Why P/E Ratios Are Broken In Crypto

Level 2 - Value Investor

Welcome Avatar! We’ve loosely discussed this topic in the past but with the recent article by Arthur Hayes making its rounds, we felt it was important to quickly formalize our thoughts on this subject.

P/E multiples.

Meme source: the Internet

“P/E” stands for price-to-earnings, and it represents the ratio of the price of a token to the earnings of the underlying protocol. In stocks, P/E is the stock price to the earnings per share.

Why do people use P/E ratios in stocks? The main reason is for purposes of comparability.

Company A manufactures widgets and generates $100 million in earnings.

Company B manufactures widgets and generates $100 million in earnings.

Which one should you buy? No way to know.

Now let’s say you can buy Company A for $300 million and Company B for $500 million.

Now you know that Company A trades for 3x P/E while Company B trades for 5x P/E. Company A is “cheaper” than Company B. Of course, that doesn’t necessarily mean it’s undervalued and is a buy - maybe Company B is growing faster so the market values its earnings higher than Company B. But for purposes of understanding the concept if we assume all else is equal, Company A is cheaper.

By looking at the P/E multiple of each company we were able to establish a basis for comparison between these two companies, allowing us to make a better informed investment decision. You can also look at the “forward P/E” multiple, which is the current price to projected earnings. Forward P/E tells you how the market values a future year of earnings.

Even when it comes to stocks, P/E has a lot of limitations. For example, P/E ignores debt - what if Company A is trading at a 3x P/E because it has a huge loan that is maturing in a few months that it doesn’t have enough money to pay for?

It also ignores growth. People use the PEG ratio to account for this, which is equal to P/E divided by growth rate.

P/E multiples in a silo ignore the quality of the earnings. A company with stable revenue sources and high earnings visibility (lower risk) should be more valuable. When looking at P/E, and multiples in general, to compare companies you need to understand these differences at a minimum. Some companies are cheap for a reason. Some companies should trade at higher multiples because they are higher quality fast growing companies. Context is extremely important.

This brings us to crypto, where people are trying to make sense of entirely new types of business models. We’ll go over some pitfalls of using P/E ratios in DeFi.

Speculative Nature of Crypto

Paid subscribers will have heard us say this many times by now: DeFi protocols are highly reliant on speculative activity in crypto markets to earn cash flows.

And speculative activity booms and busts in spectacular fashion in crypto markets. That means a DeFi application can earn millions of dollars quickly and then end up earning nothing, all in a matter of months.

BowTiedNightOwl @BowTiedNightOwl

One of the (many) issues with using P/E to assess crypto protocols is that you're looking at cash flows that can go from $10 million in one month to $0 the next month but you're annualizing the last 30 days. Then, Arthur goes as far as to comp to P/E of traditional indices.

Remember that P/E in traditional markets is current price to annual earnings. In crypto, we see people take the last month or last 3 months and then annualize (multiply by 12 or 4) to arrive at annual earnings, then slap a P/E ratio on it for comparability purposes.

Here’s the thing: more likely than not, recent earnings are going to be a terrible measure for what one year’s earnings look like in something as volatile as DeFi. DeFi earnings are a function of value capture from speculators. While there will always be *some* speculation, the copy-paste nature of crypto means the speculation gets cannibalized by similar project and always eventually fizzles out.

If you’re hellbent on using P/E, use forward P/E and model out what earnings look like in a more realistic scenario. For example, it’s more realistic to assume that out of one year, maybe 4 months will have high speculative activity, 4 months will have moderate activity and the remaining 4 months will have low speculative activity. You can then take recent earnings and adjust your forecast up or down depending on where you are (e.g. we are currently in a period of low speculative activity).

Perhaps your view is that earnings have stabilized, or that we’re entering a multi-year bear market. Whichever it is - the important thing is that you should have a view beyond “annualize last 30 days earnings” if you want to build a strong thesis.

Tokenomics

Crypto tokens are not directly comparable to equity. They have their own unique economic structures (such as ongoing issuance which we’ll discuss later).

Equity is a simple construct in comparison - you own a piece of the company which entitles you to its cash flows. Your risk is the company underperforms expectations and/or you get diluted.

Crypto tokens are not so simple. People use crypto tokens to transact, vote on governance (equity-like), deposit to earn more tokens, and may also have utility with an application. For example, you don’t buy Curve’s CRV token as a play on underlying earnings, so why would you use P/E to measure its value? Instead, you might look at how much in liquidity you can incentivize with a certain amount of CRV and then make assumptions on how much users would want to buy.

The fact that tokens can be used for a lot more than cash flows makes comparing tokens on the basis of P/E (or any price based multiple) challenging.

The “E”

The “E” in P/E stands for earnings. Earnings are revenues minus expenses.

Protocol expenses can vary widely but the largest *cash* expense will likely be paying the team and other contributors.

The largest overall expense in most cases is token issuance. These tokens are usually spent on incentivizing some type of behavior, such as providing liquidity for a protocol’s tokens (the circularity here is not lost on us). This should be considered an expense. However, it is *non-cash* meaning the protocol does not come out of pocket to pay (tokens are minted out of thin air - super shadowy air).

Some people say you should subtract this non-cash expense to get to the most accurate P/E multiple, similar to stock based compensation.

We don’t agree, and here’s why.

What matters most in valuation and financial analysis as a whole at a fundamental level is cash flow. What is the intrinsic value of a company? The present value of future cash flows. While people do use P/E multiples in traditional valuation, P/E is inherently flawed (as briefly discussed earlier). Cash flows are a superior measure. If we had to choose between a partially accurate non-cash measure or a fully accurate cash measure that excludes tokens, we’d choose the latter.

The question then becomes, how do you account for the token issuance? It’s a real cost not only to the protocol but also to token holders who are getting diluted. That dilution is offset by the token holders’ participation in staking/liquidity mining and other token distribution programs, but it’s still a cost.

Our answer: ignore it for purposes of P/E. The reason is practical rather than technical - if you included token issuance in every P/E calculation you’d end up with too many unusable numbers (negative or extremely high P/E), making the metric useless.

Instead, look at P/E multiples and make a separate assessment of token distribution schedules (that’s what we do). If a protocol is reliant on high emissions, you should discount its value or even ignore it entirely if you anticipate the sell pressure from emissions to offset.

Alternatively, you can ignore P/E multiples entirely and focus on Price to Sales (“P/S”), which sidesteps the debate entirely (although is of course a worse proxy for cash flow). Crypto applications are high risk, growth oriented and rarely reach a point of stability. Whether you use P/S or P/E multiples at this stage of the industry’s life cycle is not going to make or break your analysis. Analyzing protocols thinking they only have a 2% profit margin because of token issuance will.

Note that if you are using P/S and ignoring expenses, you should have an idea of a protocol’s financial runway. Months of runway = Cash in Treasury / Monthly Cash Burn.

And that’s a wrap. P/E has major limitations in crypto. When using it, make sure to understand these limitations and make adjustments accordingly. Often times, that adjustment is ignoring P/E and focusing on other measures.

Also, since this a free post we have to say this: obviously there are nuances and some protocols abuse their tokens by issuing tons of tokens and internalizing all the earnings. But in general based on all the things we’ve dug into over the last 1+ year, this is the most intuitive approach for us.

Thanks for reading. If you’re reading this and haven’t subscribed, join our community below. We’ve written over a 100 articles in the last year that you can access and put out ~3 posts a week on everything DeFi.

👇👇👇👇

Disclaimer: None of this is to be deemed legal or financial advice of any kind. These are opinions from an anonymous group of cartoon animals with Wall Street and Software backgrounds.