Updated: Apr 8, 2021
We wrote about why macro is important for stock pickers in our last post. This inevitably brought us to a component of our macro analysis we use to identify trends. An understanding of the forces that drive demand for the product or service of an underlying firm in our portfolio is a necessary ingredient for us to form conviction. We share with you how we define a trend, other forces that are often confused with trends, and finally how this knowledge translates into portfolio positioning.
Webster defines the word “trend” as a “general direction in which something is developing or changing.” Since this definition is far too vague for any practical use in the investing world, we took it upon ourselves to torture and adapt the meaning to suit our purposes. We find more utility to re-define a “trend” by the following: A phenomenon that exerts an irreversible pressure on penetration rates for a good or service. (Penetration rate = number of customers divided by the total addressable market). Since the total addressable market (TAM) can sometimes be hard to estimate, especially for emerging trends, we tend to be biased towards the broadest definition possible. For goods and services that sell to consumers, our TAM will tend to be the global population and for those that instead sell only to businesses, our TAM will tend to be all businesses. Because our denominator is so large, the level of the penetration rate is of minor importance relative to the shape of the penetration rate over time. Below is the penetration rate of mobile phones over the past decade. All trends over their lifetime tend to exhibit a similar shape. The penetration rate grows faster in the beginning then, after an inflection point, begins to slow.
Trends reflect how our world is changing. If the phenomenon in question can potentially reverse, i.e. the penetration rate starts to go down, it is not a trend. Trends may either accelerate or decelerate but they do not reverse or die, they mature. For example, Henry Ford invented the Model T in 1908. It was the first affordable automobile built for the middle class. Assisted by sociological and industrial realities, the Model T ignited the trend of automobile ownership. Automobiles were once a big idea, but now you wouldn’t think twice about your neighbor driving the latest Honda hybrid.
“I will build a motor car for the great multitude..constructed of the best materials… the simplest designs that modern engineering can devise…. so low in price that no one making a good salary will be unable to own one and enjoy of pleasure in God’s great open spaces. When I'm through, everybody will be able to afford one and everyone will have one. The automobile will be taken for granted...” H. Ford
Henry Ford succeeded in changing what feels “normal” to us. The above quote perfectly sums up the beginning of a trend and the vision of how a trend matures.
What else influences penetration rates?
Other forces that influence penetration rates that are just as important to understand but routinely confused with a trend are either: cyclical, faddish, or one-time shifts(1). Since most of the purported consumer behaviors tend to be either cyclical or faddish, we focus on distinguishing them from trends.
Trends vs. Cyclical forces
Cyclical patterns are easy to spot in a time series with penetration rate on the y-axis and time on the x-axis. An undulating pattern is usually always an indication of cyclicality. Seasonality in certain pockets of the retail market, for example, is a clear example of this. Everybody knows a higher proportion of people own Christmas trees in December, but there does not always have to be a clear intuitive reason for the cyclicality. Take homeownership rates in the U.S. below. Although homes are considered an investment asset by many and the cyclicality will inherently be driven by a change in price, it is not the only factor contributing to the cyclicality.
The reasons for this undulation are multi-variate and require an understanding of policy-created incentives, interest rates, securitization of mortgages in the banking sector, economic cycles, and the effect rising and falling prices have on the psychological disposition of buyers. Whatever the reason, homeownership rates are indeed a cyclical force with indeterminate upcycle and downcycle lengths and will never represent a trend. It is important to stay mindful of this because during extended upcycle or downcycle periods, like from 95’ to 08’ with home ownership rates, well-reasoned explanations will start to seep into the zeitgeist claiming that we are in a “new normal.” In 2005 David Lereah published a book famous for all the wrong reasons called “Are You Missing the Real Estate Boom?: The Boom Will Not Bust and Why Property Values Will Continue to Climb Through the End of the Decade.” Many believed more and more people would continue to own homes and that it was virtually impossible for home prices to decrease.
