We don’t like Numerical Screeners

That’s probably not what you were expecting to read in the first sentence. But it’s true. We don’t like numerical stock screeners because they limit our worldview. They often guide us towards numerically “cheap” stocks, most of which turn out to be cheap for a good reason – “value traps”, in Wall Street parlance. A sort of Anchoring Fallacy takes place – where we try to justify an incredibly “cheap” price with an unlikely growth story.

Our favorite way of finding investment ideas is through observation, conversations and reading – out here in the real world. As we’ve said a few times over the years, “investing is not divorced from reality”. But we realize that this method is inefficient. Over time, we’ve tried to become more efficient.

We used to screen by simple, litmus test type metrics just to filter out bad apples. We used metrics like ROE, Debt/Equity and Free Cash Flow. This narrowed the field, but it was still too inefficient. Over time we realized that this method also screened out a lot of promising companies that didn’t meet our (maybe too) stringent numerical tests. Many of these promising companies didn’t (yet) generate free cash flow thereby making metrics like ROE useless. And many of them have believable pathways towards massive free cash flow generation. But we missed out on them. So, we had to adapt.

At The Buylyst, we invest in near-inevitabilities – technologies and trends that will change our civilization over the next few decades. We call them investment themes. That’s our starting point – because we don’t want to play the wrong sport. We like tailwind. It’s a very long-term game. That itself drives us towards many companies that don’t yet generate a lot of free cash flow, but many of them are either Global Dominators already or have very defensible moats. We’re happy investors in companies like Spotify, Netflix and Pinterest, for example. That happened because we adapted to make our Watch Lists more useable.

The big mindset shift came when we rediscovered a concept called Expectations Investing – a term coined by Michael Mauboussin. We had used its principles in a few of our investment theses – like Lyft – which worked out well for us. But this mindset shift – from seemingly safer “value” stocks to more growth-oriented stories – wasn’t yet part of our Watch List process. So, we incorporated it, and it guided us to some companies that actually ended up in our portfolio. So, we altered our Watch Lists considerably to fit in this growth mindset by answering the question, “what do we need to believe to buy this stock?” This is how our Watch List looked:

But recently, we did a deep dive on one of those companies that “screened well” based on this new criterion but revealed some shortcomings. A deep dive into Salesforce revealed that we could further improve our Watch Lists to make our (and your) lives easier. We did.

We’re calling it Watch List 3.0. That’s what this article is about.

Ultimately, we want our Watch Lists to:

1. Give us a picture of what’s potentially overpriced or underpriced in the markets, and within our themes.
2. Guide us towards the most believable stories so we can use our limited time in the best way possible.

We think we’ve finally achieved this. But we’re completely aware that this is a work in progress – all of investing is.

Watchlist 2.0: What do we need to believe?

We used to maintain a Watch List that mirrored our portfolio in the way it looked and felt. For a long time, our Watch List was a repository of companies we had already researched – complete with a detailed thesis and valuation. Our Watch List was essentially a repository of “comfortable companies at uncomfortable prices”. This approach serves us well for a while. Many companies from this list slowly found their way into “comfortable price” territory. When that happened, we pounced. In most cases, this proved to be profitable. One of our best investments since inception – Nvidia – was a result of maintaining this type of a Watch List. We watched Nvidia for a while. And when the price was comfortable – in late 2018 – we pounced.

Then in the summer of 2020, we added a new dimension to our Watch Lists. We incorporated the core tenet of Expectations Investing – back solving into what’s priced into the stock rather than spending days on a company to estimate that “uncomfortable price”. The goal of doing this was efficiency, so we could get better at using our time on the best possible investment ideas.

We gained from this enhancement. Some of our recent investments like Qualcomm screened through this process. We zoomed in on them because they appeared to be the most reasonably priced story within our AI & Big Data investment theme. But during the process of zooming into Qualcomm, we inadvertently made another enhancement.

