Determining how to value companies in a cyclical industry is a special case, because these earnings are generally more tied with the booms and busts of the economy.
The economy moves in cycles.
There are periods of economic expansion, where businesses are started and credit flows freely, driving up prices and prosperity. There are also periods of economic contraction, characterized by recessions, bad loans, job losses and flat or downward price movements.
Similarly, the stock market cycles through natural boom and bust periods, which generally follow the economy over the long term.
Some companies, and industries, are more affected by these economic and stock market cycles than others.
Those are the companies which we define as cyclical.
Which Industries Are Cyclical?
What’s interesting about business is that it evolves over time. Industries that were once considered cyclical may no longer be cyclical anymore, and vice versa.
For example, until the 2007 housing bust, real estate was considered counter-cyclical to the stock market. It was believed (and demonstrated for a long time), that real estate prices always went up.
Until they didn’t of course, but you can see how the definition of a cyclical industry can be opaque.
Just because something has been cyclical in the past, doesn’t mean it will necessarily be cyclical in the future. And various commodity industries may move through their cyclical patterns at different times, and some may never experience a down cycle due to limited supply and/or continuous demand.
Noting that there’s no perfect definition of a cyclical industry (or company), here’s a good reference from Morningstar to categorize the cyclicality of the 11 sectors of the stock market:
Within these stock market sectors are specific industries which can vary in their cyclicality nature compared to the overall sector. From the cyclical sectors in the screenshot above, some examples of a highly cyclical industry can include:
- Specialty Chemicals
- Metal Mining
- Internet Retail
- Auto Manufacturing
- Retail REITs
- Credit services
Each of the industries above can display cyclical characteristics, but this doesn’t mean that each industry will perform in-line with the stock market.
For example, specialty chemicals can be cyclical because the demand for those as raw materials can be higher during an economic boom. But, if the chemical companies within the industry expand their supply too much compared to the economic cycle, the oversupply will push the price of that commodity lower despite the high demand.
That can make the stock prices of those chemical companies perform poorly despite its cyclical nature, which is an important part to remember about investing in cyclical industries.
Just because a company is cyclical, doesn’t always mean it will perform as well as the economy.
Remember the relationship between price, supply, and demand.
Higher prices helps profits, but those tend to fluctuate in commodity-type businesses as supply is over/under built.
Why Are Certain Industries So Cyclical?
This great explanation in McKinsey and Co’s textbook on Valuation explains:
“Cyclical companies are characterized by significant fluctuations in earnings over a number of years. Earnings of such companies, including those in the steel, airline, paper, and chemical industries, fluctuate because of big changes in the prices of their products. In the airline industry, for example, cyclicality of earnings is linked to broader macroeconomic trends. In the paper industry, cyclicality is largely driven by industry factors, typically related to capacity.”
When businesses tend to compete on price instead of any other characteristic of a product or service, that price can fluctuate at a much higher rate.
The reason behind this is that when there’s no differentiation between a product, then a customer’s only logic is to buy the one at the cheapest price.
Understanding Supply, Demand, Capacity, and Cyclicality
Products with no differentiation, like commodities which are inputs to a larger application, will need to consistently lower their prices to capture customers.
At a certain point though, a price can’t continue to freefall (or it will put all of its producers out of business), so a floor gets set to cover the costs of producing/manufacturing. If price drops so low as to repel competitors from entering, as they see little incentive to participating in the industry, then the producer eventually absorbs enough demand at this lower price to run at full capacity.
In other words, once running at full capacity, the producer then thinks about raising prices to increase profits, and customers will pay it for as long as no competing product is available. If a competitor enters and also reaches full capacity, this also supports the price until the next price hike, and this continues in a virtuous cycle.
Eventually, supply will tend to outpace demand as high profit margins incentivize high building and overbuilding, which causes the cycle to turn downwards in the opposite direction.
These cycles do not tend to happen over days but rather over years, and is particularly pertinent due to the long construction times and investment costs to build out capacity that is present in many of today’s commodity businesses.
How to Value Cyclical Companies
Simply put, there are two major suggestions that I’ve seen on valuing the companies in cyclical industries from the textbooks (both McKinsey’s and Damodaran’s).
- Take a normalized average of earnings or FCF to use in a DCF.
- Construct projected performance scenarios on multiple macroeconomic conditions, and assign probabilities and weighted values to each in order to estimate a total valuation.
A third idea is to use a higher discount rate when estimating the valuation for a cyclical company. This also makes sense but I have not seen it quantified in how much to add to the discount rate, so it makes it hard to rationally apply.
Let’s take a real life example of valuation using a normalized average of performance, ideally from a time of both boom and bust for a company in a cyclical industry.
I’ll be taking FCF/share estimates and inputting them into a DCF valuation model; familiarize yourself with that concept first if you haven’t yet.
