倔犟的小土豆

While reading the Babylon documentation, I saw the section about the Covenant Committee, and my first reaction was, “Oh, another committee.” To be honest, whenever an on-chain project sets up a multisig committee, I automatically switch on my defensive mode—this stuff is too easy to be packaged as “decentralized,” but in reality it can become a backdoor. Later, I went through the entire staking and exit flow, and found that I might not fully understand where the bottleneck really is. @BabylonLabs_io

In normal cases, when the term expires and the redemption happens, there’s no need for committee signatures—the timelock is on your side, and you can just go. The committee is stuck on two things: early unbonding and malicious behavior slashing. That means only non-standard paths go through them. For a normal BTC redemption, the committee doesn’t touch it at all.

So why design it this way? Because Bitcoin scripts are too “dumb.” If you lock some BTC and want to leave early, the Bitcoin mainnet doesn’t understand the concept of “early.” It only knows it can unlock when the time comes. What the committee does is actually quite straightforward—adding an authorized path for “early exit” using multisig. It’s not permission itself; it’s a conditional executor. Of course, there are risks here. If the committee’s signature mechanism were compromised, in theory they could sign a fake unbonding transaction. The documentation says it’s a 3/5 multisig with members that are publicly known. But whether this threshold is sufficient and how members can be replaced—there’s currently no running data to verify.

So the Covenant Committee isn’t a backdoor, but don’t treat it as a perfectly decentralized design either. It’s more like an engineering compromise stemming from the fact that Bitcoin scripts are too clumsy—it works, but it’s not elegant enough.

Personally, I think the truly important things to look at are: (1) whether the permission boundary will expand with upgrades, (2) whether the member composition will become increasingly concentrated, and (3) whether this mechanism can be replaced once Bitcoin script capabilities improve. Until those three questions are answered, saying it’s “fully decentralized” or calling it a “centralized backdoor” is too early. #baby $BABY

Hello everyone, I’m your old-stock-trading friend. In the past couple of days, the US stock market has been absolutely exhilarating. Today, I have to chat with you about the overall market and the SanDisk (SNDK) that I’ve been watching recently.

First, let’s look at the market: On August 4, the Dow closed at 54,085 points, up 1.71%. The S&P 500 hit a new all-time high, and the Nasdaq surged even more—up 2.59% to 26,584 points. With geopolitical tensions easing + oil prices falling + AI earnings beating expectations, three major positives directly pushed market sentiment to the max, and money flooded into tech stocks.

Now let’s focus on SanDisk! This stock was basically a roller coaster last night. It opened down 7%, with a low of $1,125. I almost thought I would have to cut my loss. But then it completely reversed in the afternoon—closing at $1,427.62, up 10.84%. The intraday swing was over 18%, and the shorts were crushed.

Why is it so strong? The core driver is the explosive growth in AI storage demand. SanDisk’s Q3 data center business increased 645% year over year, and its gross margin soared to 78.4%. Goldman Sachs even raised its price target from $1,200 to $2,200—implying a bullish upside of nearly 60%. Market consensus now is that AI training needs massive storage. As the NAND flash leader, SanDisk has orders booked out through 2028 and doesn’t seem to have to worry.

One quick reminder: this stock is extremely volatile. Over the past three months, it’s still down 8.62%. If you’re thinking of jumping in, don’t chase—waiting for a pullback may be safer. The market may be strong, but the Fed’s policy is not set yet, so position sizing and risk control are crucial. What do you think—can SanDisk push up to $2,000? #TradFi晒单

Bitcoin’s mainnet script language has always been “dumb” and can’t handle complex computations. That’s both its security foundation and its functional ceiling.

Babylon’s Trustless Bitcoin Vaults (TBV) aims to resolve a key contradiction: how can the Bitcoin mainnet verify a complex off-chain operation without changing a single line of code in the mainnet?

The answer lies in a “garbled circuit.”

In simple terms, your withdrawal logic is packaged into a garbled circuit via the BitVM3 framework. The circuit outputs a result—whether the claim is “true” or “false.” The mainnet only needs to verify this final outcome; as for how many conditional checks occurred in between, which branches were triggered, and so on, the outside world knows nothing.