Watchlist 2.5: How believable is it?

On top of our “this is the revenue growth we need to believe” estimate, we added a “how believable is this?” layer. We did this by incorporating a measure for “scalability”. Why? Because it’s a lot easier to believe future revenue growth when the business is easily scalable – meaning that it doesn’t cost much to produce and sell an extra product.

To account for Scalability, we simply calculated the proportion of Total Operating Costs that were Fixed Costs. Wall Street calls this Operating Leverage. As always, they put a fancy term on a simple concept. The idea is that revenue growth will outmatch cost growth if most of the costs are fixed. Software companies, in particular, enjoy this peculiarity. We love this feature because it means that more of the potential revenue growth would flow down to actual cash earnings – the main driver of stock prices over the long term.

We then combined the two factors – what we need to believe AND how scalable is this – to come up with a Priority Factor. We literally divided [Fixed Costs as a % of Total Operating Costs] by [Revenue Growth we’d need to believe]. The absolute number doesn’t matter. 0.5 or 1.2 doesn’t mean anything. But in relation to other Priority Factor numbers, it’s very useful. Here’s a comparison in our AI & Big Data universe.

We’ve been following this priority list systematically. We dug into Qualcomm, and then Salesforce. But when we took a close look at Salesforce, we realized something else. We need better measures of Believability and Scalability. Sure, we already estimate “how much revenue growth we need to believe”, followed by prioritization based on Scalability. But our work on Salesforce suggested that we need another layer to estimate “how believable is it?”.

The answer to this question is mostly subjective. It depends on the subjective things we usually think about in our detailed investment theses – durability of a company’s competitive advantage, management strategy etc. But before we get to that stage, it is possible to estimate a numerical indicator in that direction. This would be immensely useful. Again, one of the main goals of our Watch Lists is to ensure that we’re using our time in the best possible way.

So, we came up with cool, new, useful rating system.

Watchlist 3.0: The Investability Rating

The basic message of Watch List 2.5 was this: let us focus our time on companies for which the “revenue growth we need to believe to buy the stock” number is believable. We measured believability by Operating Leverage (we used fixed costs as a % of total operating costs) as our measure of operating leverage. Not too bad.

But when we dug into Salesforce, we noticed a few shortcomings of our 2.5 approach. Sometimes, revenue growth doesn’t trickle down to Free Cash Flow growth. Why? In Salesforce’s case, there were 2 reasons:

1. Fixed Cost growth kept up with revenue growth.
2. Too many new shares were issued to fund acquisitions, thereby diluting Free Cash Flow per share.

Usually, we don’t need to worry about these factors. But in many high growth companies, this is a regular feature. That’s fine. But the problem then is that assigning our standard 20X multiple to our estimate of future cash flow – to value the company – becomes harder to swallow. If free cash flow doesn’t grow much, we may as well assign that 20X multiple to CURRENT free cash flow. That usually implies a much lower valuation. No matter how impressive revenue growth is, it's hollow without most of it trickling down to actual cash earnings.

Revenue growth deserves a premium, when we’re reasonably confident that it will lead to “value accretion”. That means growth in Free Cash Flow per share. This is the extra layer we’d like to introduce in Watch List 3.0. Here’s how we’ll do it:

1. We’ll keep the “Revenue Growth we need to believe” number as is. This is our estimate of revenue growth that we need to believe to buy the stock at current price – IF Fixed Costs remain, well, fixed, AND IF there is no more shareholder dilution.
2. We’ll add a “Believability Rating” by dividing “Revenue Growth we need to believe” by “Current Annual Revenue Growth”.
3. We’ll keep the Operating Leverage number as is – Fixed Costs divided by Total Operating Costs. But let’s call this Scalability Potential.
4. We’ll introduce a measure to factor in Free Cash Flow per share growth. We’ll do this by dividing Free Cash Flow per share Growth by Revenue Growth – over the last 12 months. This gives us a sense of how much of those incremental dollars are trickling down from the top-line to the bottom-line. Let’s call this Scalability Evidence.

