‘What we need to believe’ for an investment thesis — a case study on Peloton

As growth investors, a necessary part of our jobs is to make inferences about the future.


While this is part and parcel of our role, it is certainly no easy feat. Predicting the future is close to impossible, as by definition it implies making guaranteed statements about something that has not yet happened. Rather than making predictions, like Howard Marks often says, we believe it is more important to make inferences — conclusions based on facts, observations, and probabilities. The distinction is subtle, yet critical to how we think as investors.

For example, it doesn’t matter if they are seen as the best team in the competition- no one can definitively predict whether the Golden State Warriors will win the NBA championship…but we can infer a probability of them doing so, by triangulating indicators such as team form, home court advantage and historical winning percentages. Put another way, rather than saying they will win, we would prefer to say, ‘based on X,Y,Z, there is a 4 in 5 chance of Golden State Warriors winning’ (but also a 20% chance that they lose).

One method we can use to think in inferences, especially to test investment theses, is to ask ‘what we need to believe?’. This mindset involves breaking down forecasts into objective implications, relative to things we know today. Effectively, we frame the future around concrete facts that we can verify — which makes it easier for the human brain to assess probability.

This mindset becomes even more important when we look at fast-growing, generational category leaders. At TDM we have been proud to compound our client’s capital in excess of 25% per annum for over 16 years. Because we have such a high investment hurdle rate, in instances where a business may be priced richly relative to its near-term sales / earnings, we need to believe that the company will grow at very high rates for long periods of time to meet our return profile.

To provide more colour, below is an example of how we applied ‘what we need to believe’ thinking to a potential investment opportunity in Peloton — a company we view as the category leader of Connected Fitness.

The question with Peloton (PTON):


Led by a visionary founder John Foley, Peloton is the first company to have masterfully combined hardware, software, logistics and content at scale, to deliver a seamless connected fitness experience. Founded in 2012, Peloton now helps millions of consumers around the world to become better versions of themselves, allowing them to exercise at home and on-the-go, as part of a highly engaged, virtual community.

Peloton has a lot of the characteristics that we look for in a generational business — “one mission, one platform”, outstanding customer love (Net Promoter Score (NPS) greater than 90), strong unit economics, and a first-mover advantage in disrupting a huge consumer market.

Anyone who has followed the incredible growth trajectory of the business — $200m to $4b in revenue in the last four years — likely knows though, that Peloton is what investors call a ‘battleground’ company. For as many bulls (“Peloton is revolutionising fitness!”)…there are just as many bears (“Only rich people can afford it”, “it’s just a COVID fad”, and perhaps our favourite— “it’s just a bike with an iPad!”).



The contention between these two corners — and of course there is middle ground — typically revolves around the debated size of the total addressable market (TAM). How many subscribers can Peloton in fact achieve globally? For if you are to be a buyer of the business at today’s prices (c$35b market cap), then you must believe that the size of the Connected Fitness market is not just large today but will grow well into the future.

Not only does this require analysing Peloton’s competitive advantages, but it also requires us to infer the future of a relatively new market. Specifically, how do we get comfort around the number of subscribers that Peloton could generate 5 to 10 years from today?

The answer is not an easy one, especially when it is a category-defining business.


Asking ‘what we need to believe?’ The 5-year view

There are many things you probably need to believe for Peloton to be a successful investment at today’s prices. This could range from your beliefs on consumer behaviour, to Peloton’s product strategy and even to macroeconomic conditions. Ultimately these factors (and others) feed into the future number of Peloton’s connected fitness subscribers, which we see as the most important metric in assessing Peloton’s long-term value creation. Thus, we wanted to focus our analysis on this metric.

Peloton has c2.3m connected fitness subscribers today. Core to achieving a 20% plus return is the belief that Peloton can globally expand its Connected Fitness platform over the next 5–10 years, resulting in substantial growth in subscribers.

To achieve this return, we believe you need to assume a potential of 12 million and 25 million subscribers in 5 and 10 years respectively.

Working backwards, we wanted to infer what these subscriber numbers meant in terms of global household penetration and TAM. We did this through the following steps:

1. Identify key fitness markets by country, including over 20 countries that PTON had filed trademarks in.

2. For each country, collect a host of demographic information including number of households, number of gym members, average household wealth and average population density.

