Jake Koch-Gallup

this bull market feels like one where value investors actually get rewarded.

i'm the type of investor who finds a project i like and holds it for years. in past cycles, fundamentals and revenue were a meme.

that's changed with regulation passing and more projects producing real, sustainable revenue. holding high-quality projects should finally pay off.

so i built a fundamental value investing dashboard.

it ranks tokens mainly on tokenholder value accrual, then on product traction, team and governance, valuation, and dilution/supply.

each token gets bear/base/bull fair values, comps, a catalyst and unlock tracker, and technical analysis. it all rolls up into a quality vs value matrix.

of the 23 tokens in the first run, only $BTC and $HYPE make the buy tier on a 6-month time horizon.

it was built 100% on the @blockworks MCP: price, supply, unlocks, revenue, research, etc.

let me know if you'd want a live, interactive version, and which tokens i should add.

this is my last week at @Blockworks.

after 2 years at Messari and another 6 months at Blockworks following the acquisition, i'm really grateful for the people i've worked with, everything i've learned, and the research i've published.

it's sad to see one of the leading crypto research firms shift its focus away from research. i understand the decision to prioritize product, but i think independent research matters more than ever, especially as AI makes original human thinking and analysis increasingly valuable.

for now, i'm taking a little time to reset and finish up my personal crypto asset valuation framework, dashboard, and whitepaper, which should be launching in the next week or so.

after that, i'll be looking for new opportunities in crypto research, investing, and beyond. excited for what's next.

great report on $PEARL from @minnus & @0xMetaLight.

they're net bullish, but the report backs up a lot of my bear case. some key points:

1) the "useful work" isn't happening yet. fewer than 1% of blocks match a real AI model, and gaming cards run ~82% of hashrate.

2) AI servers have no reason to mine (at today's price). an AI server gives up $1.33-$3.15 of paid AI work for each coin it mines, while a gaming card pays $0.14-$0.23 in power.

3) the only real-world example didn't last. together AI's pearl-powered endpoint launched in may at a 25%+ discount, and was deprecated a couple months later.

4) the useful mining fix has no timeline. the FP8 upgrade would make mining on an AI server about as cheap as mining on a gaming card, but there's no published date, and the draft only works on nvidia's newest chips. so older gaming cards and every AMD card could get shut out, and gaming cards run most of the network today.

5) miners sell fast. about half of pool payouts go straight to one custodian's deposit addresses, and only ~10% is still sitting in miners' wallets a week later.

6) miners who already own their gaming GPUs only pay for power, so they'd stay profitable even if $PEARL fell ~80%. which means their costs give the price little support.

7) the network is concentrated. ~90% of blocks go through pools, kryptex alone mines ~50%, and the top 1% of addresses hold ~71% of supply.

8) transparency is still missing. ~125M $PEARL was issued in launch week, 98% of it has never moved, and the lab still hasn't disclosed its holdings.

the bull case depends on FP8 shipping and AI providers actually mining. until then, holders will have to absorb ~84% supply growth, about $540M of new $PEARL, over the next year.

got some good pushback on my $PEARL fundraise post. here's my response to the main points, and how i think the team could fix them.

1) "this fundraise has been known for months."

no it hasn't. it's been a rumor, and an unverified one.

one of israel's largest business papers, @TheMarker, published the exact details yesterday ($55M in equity and a $75M GPU fund). that's the first time we’ve gotten actual numbers on it, and the team still hasn't confirmed it.

this goes back to transparency. why isn't the team announcing or confirming this? why do holders have to find out through rumors and business papers instead of directly from the team?

2) "pearl's whitepaper clearly states that difficulty was lowered for the first 10% of supply to fund operations."

the issue is again transparency.

that's 210M $PEARL, ~59% of today's circulating supply, or ~$380M at today's price.

we don't know how much of that the team mined, how much they still hold, or how much they've sold.

one reply says the team mined less than 5% of the network. maybe, but the team hasn't published anything that lets holders check that.

