert. When this happens the revenue of PHM will also mean revert. The long-term performance and value of PHM will then be based on the long-term average values of the various metrics.

There is no margin of safety based on such a valuation scenario. I would not invest in PHM from a long-term value investment perspective.

Rationale

•     PHM is basically a homebuilder. The homebuilding industry is cyclical and this is reflected in the Housing Starts. The sector is currently in the uptrend part of the Housing Starts cycle. I expect the current Housing Starts to overshoot its long-term average but it will eventually decline and mean revert.
• PHM revenue performance has not been great. Its projected 2021 revenue is still below that of 2005. It has also been losing market share. Over the past 17 years, there is a 0.9 correlation between the Housing Starts and PHM’s revenue. When the Housing Starts revert to the mean, I would expect PHM’s revenue to follow suit.
• The key components of returns – operating margins, asset turnover and leverage – should then be based on the values over the housing cycle. Looking at the past few years data would not be realistic given the cyclical characteristics.
• The valuation of PHM based on the average values over a cycle shows that there is no margin of safety at the current price.

PHM is in a cyclical sector

PHM is basically a homebuilder. The Group itself had mentioned in its Form 10k that the home building industry is cyclical. The cyclical nature of the sector is illustrated by the Housing Starts chart below. Over the past 70 years there have been several cycles. 

Chart 1: Housing Starts   Source: Trading Economics

The key takeaways from the Chart are:

•       There does not seem to be any uptrend in the long-term average Housing Starts. We have a “stable” long-term average of about 1.5 million units per year.
•  The 2020 Housing Starts is around the long-term average.

Any long-term projection of the Housing Starts should thus be based on the reversion to the mean. This is not a growth sector from a long-term Housing Starts perspective.

If I consider a cycle as peak to peak, the current cycle seemed to have started around 2005. The bottom of the current cycle appears to be around 2010. We are currently in the uptrend part of the cycle. Even though it has been 17 years, there is still debate about when the current cycle will peak.

Will Housing Starts overshoot the long-term average?

There are two key macro trends driving the current demand for houses in the US:

•       We are currently in the tenth year of the uptrend part of the Housing Starts.
• US housing deficit. There is currently a national housing shortage with estimates varying from one million homes to 5 million homes.
• Looking at both of these trends, I would conclude that the Housing Starts would overshoot the long-term average in the current cycle. To get a picture of how it will overshoot, I used the following model:
• The Housing Starts reached the long-term average of 1.5 million units in 2020/21.
• The Housing Starts will continue to grow at CAGR of 8.9 %. This was the CAGR from 2010 to 2020.
• The Housing Starts will then peak and decline with the same profile as during the uptrend.
• I looked at 2 scenarios. The first is that the total excess is about 1 million units – Conservative scenario. The second is when the total excess is about 5 million units – Optimistic scenario. The excess each year is the number of Housing Starts greater than 1.5 million units

The chart below shows the Housing Starts projections under the 2 scenarios.

Chart 2: Housing Starts Projections

•     If you take the Conservative scenario, the peak will be around 2024. The current cycle will then be 20 years long with the peak level below the 2005 peak
•      If you take the Optimistic scenario, the peak will be around 2027. The current cycle will be 23 years long with the peak around the same level of the 70 years highest peak.

I do not expect any changes to the long-term average given that it is a 6-decades Housing Starts average.

Industry revenue growth to be driven by price increases

While long-term average Housing Starts did not show any growth, house prices behaved differently. Based on the Federal Housing Finance Agency House Price Index, house price increased at a 4.3 % CAGR from 1992 till Aug 2021.

While there is no growth in the long-term average Housing Starts, the sector revenue would still have growth due to unit price growth. But this is growth at about the US long-term GDP growth rate. Again, it showed that this is not a growth sector.

Note that house prices are influence not just by the Housing Starts, but also by the supply and demand of existing properties as well as the economic situation.

Chart 3:  US House Price Index   Source: Trading Economics

PHM has lost market share

At the start of the current cycle in 2005, PHM achieved a revenue of USD 14.7 billion. The LTM PHM revenue for 2021 is USD 12.8 billion. In other words, PHM did not achieve any revenue growth over the past 17 years. This is despite undertaking a number of asset acquisitions.