3. Analyse Peloton’s expansion and membership numbers over the last 5 years to deduce how many households Peloton had penetrated already and what the ‘ramp-up’ curve looks like in each country they had entered (USA, Canada, Germany, UK at time of analysis).

4. Infer (based on the above) what future ramp-ups could potentially look like for new markets, acknowledging that COVID both accelerated penetration but also increased brand awareness.

5. Think through (using our Peloton Strategy Officer hat) what the most likely next markets for roll-out could be — informed by management calls, attractiveness of different markets and logistical synergies.


Exhibit 1 captures the output of our work. It shows one scenario of future expansion and the penetration rate for households Peloton would have to attain by 30 June 2026 (red border bubbles) to believe PTON can reach 12m subscribers. nb: Household income % are estimates only based on best available information


Of course, we would not be surprised if what Peloton ends up doing looks quite different to the above. There are many permutations of roll-out and penetration possible. The point of this exercise was not to predict the future, but to come up with a reasonable scenario that challenges us to ask whether we believe this is possible or not. In other words, given what we know, is this a bet we are willing to make?

By plotting 3 additional metrics for each market — % of wearables penetration (yellow bubble), % of gym membership penetration (blue bubble) and % of households above $75k USD income (green bubble), we were able to better triangulate if Peloton’s inferred household penetration in each market was feasible.

This analysis helped to surface several take-aways;

  • Despite being the most mature market, the US still accounts for almost half of the incremental subscriber additions over the next 5 years and is still the most important market over that timeframe.
  • Germany is a key market and will require strong execution to achieve our targets, especially as the first non-English speaking market.
  • China remains a wild card with even very small penetration assumptions amounting to big subscriber additions. Understanding and scrutinizing management’s China strategy will be key over the long-term.
  • Fitness is truly a global market, and we would have to believe Peloton and Connected Fitness has the traction to expand into most international markets just as smartphones proved they had universal appeal.

From this exercise, you can start to see what someone might need to believe in order for Peloton to reach 12 million subscribers in 5 years. In many countries, the required household penetration rate for Peloton (to meet our target returns), is well below the current % of households with gym memberships or wearables. By anchoring the future around observations that are true today, you can see how it may be easier to think about probability & ambiguity.


Asking ‘what we need to believe?’ The 10-year view

When we get to a 10-year horizon, it’s hard not to venture into the realm of crystal ball-gazing! This timeframe is extremely difficult for humans to process. Just consider several things that weren’t around 10 years ago:

  • Hungry? Want food delivery to your door? Sorry- no such thing as Uber Eats, Doordash or Grubhub…
  • Struggling with love — want to swipe right? No Tinder either…
  • Listening to music on your run? Definitely no Apple Watch or Airpods
  • Want to Zoom your colleagues? Sorry again, but Zoom wasn’t even a company back then, let alone a verb!

Even the idea of jumping in an Uber with a stranger 10 years ago would have raised a lot of eyebrows. All of this is to say, ‘what you need to believe’ over a 10 year horizon can be very hard to fathom…

For most of TDM’s investment cases — and certainly for next generation businesses like Peloton, we challenge ourselves (as hard as it may be) to think about the vision 10 years out, as it helps to inform what ‘steady-state’ financial profile and cash flows might look like.

When thinking 10 years out for Peloton, we wanted to solve for the same metrics (number of connected fitness subscribers and % of household penetration) but had to push our thinking to consider more markets and more broadly, what technology adoption curves look like.

We identified 22 markets (large enough to move the dial) where we could see Peloton eventually being in if it executes well. We also realised that household penetration rates in each market could be vastly different due to country-specific factors, including:

  • Disparities in average household income
  • Cultural proclivity towards fitness and exercise (as proxied by % of the population who attend the gym); and
  • Average housing size (those Peloton bikes are quite bulky!).
Exhibit 2 shows a framework we crafted that groups each of the 22 countries into 5 different ‘classes’ based on propensity to adopt Connected Fitness.

