Learning from Yen Liow: Game Within the Game
Learning from Yen Liow: Game Within the Game
(My Notes from Yen's 2021 MicroCap Leadership Summit Presentation)
I recently watched a presentation with Aravt Global’s Yen Liow, where he spoke for ~30 minutes about the fund’s growth equity strategy, with a particular focus on micro and SMID cap zones. Aravt Global is a $500M fund based in Manhattan.
In this investment memo, I plan to share my takeaways so that I can come back to this and extract value from it for years to come. As such, the formatting is going to be more note-like than my typical writings.
In reading this piece, please assume that the majority of the takeaways are paraphrased snippets from Yen’s presentation. I will quote directly when appropriate, but I don’t want the reader (nor future me) to think these are my own insights. This is essentially my note-taking from this presentation and will be presented as so.
You can download this as a PDF below.
Section 1: Hunting for Horses - Understanding what exactly Aravt Global are looking for in their hunt for long-term compounders.
Section 2: Game Selection - The importance of deciding where to hunt.
Section 3: Exploiting Volatility - What makes volatility exploitable, and the relationship between business and price volatility.
Section 4: The Power of Transitions - Alpha explosions in transition case, and the emergence of transitioning monopolies.
Section 5: Concluding Remarks - My closing remarks.
Section 1: Hunting for Horses
Framing the preface to the discussion, Yen asked how many 100x or 10x investments do you need in your lifetime? The answer was, not many or none.
The land of 100x investments typically takes place at the proof of concept stage of the business cycle. Using the classic business cycle that Aswath Damodaran likes to use, I imagine this would be somewhere around stage 2.
Using the more concentrated business cycle that Yen shared in his presentation, he shows three distinct stages in the proof of concept, replication (high growth), and maturation and decline.
The 100x type investments tend to come from what Yen calls “venture type investing in public markets or turnaround strategies”. This is vastly different from a strategy that seeks 10x investments, which tends to be more aligned with Yen’s investment style. In this growth equity space, there are a handful of strategies that can work, and Aravt tends to focus on just three.
The two most critical questions that were asked in pursuit of finding these 10x investments were:
“Can you find hunt inefficiencies in a way that allows for repetitive exploitation? And can you survive that ride, enduring tremendous volatility whilst staying engaged for between 5 to 10 years?”
As such, the topics for discussion in this presentation centred around:
Actionable Volatility
Hunting Early Stage Horses
Surviving The Ride
Two frameworks to start with:
Pattern Recognition: How do you systematically break down the vast universe of investment opportunities to a finite set of ideas that have a chance of generating sustainably superior returns?
Focus: How do you maximise your return on time?
To solve that equation, Aravt looked to dig into the base rates of $2B+ companies listed in the US (using a study from 1990 to 2016) that have compounded at over 20% for 10Y and 5Y rolling cohorts.
The study would show that just 17 of the 90 companies studied (3% per cohort) were able to compound at >20% (a return of greater than 6.2x) over a decade.
On a shorter time frame (5Y rolling cohorts) they would find that 96 of the 645 companies studied (14% per cohort) were able to compound at >20% over that period of time (a return of greater than 2.5x).
After breaking down the data, Aravt identified three repeatable and scalable patterns/strategies that enabled this type of superior return.
Horses: Compounders that grow EPS more briskly than the broader market
Commodity Framework: A demand shock that hits a fixed supply curve
Quality Transitions: An industry structure or company mix is upgraded through M&A
“We launched with three [strategies], we believe only one matters in the 5Y to 10Y+ duration hunt for extreme performance”. The one that matters is the hunt for horses.
What is a horse, exactly?
“We focus on one strategy, we call them horses. We do it across all market cap zones. They are specifically monopolies and oligopolies that can compound earnings per share much more briskly than the market can for long periods of time and trade at what we believe are reasonable prices”
This plays into what Aravt call their ‘true north’; where they believe that over time intrinsic value and price converge, specifically EPS and FCF in more mature states, whilst sales and gross profits can drive that price in the interim. Without knowing exactly when that happens they believe the price will eventually converge.