A recent example of how we have used our understanding of cyclical forces to help initiate a portfolio position was our thesis on Donnelley Financial (DFIN). A large part of Donnelley’s business is linked to capital markets activity such as IPOs that are cyclical in nature. Toward the middle of 2020, IPOs had been in a downcycle for quite some time and there were a host of explanations claiming that the private markets had too much capital, appeasing shareholders was too onerous and there wasn’t an advantage to going public. Wall Street Journal published an article in 2017 titled, “America’s Roster of Public Companies Is Shrinking Before Our Eyes | Gusher of private capital, IPO slump and merger boom cause listings to plunge; There’s no great advantage.” This was the prevailing sentiment at the time. While these reasons held some water, it was clear to us that public markets still offered the most accretive way for founders and employees to monetize their ownership. Not to mention it is the biggest scoreboard in town. IPOs were here to stay and although we didn’t know when we knew there would eventually be an upcycle. This understanding helped add to our conviction that the future of DFIN would be brighter than its past.
How to distinguish between a Fad and a Trend?
When a trend is in its embryonic stages it can be indistinguishable from a fad. There is generally a lot of associated speculative activity and bubble-like behavior. Internet business models did indeed change the world but were accompanied by euphoria and subsequent panic selloffs in their early years. Sometimes a fad is incredibly obvious and hysteria and delusion seem to be the primary driving force. The most obvious example of this we know of in the recent past is the Beanie Baby craze. This episode was purely speculative. To quote Zac Bissonnete, author of The Great Beanie Baby Bubble, “The first buyers had been children with allowances. Then their moms had started collecting. By the time of the 1998 Ty Christmas party, a store owner remembers, it was mostly ‘creepy, belligerent men’ she saw lined up when she dropped in to check on retailers.” Unlike the Beanie Baby craze, most fads will tend to have plausible explanations behind their adoption. There isn’t an easy answer that helps us distinguish between fads and embryonic trends in real-time. We believe wearable technology (sans watches) to be a fad but this could be laughably wrong when we look back at this prediction in distant the future. There are two lessons we can offer: 1) Sometimes fads are only known to be such in hindsight. 2) Pay special attention to fad-like forces that should die but keep persisting or even growing stronger.
Examples of current trends
Health and wellness – A larger and larger percentage of the global population is increasingly health conscientious
Consumerization of enterprise technology – higher % of enterprise technology will have started in the hands of everyday consumers.
E-Commerce – Percentage of retail transactions done online will continue to increase
Rise of the Gig economy as a result of online platforms – the percentage of contractors vs W2 will continue to increase
Digitization – More businesses will translate offerings to digital.
Cloud Computing – More businesses are moving to the cloud.
Work from home – More businesses will offer WFH employment
Artificial intelligence – AI will be used to make more and more decisions. Effectively lowering the cost of prediction.
Cybersecurity – More companies will focus dollars on cybersecurity
Electric vehicles– A higher proportion of cars will be electric
Open source & open access – More technologies will evolve to be open source and open access.
Cryptocurrency usage – Higher % of online transactions will be completed in cryptocurrencies.
Now that we have identified some trends, now what?
Most of the trends we listed are widely understood and priced in by markets. Others, like the consumerization of enterprise technology, are less widely known and may not be easily understood. Typically for investing purposes, the less widely known the trend the better. However, knowledge of trends is not a sufficient condition to make a good investment. As we already mentioned with the advent of the internet, lots of people knew it was going to fundamentally change the ways business operated but many lost their shirts during the internet bubble. Maybe you understood that social media was a surging force but bet on Myspace. The point is there exists a chasm between identifying a trend and making a viable investment in a company that complements that trend. The root-level work to understand which companies will benefit and have staying power is necessary. Even then our job is not complete, we still have to reconcile all this with the price the market is offering us. The longer the runway and the lower the level of the penetration rate for the trend, the less entry multiples might drive our decision. Each situation is different and involves a thorough excavation of company strategy, fundamentals, and risks. We wouldn’t have it any other way.
1. We do not focus on one-time shifts given they are less frequent than cyclical and faddish forces. One-time shifts are a result of exogenous shocks that have a step-function effect on penetration rates. The percentage of people who own N95 masks pre and post COVID is an example of this.