
DCF is totally broken for most L1s
It's always been broken and will remain so for a long time. Maybe decades. Using DCF to "value" L1s is not correct
Here's a full explanation of why this is true -
DCF stands for "discounted cash flow [analysis]". It means you are trying to calculate a floor value for your asset based on the total value of all future cash flows that will accrue to that asset.
Any DCF valuation for any asset/stock/real estate/etc is a floor valuation, ie. a minimum and not maximum valuation.
A DCF valuation might be substantially less than the market value for the asset - maybe because investors expect the cash flow to grow, or because the asset has value beyond its cash flow.
For DCF to work mathematically, the relationship between the asset being valued and its cash flow must follow all of three rules:
1. The cash flow must exclusively belong to the asset. This is the "value accrual" part.
For stocks or bonds, this comes from the legal right the asset holders have over the payments/retained earnings /dividends.
For L1 REV, cash flow value accrual comes from the fact that REV is either burnt (and the thing burnt is the L1 token itself) or the REV is directly paid out to validators, and any L1 token holder may readily become a validator, and only L1 token holders may become validators, so validators are just L1 token holders. L1s pass this DCF rule of value accrual, but fail the next two:
2. The cash flow must be the independent variable
The asset's floor valuation must be a function of the cash flow, and not the other way around where the cash flow is in whole or part a function of the asset's current market price.
A bond (which is a loan that might have interest payments and then principal payback is due at the end of the fixed loan term) usually has an asset market price that is very close to the DCF floor valuation because a bond doesn't have much going on besides its cash flow payments.
A bond is not trying to develop breakthrough technology. A bond is not a meme stock that sustains a level of investor confidence or enthusiasm far beyond its value of debt repayments.
So a bond is on the edge of the "DCF spectrum" - it's nearly a pure DCF asset.
A share of McDonald's has a DCF floor valuation based on the cash flow produced by people buying burgers.
Like all equities, McDonald's has a "P/E ratio" (price-to-earnings ratio) that basically measures how confident and enthusiastic investors are for McDonald's shares above and beyond the DCF floor valuation.
Tesla is an example of a stock with a very high P/E ratio where the TSLA asset market price is way beyond its current cash flow floor valuation. McDonald's P/E ratio has been relatively stable across many decades, reflecting a fairly direct relationship between the cash flow produced by McDonald's and the market price of a share of McDonald's.
This puts McDonald's in the middle of the DCF spectrum - it's mostly a DCF asset; not as much as a bond, but the market price is substantially derived from the cash flow.
Now, not only is the share price of McDonald's mostly derived from its cash flow, there is another crucial element that gets to the heart of this DCF rule that cash flow must be the independent variable:
Changes in the McDonalds share price have very little if any direct effect on the McDonalds cash flow. The relationship only works one way (cash flow feeds into share price; not vice versa). This is key to DCF.
Imagine if somebody bought half of all McDonalds shares and, in doing so, pumped the share price up 2x in one month. Would the cash flow also 2x? Would people tend to buy 2x the amount of burgers after the share price 2x'd?
There is a big difference between an investor having a thesis that McDonald's is about to experience a fundamental catalyst that will 2x burger purchases, vs. the burger purchases going 2x because of a big share purchase.
A large McDonalds stock price increase might make the news and drive a bit of extra foot traffic to the stores, but is the cash flow going to 2x because the share price 2x'd? No, definitely not. Because for McDonald's and most other reasonable equities, the cash flow is the independent variable that's upstream of the share price - the tail does not wag the dog.
However, today's L1s and their cash flow work totally different - the tail does wag the dog. L1 cash flow is not the independent variable.
When an L1 native token price goes 2x in one month, it pretty much always in history has coincided with rising L1 user exuberance. This is not just historical data, it also lines up with the theory and observation of how a rising L1 token price excites people to buy/trade high beta assets on the L1.
It's not just coincidence/correlation that L1 speculation follows L1 token price. It's also because a rising L1 token price creates/causes net new L1 speculative activity, due to reflexivity.
This means that changes in speculative activity on the L1 (and thus also L1 REV because speculation mechanically drives congestion pricing and MEV extraction) is both correlated with and is caused by changes in the L1 token price. So, L1 cash flow is not anywhere close to being the independent variable and totally fails this DCF rule.
Eventually, in distant future eras of crypto, some L1's USD-denominated cash flow may become almost entirely independent from its L1 token price. Maybe this happens because the majority of fees end up being paid in relatively fixed USD amounts by institutional customers for non-speculative purposes. Or perhaps because L1 speculative activity ends up becoming mechanically decoupled from L1 REV, i.e. congestion fees somehow don't go up during manias and MEV is almost entirely solved so speculation isn't taxed. This is science fiction by today's standards.