Once we landed on our adoption framework, we back-solved what the required penetration rate for each of these ‘classes’ of markets would need to be, to be within the ballpark of the 10 year subscriber target of 25m.

We concluded that the 11 most attractive markets for Peloton (green box in Exhibit 2 ) would need to hit around ~8% of all households in 10 years, if PTON was to achieve 25m subscribers and thus a 20% plus investment return.

The inference that so follows, is that we have to believe Peloton is on its way to becoming a global, mass-market brand by ~2031 and not just appeal to affluent consumers.

While it is still early, Peloton is starting to demonstrate this in the US. Exhibit 3 (below) is an extract from the Peloton Analyst Day in September 2020 showing that Peloton has made significant strides towards appealing to <$100k annual income households over the past few years. This is very encouraging and suggests an 8%+ household penetration could be realistic.


Exhibit 3 A chart from Peloton Analyst Day (Sep 2020)

To achieve 25m subscribers, you’d also need Peloton to achieve penetration within the remaining 11 ‘lower propensity’ markets ranging from <5% to as low as 0.5% of households. Is there a probable world where Peloton’s flywheel leads to cost advantages through scale? A world in which the business then passes this back to consumers, and ultimately lowers upfront equipment prices to prioritise subscription growth? We have strong confidence that this is a playbook that management has front of mind.

Technology adoption


Lastly, as we went through this exercise as a team, we found ourselves asking what Connected Fitness penetration might hypothetically look like in the ‘end-state’.

This led to a fascinating debate around the pace of technological adoption more broadly. We uncovered data which indicates that new technologies are being adopted along their S-curves at a pace far quicker than any prior time in history. This is captured visually in Exhibit 4 below, which is taken from a fantastic research paper by Michael Mauboussin and Dan Callahan.


Exhibit 4


We are not claiming Connected Fitness is as comparable in importance or universality as some of the other technologies shown; but the underlying principle is clear — technologies are diffusing more quickly and growth typically ‘takes off’ in an exponential trajectory, making it hard for most human brains to process.

This suggests that while the required household penetration rates for Peloton may seem difficult to fathom at first, one could conclude that there is a reasonable probability of success if Connected Fitness goes mainstream and Peloton continues to innovate and execute.

An interesting analogy we’ve heard on Connected Fitness is one that links it to the disruption that came for Gaming 15–20 years ago.

At first, the advent of consoles such as the Nintendo, Xbox and PlayStation piqued the attention of mainly hardcore enthusiasts (and broader social skepticism). But sure enough, as with most technology S-curves, the majority of consumers eventually realised the comfort and convenience of gaming at home and migrated away from physical arcades. Fast forward 20 years, and you’d be hard-pressed to find many arcades still around.

While it’s not the perfect parallel, it begs the question if something similar could happen to fitness? We’ll leave that one for you to ponder…


Exhibit 5 A taxonomy of technology adopters by Everett Rogers (2003)



Thinking in terms of ‘what you need to believe’ and making inferences about the future is a critical part of how we think about investments.

By laying out snippets of how we approached Peloton’s 5 and 10 year opportunity, we hope to show some of the key principles of this mindset:

  • Focus on the big picture and implications of what your forecasts mean — in this case, what % of households can Peloton get into?
  • Triangulate these implications with verifiable data points today — For example, referencing current gym membership and wearables adoption data helps the brain to better evaluate the extent to which Connected Fitness could penetrate households over time.
  • Distil key inferences about the future and think in terms of probabilities  Doing the above exercise made it all the clearer to us how important traction in certain markets would be (eg., US, China, Germany) to the long-term Peloton story, and helps us narrow our focus on the probabilities of success in these markets.

While it’s not the only thing that goes into our investment decision, asking ‘what you need to believe?’ is a powerful way to break down the ambiguity of the future, especially when combined with quantitative analysis.

When it comes to investing, making inferences rather than predictions can be a useful distinction to remember.

This first appeared on our regularly updated Medium page – TDM Tidbits, where the team share their thoughts and experiences with the world. 

Read The full Article
What Makes Great Investor Relations
TDM Market Memo 2021
TDM is currently not open to new clients
Thank you! Your subscription has been confirmed. You'll hear from us soon.