In light of that, they assume that portfolio return will converge with the realised earnings power of their portfolio over time. The job at hand is therefore locating those companies which will generate sharp, durable, and persistent FCF, all whilst being prepared to pounce on actionable volatility when the opportunity arises.
Typically, Aravt is opposed to highly competitive markets, citing that competition is great for the consumer but less desirable for the capitalist. In an oligopolistic or monopoly-like market structure, the potential to buy into “unfair fights” is greater, and more desirable. Imagine taking Mike Tyson (in his prime) to a tenth-grade sparring session. The odds are certainly in your favour.
A key notion that Yen shared as to why they tend to favour these horses. An investor will spend hundreds, if not thousands, of hours on researching their positions. If those companies are not monopolies or oligopolies, or at least in highly advantaged business models, the link between the historical analysis you conduct and the projection of future earnings in order to arbitrage price will be very weak.
Taking advantage of the pricing inefficiencies that firms in monopolies and oligopolies can exert has allowed Aravt to exploit all the volatility and stay engaged for long periods of time during highly volatile periods of price activity.
Bringing back that visual of the business cycle for a moment, Yen remarks that Aravt focuses its business model on the replication phase. Businesses in this phase have won their niches and the relative size of the niche to the company is very large.
Across Micro to SMIDcap names, there is a strong mix of proof of concept stage and replication stage investments, with the replication phase in microcap being particularly exploitable. However, proof of concept investments in any market size cohort (micro, SMID, large) has a much wider dispersion of outcomes and is a harder call to get right.
Some of Aravt’s Key Findings from SMIDcap:
Very inefficient market.
Both tails are in play in terms of the distribution of outcomes.
A lot of businesses that are ill-formed or should not be listed publicly exist in this space.
There are a large number of proof of concept stage investments, but also a surprising number of monopoly and oligopoly exhibiting firms that they believe fit their ‘growth equity’ style of investing.
Surprisingly, Yen admits that the number of business trade-offs has been lower than anticipated in their hunt for oligopolies and monopolies in SMIDcap. Amongst the firms they are looking for, there has been low to no regulatory risk which, portrayed against a global backdrop, is important right now. They also found that small does not equate to immature, with many of these firms having long and fruitful histories and displaying the characteristics that they can compound for very long periods of time. And finally, they found a large number of quality owner-operators that are aligned with Aravt’s objectives.
Know Your Terrain in MicroCap Universe
1) Huge Investment Universe: The market for microcap companies is, by far, the largest in numerical terms for the number of businesses to sift through. So, a significant top-funnel, that has to be filtered for quality.
2) Massive Outcome Distribution: The widest of all deciles within all market cap zones.
3) Wild Return Skew: The widest set of alpha to be able to exploit exists within the microcap zone.
4) Bone Crunching Volatility: I think this one speaks for itself.
Section 2: Game Selection
The most important thing an investor can do at the outset is deciding where they hunt.
For Aravt, the objective is to compete in less efficient and less competitive markets, with the goal of extracting profits repetitively and comfortably.
Within efficient markets, any edge that a firm may have will quickly disappear as a result of the level of competition.
Inefficiency exists in both competitive and less competitive markets, but the key question is what drives that inefficacy?
Competitive → Inefficient = inefficiencies that are more temporal in nature, not structural. Being so widely covered, there are few structural inefficiencies, but there are always times when firms are “chronically mispriced”. Investors need to be set up to be able to exploit those temporal inefficiencies.
Less Competitive → Inefficient = inefficiencies that are more likely to be structural in nature when compared to efficient markets within the less competitive environment. “That’s where we want to be spending our time. For efficient zones within the microcap zone, there is an enormous universe of stocks which frankly are completely uncallable and require a very different industrial design if you choose to pursue them”.
Size of the Business Profoundly Impacts Skew
Below is a graph compiled from the same dataset of all US-listed companies from 1990 through 2016, split for annual and rolling decade distributions of what the outcome skew looks like for those stocks.
One a one year skew (left) it appears to be a normal distribution curve, but on the ten-year rolling cohort side (right), as investors start to hold things for a longer period of time, the number of ‘zeros’ (ie business failures) begins to appear in a more dramatic fashion, but you can also see the right tail being dragged out a little more in tandem.