I also compared PHM revenue from 2010 till 2020 with those of the top 10 homebuilders and found that PHM share had declined from 20% in 2010 to 12 % in 2020.

Chart 4: Peer Revenue

Note: The top 10 companies are the top 10 companies by their 2020 revenue. The total for the top 10 in the chart included PHM.

Given this history, I do not expect PHM to grow by taking market share away from its peers. Rather any growth then has to be due to industry growth. But there is no growth in the long -term average Housing Starts.

During the period from 2005 to 2020, there is a 0.9 correlation between the Housing Starts and PHM’s revenue.  These mean that when the Housing Starts mean revert, we should also expect PHM’s revenue to mean revert.

In other words, PHM long term revenue over several cycles should be pegged to the average long-term Housing Starts. The 2020 Housing Starts is around the average long-term level. As such the long-term average revenue for PHM (assuming no price increases) should be equal to its 2020 revenue.

This is not to suggest that there would not be any revenue growth for PMH.  Rather any long-term revenue growth would mainly be tied to a price increase rather than an increase in unit sales volume.

Is PHM becoming more efficient?

Over the past 17 years, the ROE for PHM ranged from -52 % to 56 % with an average of 2 %. This is not a good average return.

There are 3 major financial metrics that drive ROE – net margins, asset turnover and leverage. These represents operating efficiency, asset use efficiency, and financial leverage.

The first two components assess the operations of the business. The larger these components, the more productive the business is. The last component captures the company’s financial activities. The more leverage the company takes, the higher the risk of default.

I wanted to see how these parameters for PHM have performed since the last Housing Starts peak. I plotted the Gross Profit (GP) margins, Selling General and Admin (SGA) margins, Asset Turnover and Debt Equity ratio relative to their respective 2005 values. Refer to the chart below.

•       GP margins in the uptrend part of the Housing Starts cycle were above that of 2005.
• From 2005 to 2021, the SGA margins and were worse than their respective 2005 values.

The picture is one of deteriorating performance for 2 of the key drivers of returns during the past 17 years.

Chart 5: Operating Trends

For a cyclical company, the debate is how much the past few years performance is due to being in the uptrend part of the cycle. If a considerable part is due to the cyclical position, then using the past few years values would not present an accurate picture.

To overcome this debate, I use the average values over the cycle. This is especially for revenue and GP margins where there are strong links between their performance and the position of the cycle.

Leverage is reducing

As can be seen from the chart below, PHM has been able to reduce its leverage as measured by the Debt Equity ratio over the past 17 years. Unlike revenue or GP margins, capital structure is more within management control compared to the effects of the cycle.

Chart 6: Debt Equity Trend

Valuation.

The reasons I focussed on the above operating metrics is because they feed into my valuation model. At the same time, if you accept that PHM is a cyclical company, then its valuation should be based on the average values over the cycle.

I used the following financial model to perform such a valuation. In the model the profits and capital structure are linked to Revenue as follows:

EBIT = GP – SGA

Debt = TCE X Debt ratio X market cost of debt factor.

Where:

GP = Revenue X GP margin

SGA = Revenue X SGA margin

TCE = Total capital employed = Revenue X TCE margin. I defined TCE as SHF + Debt – Cash.

Debt ratio = Debt/TCE

Margin = respective metric as a % of Revenue. For example, the TCE margin = TCE/Revenue.

Market cost of debt factor = 1.06. It represents the factor to convert the Book Value of Debt to the market value of Debt. This is assumed to be the same as that used to derive the WACC.

I used a single-stage Free Cash Flow to the Firm model to value the group. I then derived the equity value by deducting the derived Debt and adding back the current (Sep 2021) cash.

You will note that the equity value depends on the efficiency of the operations (as measured by the GP and SGA margins) and the capital structure.

In the Free Cash Flow model, I have assumed the business is at the long-term average level. I defined the long-term average revenue as equal to the revenue with the Housing Starts at 1.5 million units per year. For PHM, this is equal to the 2020 revenue. The other assumptions for the model are summarized in the table below.

