As per the last 30D rate solana:pumpCmXqMfrsAkQ5r49WcJnRayYRqmXz6ae8H7H9Dfn buybacks annualise to roughly 18% of mcap vs 4% for hyperliquid:native. For the last 16 weeks it earns about as much protocol rev as HL, $65M against $68M over 30D, at roughly 1/10th of the mcap (~$2B against ~$21B) solana:pumpCmXqMfrsAkQ5r49WcJnRayYRqmXz6ae8H7H9Dfn looks cheap on current cash generation. h/t to @DefiLlama for the data

zcash:native holding up relatively strong through this drawdown. Custody scares will bring attention to sovereign pvt money and market seems to value that potential before here. More adoption soon as TAM is huge.

Samsung’s 2023 chatgpt restrictions showed how sensitive data can block adoption. > Roughly 71% of hacked value under infra key leaks > Cisco’s 2026 survey found 43% increased privacy spending in the prior year and 93% planned more investment @near_ai RED’s opening will be helping security teams use AI without creating another store of readable internal data and those losses don’t prove AI caused the hacks, but they sharpen the case for limiting exposure. This will open other department's budgets.

Added good size of $ZRO in the last 24H. After FTX, DEX volumes nearly 2xd in Nov 22 and I think this Binance hack will push venues to rethink custody and not just traders to switch apps. ATLAS lets them keep their customers while sharing trading infra, with 75% of fees after venue rebates earmarked for ZRO buybacks + burns.

This reminds me of Sept 24 when BTC fell ~10% while coins outside the top 10 gained ~5%. Alt leverage had already been flushed, so BTC selling didn’t trigger another wave of forced alt selling. Next 1 week will be crucial as TOTAL2 is still below its 3D trend band. If we reclaim $1.1T while BTC stalls then the stronger alts can run before BTC recovers.

Makes completely sense!!! > ATLAS lets trading apps keep their customers without building an exchange > It handles matching, clearing, settlement and risk underneath their own interface > The venue that brought the user gets a rebate of 20–65% > A bigger rebate requires staking more ethereum:0x6985884c4392d348587b19cb9eaaf157f13271cd > Of what's left, 25% goes to whoever created the market > The other 75% is used to buy ethereum:0x6985884c4392d348587b19cb9eaaf157f13271cd on the open market and burn it Could be a really good cook

So @HyperliquidX generated $74.4M in 30D revenue vs @Pumpfun' $58.6M, yet Pump leads all time at $1.46B vs $1.19B. These are 2 proven monetisation engines connecting. Assuming HL averages ~$6.9B daily volume, routing 5% through Pump at a 0.05% fee means ~$63M in annual gross builder fees, before costs. > For 1% it'd be ~$12.6M annually > For 10% it'd be ~$126M annually This is just more than distribution. h/t to @BlockworksData for the data

Yeah it's the first step as a protocol buying its token creates identifiable demand, but that alone does not prove it is the dominant marginal buyer, that purchases exceed selling pressure or that the token is attractively valued. Basically👇 Customer payments → protocol rev → actual token purchases → identifiable token destination There's a reason recently why we're seeing serious interest around businesses that turn customer spending into demand for their own token. $HYPE $PUMP $LIT $NEAR $STONK and $VVV are some good examples of it. Also $UNI $PONS $RAY $SYRUP also have rev funded purchase mechanisms worth tracking (but doesn't make it to the main list) Buybacks will be a norm soon which is why we need to find businesses where recurring rev can grow the buying budget before the market prices that improvement. Yes treasury purchase still creates demand, even without a burn but where those tokens go determines whether that demand is durable or temporary.

I really want @AskVenice to move from the AI destination people visit into an inference supplier embedded inside products they already use cuz that’s the bigger TAM for em. Basically serving devs who want to offer pvt AI without operating models themselves. As agents handle more personal context, privacy will shift from a preference to a purchasing requirement and my bet is devs would rather buy that capability through an API than build and maintain the infra. Won't be surprised if Venice serves thousands of applications without acquiring every end user directly an that’s why I’m interested in $VVV which is a route to recurring customer funded demand.