Thread by @SolBeachBum

Bullish on #Solana’s tech?

But bearish on $Sol the asset?

Myself and @StrategicHash have built a 🎉Solana Economic Model🎉 to put these theories to the test.

TLDR?

Sol can become one of the most expensive assets in the world…

A thread

I have previously written about the ways to think about valuing a Layer 1 Asset in the below thread

https://twitter.com/SolBeachBum/status/1659977808880230406?s=20

Ultimately, Sol’s price accrual will come from:

👉Users buying Sol to pay for tx fees

👉Investors buying Sol to stake and earn yield

👉Hoarders storing value in Sol as a SOV

For the purposes of our valuation, we have assumed revenue growth to 2035, and applied a 33x multiple

🎉TPS🎉

User tps will be a major driver therefore of Sol's valuation

We have assumed this will hit 200k TPS in 2035

This may seem high, but Sol is increasingly becoming the home for high-touch use cases, and take on-chain gaming as an example:

The top online games have 100m daily active users.

If a fully on-chain blockchain based game achieved the same numbers, and every player performed 100 total transactions a day

- that would equate to 100k tps – for just one game.

Now consider all the bot-driven activity that Sol's low fees allow

In tradfi markets, around 25k trades are executed every second

But HFT market makers place / cancel orders at a ratio of 20:1

So if global trade moved onchain, then CLOB TPS could total 500k tps +

If one believes that total on-chain transactions will never reach 200k tps - this makes the case for Solana even more compelling

Since it would be able to capture all transactions on its single shard

And provide improved UX, reduced finality and cheaper fees vs Eth's L2 layer

🎉Fees🎉

Sol's current fee model MUST change.

Here is how they currently work:

All transactions incur a standard base fee, set in Sol, at 0.000005 Sol

This equates to 1/100th of a penny, with Sol at $20

So what's the issue?

Validators pay for votes on chain. These are very computationally light txs - at around 2,000 Compute Units (CUs)

At the same time, complicated MEV bots can post txs which take up 500k CUs

Each block currently has a maximum of 48m CUs

So a couple of things

1. Despite there being a limited resource (48m CU block limits), light txs are charged the same as heavy txs

2. Validators profitability takes a massive hit, as they have to pay out huge sums in tx fees.

Toly has visions of increasing the validator set to ~20k

But these fees mean that Solana can only support ~2000 validators.

It needs to make changes, so as to help grow the validator set going forward.

See below validator avg 'profitability' if increasing to 20k over time:

👉We therefore believe that base fees need to be set upon CUs used by each tx

👉However, we also believe that the base fees need to be able to flex up or down dependent on demand

If this second point does not happen, then Sol's price can not accrue SOV premium

This is because base fees are priced in Sol

If Sol's price were to 100x from $20->$2k

The base fee would also increase from 1/100th cent to $0.01

Another 10x, fees are at $0.10, thereby ruling out most use cases + killing revenues

Sol needs to be able to ⬆️ w/o fees following

In 2035 - we believe base fees will need to REDUCE from current levels - to average 0.000002 Sol

With validator fees only costing c. 1/10th of that amount

And computationally heavy MEV bots etc costing a lot more

To these assumptions we have also allowed for priority fees on 15% of all transactions

And MEV at 0.1% of the assumed TVL ($1tn) - accounting for $1bn of revenue pa.

With these core revenue assumptions in place, the results are as follows:

Inflation

Forecast to be ~0% by 2032

👉This is driven by the reduction in inflation rate to a terminal rate of 1.5% in 2031

👉And increase in fees received. 50% are burnt and reduce the total supply

If avg base fees increase from our assumed levels, this goes deflationary

Total Revenues

Assumed to hit $43bn pa in 2035, made up largely by the base fee from 200k tps of user transactions

The received staking yield will be c. 5% in 2035

However 1.5% of this is inflation rewards which are offset by 1.5% of inflation dilution (nets to 0)

So real staking yield (net of inflation) = 3.6%

NB this assumes the current staking rate of 84% of the total Sol supply remains

Worth noting that Eth's current staking yields are based on a staked rate of only ~20%.

If more Eth is staked, that yield will go down.

Conversely is less Sol is staked on Solana, the yield will go up

Validator Profitability

If our proposed fee changes are put through, then we estimate that 20k validators could be profitably maintained (with scope to increase this number further)

Each would earn c. $60k of profit on average

P/E Ratio

Borrowing VanEck's PE ratio for Eth, they applied a ratio of 33x

This is effectively the difference between annual revenue and Valuation

If Solana produced revenues of $43bn pa

👉This would make its FDV = $1.4tn

👉With Sol price of $2,000

tinyurl.com/mr48tnse

So what does this Sol price do to avg base fees?

We have used c. 0.000002 Sol - which equates to ~$0.003 per tx

@jon_charb's recent Eth val assumes Eth L1 fees at $0.20 per L2 tx

On top of this, one needs to also add L2 fees

So Sol fees ~100x cheaper

tinyurl.com/2f563ruk

So in conclusion - if Sol can continue on its journey of accruing low value, high tps use cases

Based largely off just the base fee income alone:

👉Inflation ~0% by 2032
👉Revenues of $43bn pa
👉FDV of $1.4tn + Sol price of $2k
👉Profitably supports 20k validator set

If you enjoyed this thread -

- Pls give myself and @StrategicHash a follow
- Like and RT the first tweet
- Listen to @UnlayeredPod

You can read more on our findings here - medium.com/@LightspeedRes…

Financial Model is here - docs.google.com/spreadsheets/d…

@SolanaLegend @LoganJastremski @zen_llama @SOLBigBrain @KyleSamani @TusharJain_ @aeyakovenko @rajgokal @R89Capital @fd_ripatel @daliallamar @solanobahn @Dragonmansol @Ravers @durdenwannabe @KEMOS4BE @lui1of1 @laine_sa_ @dilettantejr @axelverse_m @0xArunk @dubbel06 @buffalu__