So, in a hypothetical far off future, the L1s of the that era might pass this DCF rule for cash flow to be the independent variable. But today's don't - it's not even close.
3. The third rule of DCF is that the cash flows must be relatively stable and predictable over long periods of time.
This is the where the "D" in "DCF" comes in, it stands for "discounted". DCF expects cash flow payments to continue ~forever (or maybe for ~10 years of a bond repayment) and then you "discount" (an accounting term) the literal flow of future cash payments by the prevailing interest rate (such as the Fed funds rate) to work out the present immediate cash value for the future cash payments (because $1000 received today is worth more than $1000 received in 5 years).
McDonald's has a pretty good idea of how many burgers Americans will buy in 1 year, 5 years, 10 years. Amazon has a pretty good idea how much stuff people will buy this Black Friday and Black Friday in 2029.
But you and I (and your favorite analyst) have very little idea what a major L1's REV might be in 3 months, 12 months, or 36 months.
This is because L1 REV today comes primarily from two sources. The first source is congestion fees under relatively fixed blockspace supply which can cause major swings in total fees if demand or supply changes in either direction.
The second source is overly extractive MEV that's primarily driven by short-term speculative token swaps and can explode/collapse if MEV protections change and/or speculative activity levels on the L1 change, eg. due to a market crash or moving elsewhere, such as to another chain. Note that in the app layer, total app fees (GDP) can also spontaneously crash for similar reasons, including if intensifying competition reduces extractive protocol take rates (eg. high app swap fees are unsustainable).
Another reason that L1 REV is unpredictable is that the L1 competitive landscape is highly dynamic. A chain might be one outage or competing chain or product launch away from a dramatic permanent reduction in REV.
So, today's L1 cash "flow" is not really a flow at all. It's more like a cash dump truck that shows up every day. You never know how much is going to be in it. The amount of cash could rise or fall by an order of magnitude with little notice, and then change again next quarter.
In the distant future, if an L1 manages to have independent cash flow that's also forecastable over long periods of time, then it can pass the DCF rule for the cash flow to be reasonably stable. But most L1s don't pass this today.
In short, for DCF to work mathematically, three rules must be followed:
1. the cash flow must accrue directly to the asset (works for most L1 tokens)
2. the cash flow must be the independent variable and not have a change in cash flow caused by or correlated with changes in the asset's market price (totally broken for L1s this era)
3. the cash flow must be predictable over long periods of time (mostly broken for L1s this era)
Someday we might be able to use DCF to create floor/minimum valuations for L1s, but today we can't.
We can say "my favorite L1 made $20M in fees yesterday and that accrued to the L1 token" because the cash dump truck did indeed show up yesterday with $20M.
But we can't say "My favorite L1 token is worth $___ because look at this amazing historical cash flow chart" because for today's L1s, past cash flows are not sufficiently independent and stable predictors of future cash flows.

great question from @alexgedevani - especially in this age when we hear a lot of 'oh I'm totally going to run this weekend' and 'bet'
i've always liked the idea of an accountability app - and would like to use this tweet to explore what could make it successfull
the example we will use here is an exercise accountability app where users put up $ against their friends / pvp with other users on the app
first let's start off with the problem, why aren't people actually using the accountability apps
1. fear of commitment
• many users know they need to build that habit, but know that it's difficult
• as a result they choose not to embark on it entirely
2. lack of access
• most habit apps are still focused purely on inculcating the habit
• there's no monetary gamified element around it
so what might increase adoption?
1. increasing ease of access
• embed widgets in upstream / downstream channels of habit creation
• examples include: check-out pages for sports equipment, supplements (e.g. protein powder), sign-up pages for marathon, ClassPass etc.
• signing up for marathon: put a stake down to stick to a running plan
• check-out page for supplements: put a stake down to stick to a diet
• the platform can also crowdsource things like marathon training plans, diet plans, to make it even more compelling and reduce the activation energy for users to participate
2. gamification / socialification
• @alexgedevani had a brilliant idea - bet on points accumulated (points can be used to redeem items)
• instead of having points expire by date, points expire based on whether users actually stayed accountable
• stompers: exercise accountability apps where users can buy 'items' to slow each other down - great way to monetize
3. simplification
• users need to be able to put down a stake easily
• how? pre-set templates
• the app shouldn't be getting users to fill up a bunch of things as thought they are customizing a burger order
why crypto
1. yield: while you lock up your capital for the bets, optimize it for short-term yields
2. high programmable: what do you want to be accountable on, smart contracts significaintly expand the potential markets users can tap into
what's cool
• data flywheel is imminent with such a product
• what products are users buying? are they sticking to their original plan? what can be upselled to them if they don't stick to their plans?