The below table, shows a more granular presentation of that data, the deciles of each market zone for that sample over a 26 year period and their respective outcomes. Decile 1 represents microcap and Decile 10 represents mega-cap.
The biggest takeaways here are:
Skewness varies inversely with market cap. The smaller the market cap, the more extreme the skew (the difference between mean and median outcomes).
Microcap is rife with business and stock failure. The number of stocks in the microcap decile that generated any positive return over that 26 year period (in rolling ten-year cohorts) was 43%, the lowest of any decile.
The divergence between winners and the average is HUGE. Despite being rife with business failure, the mean/mean CAGR in the microcap space over that sample period (11.3%) is the highest of any decile. As such, the divergence is dramatic between winners and average outcomes.
It came as no surprise to hear Yen say that “microcap, in our view, is a stock pickers dream, but you better know what you are doing”.
To demonstrate how favourable the skew can be if the investor is highly skilled at finding rapid, durable compounders he shows a table based on a 20-stock portfolio where 35% of those companies are zeros, 35% earn a 9% CAGR (market return), and 30% earn a 30% CAGR over ten years.
Despite 35% of companies going to zero, this investor would still earn a 17% gross un-levered return (5x initial capital) over that time frame. This is assuming the investor hits their base case only one-third of the time.
Section 3: Exploiting Volatility
In the SMID/Microcap zones, both price and business volatility are high, presenting enhanced opportunity and risk.
Price Volatility: relates to the price of the security.
Business Volatility: relates to the underlying value of the company.
When both business volatility and price volatility are fluxing at the same time, arbitrage is extremely difficult. Instead, Aravt tends to focus on situations where the underlying strength of the business is stable, they have the ability to confidently project the intrinsic value and allow the price to come to them.
“In the business of arbitrage, our job is to exploit price versus our assessment of value”. One of our jobs is knowing the difference between opportunity and risk. Just because prices go up and down, it doesn't mean its exploitable volatility”
Exploitable volatility is most apparent in cases where the business predictability is HIGH whilst the level of price volatility is also HIGH. Here, the investor can confidently project/assess the intrinsic value and wait for the price to come to them.
In most cases, companies within the monopoly or oligopoly structure tend to be those which have a higher degree of business predictability. Thus, price volatility is your friend. When the business quality is high but the level of volatility is lower, the available alpha is a little smaller, given the investor is granted fewer shots on goal to size their positions.
In scenarios where the quality of the business or the predictability of that business is low and the volatility is exceptionally high, this is one of the toughest environments to profitability exploit repetitively. Avoid these environments when looking for long-term compounders.
High Prediction or Low Prediction Environments
This circles back to the game selection element of investing strategy. In order to have a basis for predicting values or outcomes, it matters if you are operating in a low or a high prediction environment. Using game selection, the investor can make volatility their friend.
Low Prediction Environment: A weak relationship between history and the future. Eg, flipping coins, roulette.
High Prediction Environment: A strong relationship between history and the future. Eg, subscription businesses, annuities.
Operating in a high prediction environment aligns with Aravt’s true north; where they believe that over time intrinsic value and price converge, specifically EPS and FCF.
Investors can survive the volatility, knowing that the business itself is growing stronger, but most importantly it allows for exploitation of that volatility.
In the chart on the right, the low prediction environment is when both price and business volatility are fluxing in tandem.
Section 4: The Power of Transitions
Earlier on I noted Aravt’s three repeatable and scalable patterns/strategies that assisted in earning a superior return. One of those was “quality transitions”.
Whilst transitions are not considered to be a source for long-term compounding, the greatest source of alpha explosions came from this strategy, which tends to be more trading-intensive. Yen cites that the greatest alpha explosions are typically within transitions where businesses shift from:
Imminent bankruptcy → Survival
Poor quality business → Less bad business
Entering a recession → Exiting a recession
Cyclical business → Less cyclical business
The issue with relying on transitions is those lower-quality businesses (the state of the business pre-transition) tend to have the widest outcome spans (less callable) and return dispersion. I.e, it can be easy to get your fingers blown off when you are holding dynamite.