This company is called Westlake Chemical Company and is in the cyclical chemical industry. Essentially, the company produces plastics which are used in an expansive range of solutions. Chances are, you’ve probably interacted with a type of plastic produced by Westlake (or one of their competitors) within the last 24 hours, and weren’t even aware of it.
One of the drivers for plastics demand is construction, especially residential construction, and so that component of the demand for plastic makes it quite cyclical most of the time.
Keep in mind that the market for chemicals which form plastics is global, and so the supply/demand curves for this industry depends much more on the global economy than would something more U.S. based, like a homebuilder. Also keep in mind that $WLK earns over 30% of its revenue internationally, and so the value of the USD compared to other currencies also plays a role in its results, which can vary from the standard cyclical nature seen in other industries.
For this valuation I’ll be only concerned with a very rough estimate, just to show what an application of normalized free cash flows would look like in a basic DCF.
I like the website quickfs.net for a full visual of a company’s financials over a very long period of time, such as 10+ years. Taking a screenshot for $WLK:
We can see a wide range in free cash flow per share over this 10 year period. Because the production of chemicals is a pretty capital intensive business, we may see larger fluctuations of free cash flow vs regular earnings due to the large capital expenditures required to compete.
In this case, a look at EPS might be beneficial:
You can see how the EPS are higher than FCF, which further reinforces the idea that heavy capex is required for $WLK.
Plugging those 10 years into Excel, we can quickly get the following:
- Average = $4.49
- Median = $3.90
Using the Download Financials link with my quickfs.net premium account, I was able to calculate the following for Diluted EPS from 2000-2009:
- Average = $0.53
- Median = $0.48
Doing a CAGR calculator on the difference in Median EPS from 2000-2009 to 2010-2019 gave me an increase of 23.3%. This seemed really high, so I did the same calculation for Free Cash Flow per share, which clocked in with the following:
- 2000-2009 Median = $0.78
- 2010-2019 Median = $2.91
- 10Y CAGR (Median-Median) = 14.1%
The problem with these estimates is that the 2000-2009 period includes two bust periods for the chemical industry—namely 2002 and 2008 (for $WLK). You can see this by the negative EPS for the company in both years. And so, the 2000-2009 period might be artificially low as a starting estimate for growth.
You can see why valuing cyclical companies is hard.
There’s a lot of subjectivity in even how you normalize the EPS or FCF, which can cause your growth estimates and FCF estimates to vary wildly.
To stay conservative, I decided to first calculate the normalized FCF/share from 2008-2019 to include at least one bust cycle, taking the median value of these numbers: $1.75.
Next, I took a look at the revenue growth, which for this case the median YOY% from 2000-2019 = 13.9%. Note that this is close to the 10Y CAGR we found above so 14% as a growth rate might be a reasonable estimate.
Putting those 3 inputs together, I found a DCF valuation of $43.94.
This would be the valuation if we were to assume an equal 50-50 period of boom/bust over the next 10 years for the cyclical chemical industry.
Assuming a continuation of the last 10 years (mostly a boom), taking a FCF estimate of the 10Y median of $3.90, would revise the fair value DCF valuation to $97.93.
The likely truth for the valuation of $WLK likely lies somewhere in the middle, as the current market price of $85.27 seems to imply.
The question for how much to weight the optimistic (10Y boom) case versus the pessimistic (50-50 boom/bust) case becomes the million-dollar question in this kind of a valuation estimate, and probably is biased whether the analyst realizes it or not.
Earnings Estimates and Cyclical Companies
In the McKinsey book, researchers had found that earnings estimates were wildly inaccurate for companies in cyclical industries. In general, analyst estimates bias themselves upwards over the short term. While this might be appropriate for businesses whose earnings do not fluctuate wildly along the macroeconomic cycle, most cyclical companies did not exhibit a similar earnings behavior over a full business cycle.
Where earnings estimates may be a decent proxy for short term price performance and even longer term valuation, this relationship breaks down for cyclical stocks and should be widely considered when reading analyst reports and projections. There’s simply little incentive for predicting the booms and busts of a commodity cycle, and it’s an extremely difficult thing to do anyways.
And so, a valuation of a cyclical company based on the presumption that earnings or free cash will perform smoothly and continuously from recent trends is likely to be inaccurate, due to the real boom/bust nature of cyclical companies and their earnings.
There’s various different ways to look at valuing a company in a heavily cyclical industry, and it comes down to the fact that their earnings and free cash flows are much more volatile than a non-cyclical company.
The nature of a DCF valuation model is to assume a smooth, consistently growing level of earnings and free cash—which is obviously unrealistic but does get closer to reality with a non-cyclical than with a heavily cyclical company.
It’s very important to be careful to differentiate between a commodity business and a cyclical business. Though often the same, sometimes they’re not.
I’ll leave with a great quote from Peter Lynch, which pertains to the valuation of companies in a cyclical industry quite well: “In cyclicals, a period of silly prices is followed by a period of sobriety.”