We’ll combine numbers 2 to 4 into something we’ll call the Buylyst Investability Rating.

Buylyst Investability Rating = ([Believability Rating] multiplied by [Scalability Potential]) multiplied by [Scalability Evidence]

The formula or the Investability number it spits out for a particular company isn’t that important. How this number compares to other companies in the peer group is very important.

In the new Watch Lists, we’ve grouped the Buylyst Investability Rating in the following buckets: 

1. High: Above 1.5X Investability Rating Median
2. Average: Between 0.5X and 1.5X Investability Rating Median
3. Speculative: Below 0.5X Investability Rating Median 

To put those Investability Rating buckets in proper context, we’ll also show its components – grouped in easy-to-read buckets:

1. Believability Factor: High (> 1.5X Median), Low (< 0.5X Median) and Average (between 0.5 and 1.5X Median)
2. Scalability Potential (Operating Leverage): High (> 1.25X Median), Low (< 0.75X Median) and Average (between 0.75 and 1.25X Median)
3. Scalability Evidence (FCF Leverage): High (> 1.5X Median), Low (< 0.5X Median) and Average (between 0.5 and 1.5X Median)

Focusing on High Investability Rating companies is the best way to prioritize our time. We like this measure a lot because it’s coherent with our investment style – investing in comfortable companies at comfortable prices. But it is a work in progress. Whether or not a company will increase fixed costs or will issue shares to dilute FCF per share is a subjective call. It’s not hidden in any numbers. Yes, we’ve taken the past 1 year into account with regard to these variables. But what happens in the future can only be guessed from earnings calls, management presentations, CEO interviews, back-channel checks etc.

We’ve done our best to incorporate whatever data we have to do the following:

1. Give us a picture of what’s potentially overpriced or underpriced within our themes.
2. Guide us towards the most believable stories so we can use our limited time in the best way possible.

No Watch List is perfect. But some are better than others. We strive to constantly improve ours.

Possible Shortcomings: Assumption is the mother of all f….ups.

Our Watch Lists have some in-built assumptions that could lead as astray:

1. We always pay 20X Sustainable Free Cash Flow for our investments. We’ve incorporated this assumption in our “revenue growth we need to believe calculation”. More details here. 
2. We assume that working capital swings over the past 12 months are representative of the future.
3. We assume that capital expenditure over the last 12 months is representative of the future. 
   1. To account for some inevitable increases in Capital Expenditure, we’ve kept Depreciation & Amortization intact. We haven’t backed them out, just to err on the side of conservatism.
4. We assume a 20% combined haircut for cash taxes and cash interest – that’s 20% of EBITDA – Capital Expenditure – Increase in Working Capital.

Now all these assumptions are always revisited when we do a deep-dive into any company. A company may not have any debt or may increase capital expenditures or may have had a one-time spike in working capital. These details about how management may allocate capital in the future are not hidden in historical numbers. They’re subjective assessments based on what management teams share with investors.

Watch List 3.0 is an idea farm and a guide to how we should spend our time. It’s not investment advice or a quantitative algorithm for allocating our money. It’s our eyes and ears on the ground.

In fact, the most immediate use for you is: A sanity check on your current investments or potential investment ideas. If the Investability Index is high, it should tell you “this is a reasonable bet IF YOU LIKE THE STORY”. If the Investability Index is low, it should tell you, “Hope you LOVE the story, because numbers suggest this is a highly SPECULATIVE BET”.

At The Buylyst, we like a High Investability Rating. We like believable stories before we invest – stories that we can quantify or guesstimate with reasonable confidence and margin of error. If we didn’t care about believability, we’d just be gamblers or YOLO traders.

We do this work because we play the long game. We like to sleep well at night while our money doubles every few years.

Watch List 3.0 is available in the Portfolio Page.

Many Happy Returns.