Transitions tend to correlate with being in a low prediction environment.
So, this is less a question of whether or not there is alpha to exploit, but more so a question of how repeatable this strategy is when looking to scale and size.
What about the idea of Monopolies in transition?
All monopolies start somewhere, and Yen breaks this idea down into monopolies in the making and those which are already established and in the scaling phase. Framing this back to the business cycle, those which are “in the making” are more associated with the realm of the 100x, venture type investing, proof of concept phase, in public markets. Whilst monopolies that are already established as the winners and are looking to scale are more so in that replication phase in the land of the 10x.
• Monopolies in the making (100x): Rarely a single point of inflection. Wide dispersion until monopoly is set.
• Scaling a Monopoly (10X): One established that they are the winner, the question is how big can they get? There is a huge opportunity if they can execute, and that is a more answerable question that can be equally lucrative.
Of course, the approach you take depends on your skillset, price, and liquidity tolerance.
Section 5: Surviving Volatility
In order to survive the ride (volatility) an investor’s risk and sizing approach are critical.
Yen suggests there is no signal from the price when in the POC phase as the business volatility is still very high. As such, investors should average up, not down.
In the replication phase, investors are no longer predicting, but are now observing. The businesses have won, and price volatility (up or down) becomes actionable now that the business predictability is stronger.
Of the 100s of companies Aravt has studied over the $2B market cap threshold over rolling decades, the right tail distribution of outcomes looks like this:
15%+ drawdowns occur multiple times per year
20%+ drawdowns occur most years
50%+ drawdowns happen once or more each decade
“We cannot predict when they occur, but we are not surprised”.
Using Amazon as an example Yen points out that over the last decade Amazon has:
Generated a 30% CAGR
Spent 34% of the time at least 10% below ATHs
Spent 12% of the time at least 20% below ATHs
Faced a max drawdown of 34%
But taking a look at the period of time between when Amazon was still a proof of concept business (left) and when it became a recognised monopoly (right) draws some interesting insights.
During the POC stage, Amazon was a monopoly in the making. The monopoly was not set, and the predictability of the business was low.
During the replication phase, there was a whole slew of things that were now OBSERVABLE as opposed to what you would have had to predict such as:
Quality of the moat
Size of the TAM
Unit Economics
Quality of the management
The ability to scale
All of that volatility from the POC phase and even the GFC (shown in red) looks immaterial when shown on a lifetime return basis.
To survive volatility, Yen shares four critical pieces of advice:
1) Be aware of the base rate: The cost and the provider of high performance.
2) Game selection: Choose a game that gives you a chance to survive and/or exploit that volatility.
3) Marking Risk vs PCL: There is a tremendous difference between marking risk and permanent capital loss. Just because price action varies, it does not make it actionable.
4) Hold through vs Add down: Protocol when heavily sized is to hold through instead of adding down. The first imperative in a highly volatile environment is to ride through it.
Section 5: Concluding Remarks
That wraps things up. I have compiled my notes here in Substack so that I can come back and revisit this document for years to come and also for readers to get an itemised presentation of the key points discussed. I hope it brings someone some value.
Closing out his presentation, Yen shares his fourth source of strength during volatility.
1) Trust the quality of the team: In Yen’s case he has a solid team behind him, but for the individual investor, this translates to trust your capabilities.
2) Trust the quality of the business: If you are comfortable that you are in a situation where business volatility is low but price volatility is high, you should be able to sleep better at night.
3) Trust the quality of the management team: Find great businesses managed by great teams. Focus on the business, and let the managers run that business.
4) Valuation: This tells you how much you might make, but not anything about protection.
For those interested in listening to the full presentation, as well as the 20 minute Q&A that takes place ~28 minutes in, I highly recommend watching the below video.
Conor,
Author of Investment Talk
You can also reach out to me here:
Twitter: @Investmenttalkk/@TheITNewsletter
Commonstock: Investment Talk
Pinterest: @InvestmentTalkkk
Email: Investmenttalkk@gmail.com
Really enjoyed this one, thanks for sharing Conor