Token valuation is really complicated. Especially for beginners and experts alike, everyone makes the same mistakes. Some protocols generate $500 million in fees annually, so they divide by market cap and conclude it's "undervalued." But wait. That calculation is likely wrong in both numerator and denominator.

The reason early Bitcoin buyers succeeded long-term is because they understood the intrinsic value. The same logic should apply to token valuation, but most frameworks overlook this.

The problem is that the revenue a protocol generates and the actual revenue holders receive are completely different. For example, a protocol might have $100 million in annual fees, but only a few million are distributed to holders. Ignoring this and comparing "EV/protocol revenue" leads to completely misguided conclusions.

In stock valuation, EV/EBITDA is used because it reflects the company's actual value. But applying this directly to tokens introduces three fundamental issues. First, treasury assets. Holders do not have legal claims to them. Second, protocol revenue. Most of it doesn't reach holders. Third, the true costs, which appear as new token issuance.

The correct approach is to focus on "EV/holder revenue." This indicates how much actual profit ends up in your pocket for every dollar you pay. It also considers the balance sheet and real business costs.

The difference is clear in real examples. For HYPE, the retention rate is 89.6%, with a cumulative ratio of 100%. That means over $80M out of $900 million in fees was distributed to holders. Meanwhile, Maple has a retention rate of 13% and a cumulative ratio of 25.1%. Even with the same fees, only about 3% of the revenue reaches holders, compared to 90%. Just changing the framework can quadruple the valuation.

The term "dilution" is also overused in the industry. Team incentives are operating costs and should be deducted from profits. Ecosystem incentives are the same. But investor lock-up releases and sales are market events, not operating costs. Failing to distinguish this leads to flawed analysis.

Treasury valuation is also crucial. It's not about "how much funds are there," but "can holders withdraw them?" PUMP holds about $700 million in stablecoins, but without governance mechanisms, holders can't withdraw at all. So, the withdrawable assets are zero. The company's value equals its market cap.

In SKY's case, 99.9% of its finances are in its own tokens. Applying a 50% discount drastically reduces the withdrawable value. The protocol revenue multiple appears as 7.3x, but the holder revenue multiple is 16x. The choice of denominator can change valuation dramatically.

This framework isn't perfect. The discount rate for treasury claims is subjective, and data sources can be noisy. But comparing only with "EV/protocol revenue" is far less practical for real judgment.

It's encouraging to see the industry evolving. Fee switches are turning on, reverse buybacks are replacing inflation staking, and governance layers are voting to stop incentives. We now need tools to measure what's actually happening more accurately.

Valuing based on what holders truly receive is, I believe, the perspective of long-term investors.