Note

(a) The cost of equity and WACC is based on the default spread approach.

(b) The is based on bond pricing formula using PHM’s 2020 debt payment schedule

I then consider 2 scenarios:

•        Conservative based on the Housing Starts peaking in 2024
•  Optimistic based on the Housing Starts peaking in 2027

For each parameter under each scenario, I took the average values over the duration of the cycle. For example, for the GP margin under the Optimistic scenario, I took the average GP margin from 2005 to 2027 derived as follows.

•       The values from 2005 to 2020 were actual values
• The value for 2021 was based in the LTM values.
• The values for each year from 2022 till the Housing Starts peak was assumed to be the 2017 to 2021 average values. I assumed that values are the same for each year.

The values of the various parameters and the corresponding valuations of PHM are summarized in the following table.

The market price of PHM as of 3 Jan 2022 is USD 56 per share. You can see that there is no margin of safety under both the Conservative and Optimistic scenarios. I also reversed engineered the market price to get an understanding of what the market is assuming.

Reverse engineering the market price.

There are two sets of parameters in my financial model.

•      Those that are very dependent on the market cycle. I would put the GP margins here. Looking at the trends in Chart 5, I would put its SGA margins and TCE margins here. Note that I have defined TCE margin as TCE/Revenue.  This is effectively the inverse of the Asset Turnover shown in Chart 5.
• Those that are more dependent on management action. I would categorize the Debt/TCE here

To reverse engineer the market price, I have assumed

•     The TCE/Revenue, Debt/TCE and SGA margins will be the past 5 years average values. In other words, I assumed that the efficiency and productivity achieved over the past 5 years will be sustained.
• The GP margin is based on the Optimistic scenario.

As you can see, the current market price is based on the SGA margins and Asset Turnover as non-cyclical parameters. This is contrary to PHM historical performance for these 2 metrics.

Even then there is no margin of safety if you invest at the current price. If you believe that the market price has room to go higher based on fundamentals, you must believe that there would be further improvements in these metrics in the future cycles. 

Conclusion

If you accept that PHM is in a cyclical sector, you have to analyse and value it through a cyclical lens. The challenge then boils down to determining what to use as the average values over the cycle.

In my analyses, I have assumed that the current cycle will take a few more years to complete. I then determined the average cycle values by taking the historical values and the extrapolated ones.

Using these cycle averages, I found that there is no margin of safety at the current market price. As such I would not invest in PHM from a long year value investment perspective.

The above analyses are based on one way to determine the value over the housing cycle. I have another way to undertake such an analysis. Refer to my blog article Is MDC Holdings one of the better NYSE stocks?

Over the past few weeks, I have used the same framework to analyse several homebuilders such as Lennar and D.R. Horton. The feedback can generally be classified into:

•      I should not be using the national Housing Starts but the regional ones to get a better picture of the cycle.
• The historical trend in the House Price Index may no longer apply due to socio-economic changes.

The key point is that we can debate the cyclical values and the growth rates. But I don’t think there should be any debate about just using the past few years data as representative of the future because these do not represent the values over a cycle.

So be wary of any analysis and valuation that just looked at the past few years or even past 10 years of data. They just represent the values for the uptrend part of the cycle and not the full cycle.

***

Editor’s Note: The article is from H.C. Eu who blogs at Investing for Value. He is a self-taught value investor and has been investing in Bursa Malaysia and SGX companies for more than 15 years. His value investment experience has been enhanced by both his Board experiences and his contacts with controlling shareholders of many Bursa-listed companies. These have given him a unique opportunity to be able to analyze and value companies differently from other research houses.  If you enjoyed this piece, you can find similar pieces and other value investing tips in his blog.

H.C. Eu is not an investment adviser, security analyst, or stockbroker.  The contents are meant for educational purposes and should not be taken as any recommendation to purchase or dispose of shares in the featured company.   Investments or strategies mentioned may not be suitable for you and you should have your own independent decision regarding them. The opinions expressed here are based on information he considers reliable but he does not warrant its completeness or accuracy and should not be relied on as such. He does not have any equity interests in the company featured.