3) "they raised capital to build a GPU farm to add hashrate to their own network!"

yes, more hashrate is better. i agree.

but if that fund mines $PEARL (which it almost certainly will), it centralizes the network around the team even more than it already is.

the team ships the code and the upgrades. $75M buys roughly 2,500-3,000 H100s, which could make the lab one of the largest miners on pearl.

as someone put it, "whoever controls the majority of the network's hashrate ultimately decides which chain gets secured..."

so we're comfortable with the dev controlling a growing chunk of the hashrate, without disclosing how much? that takes away a lot of the credible neutrality and decentralization that make blockchains, and especially SoV assets, valuable.

4) "trying to accrue value to the token would invite additional valuation frameworks, which can only result in a lower valuation ceiling."

i disagree. $PEARL is a lot less credibly neutral than other PoW assets like bitcoin, zcash, or even monero. it's also missing what monetary assets need:

> low inflation (supply grows ~85% this year)
> deep liquidity (~$8M of daily volume)
> transparency from the team (undisclosed holdings + fundraise)
> years of evidence that it works (it's 6 months old)

so right now, the SoV case is weak, and a second source of value would help, not hurt.

look at ethereum. it's valued as an infrastructure token (it returns value to holders thru staking and burns) and as a monetary asset (it's the main collateral across DeFi, it's the gas currency for ethereum and most of its L2s, and ~$18B of it sits in spot ETFs).

a claim on value accrual gives a token fundamental support. monetary premium is the most fragile part of a token's price. when zcash's orchard bug was disclosed in may, $ZEC fell ~50% in two days even though nothing about its fundamentals changed.

5) “there is no equity/token split. pearl research labs is an equity business AI lab, completely separate from the token."

this isn't accurate.

they are not completely separate. the lab ships the code, the early low-difficulty window was set up to fund its operations, and its GPU fund will likely mine $PEARL.

and the part that is separate is the part that makes money. that's what an equity/token split is. here is the test i use:

> does the lab make money without the token? yes, it sells inference in USD.
> does the token share in that revenue? no.
> are any buybacks programmatic? no.
> could the lab keep growing while $PEARL falls? yes.

so the business and the token have to be valued separately.

if $PEARL went to zero tomorrow, the lab would lose its mining subsidy, but it could keep selling inference on its own GPUs.

the counter is that the team holds $PEARL, so their incentives are aligned with holders. but that only goes so far.

the team also owns equity in the lab, and so do VCs who may not hold any $PEARL. when the token and the business want different things, like whether inference gets sold for USD or $PEARL, the lab will answer to its equity holders. and we still don't know how much $PEARL the team holds, whether any of it is locked, or how much they've sold.

there are ways to fix this though:

1) TRANSPARENCY. disclose how much $PEARL the team and investors mined and hold, how much they've sold, who invested, and what share of the hashrate the lab controls. for a SoV asset, this is the most important thing a team can provide. it's why @blockworks has a token transparency framework.

2) FIX TOKENOMICS. accept $PEARL for inference, and burn the $PEARL that comes in instead of selling it. that adds demand for the token and helps offset miner selling. (i know. i know. "this is a SoV token, it doesn't need utility bro!!" but as i outlined above, it can/should have more than one source of value).

3) USE THE REVENUE FOR THE TOKEN, NOT THE TEAM'S AND VCs' POCKETS. most or all of the lab's inference profit should go to buying back and burning $PEARL. make it a fixed policy with burns sent to a published onchain address, not a verbal promise or something left to discretion.

4) CAP THE LAB'S HASHRATE. commit to keeping the lab's own mining under a set share of the network, and report it publicly.

5) TIME. supply growth drops to ~33% next year, ~18% the year after, and ~12% after that. by then the team could prove it's a credibly neutral dev, and $PEARL could have a real case as a SoV. bitcoin is 17 years old. zcash is 10. $PEARL is 6 months.

if the team addresses the first four, i'll revisit my bearish view.