Let’s use an imperfect analogy: traditional on-chain operations are like shouting “I’m withdrawing money” in an open public square—everyone can see your path and timing, and frontrunning robots are already waiting at the finish line to steal your spot. A garbled circuit moves the entire process into a locked room—while the verdict is public, the case file is permanently sealed. You know the result, but not the reasoning process.

There are two key technical supports here.

The first is the BABE protocol. In January 2026, Babylon co-founder and Stanford professor David Tse released BABE, reducing the verification cost of Groth16 zero-knowledge proofs by three orders of magnitude. In BitVM3’s early方案, a single garbled-circuit file could be as large as 42 GiB, making large-scale real-world use impossible. BABE compresses off-chain storage and setup costs by an entire three orders of magnitude, moving this from the lab to the edge of practical engineering.

The second is the cut-and-choose protocol.

At present, the generation and verification of garbled circuits depend on specific prover nodes. If the prover is attacked or colludes, the integrity of the circuit may be compromised. The centralization of the prover is itself a single point of failure.

As of May 2026, more than 56,000 BTC (about $5.6 billion) have already been put to work for yield through this mechanism. But between “it can run” and “it can run reliably at scale,” there are still two hurdles: gas costs in a real mainnet environment and the fault-recovery rate.

I’m on board with the direction—but before the data comes out, let’s wait and see, and not change anything.
#baby $BABY @BabylonLabs_io

There’s one thing I make sure to check every day before the market opens—oil prices.

Last night, WTI crude oil plunged by as much as 7%, slipping below $80. When oil prices fall, inflation expectations cool off, U.S. Treasury yields move lower, and tech stocks get excited. But the problem is that something could happen again at any moment in the Middle East. Iran has already warned that it “will never allow the U.S. to open a shipping route through the Strait of Hormuz.” This whole oil-price situation is like a time bomb—anything could set it off at any moment. So while I’m holding Sandisk, I’ve also added a bit of a gold ETF to hedge on both sides.
#TradFi晒单

Recently I came up with a stupidly simple way to trade SanDisk, and the results are okay.

Basically, set the Bollinger Band parameters to 20, 2. Buy a little when the stock price hits the lower band, and sell when it rebounds near the upper band. Last week, when SanDisk dropped to 1121, it happened to land right on the lower band. I bought a little, and today it rebounded to around 1288, and I exited most of my position. Even though I didn’t sell at the absolute peak, I still made a steady profit of more than ten points. This method doesn’t guarantee profit every time, but at least it gives me a basis for entries and exits, so I don’t have to blindly guess based on feelings every day. When combined with the RSI, once it drops below 30, it’s basically a short-term bottom.
#TradFi晒单

Talking about Babylon’s BTCVault, the phrase “self-custody” really sounds tougher—more hardcore—than the kind of custody model you get with wBTC. The entire BTC is kept locked in the Bitcoin mainnet’s Taproot scripts—no bridging, no wrapping, no third-party custody. It sounds like a one-and-done solution to the question of “who holds the coins.”

But after going through the official documentation, I found that “being able to retrieve it yourself” comes with a very real prerequisite—self-service recovery doesn’t rely on just a mnemonic phrase.

Sounds very free, right? But the trade-off is that you have to manage more things. Losing your wallet mnemonic is obviously troublesome, but losing the WOTS file and the claimer artifacts is equally fatal. The official docs state it plainly: if you lose the WOTS file or the artifacts (but not both), self-claiming will no longer be possible. If you lose both, and the Vault Provider doesn’t respond, then all you can do is contact the team to initiate the Security Council recovery process. In the testnet phase, Security Council uses five keys with a 3-of- multi-sig threshold—it can’t transfer your BTC away, but it can stop payments in an emergency scenario. This isn’t handing control to others, but it does add a layer that “requires others to cooperate.”

There’s also an easily overlooked timing issue. From submitting the request to create a vault to activation takes about 2 hours, and the activation window is roughly 48 hours. If it times out, it enters the Expired state. The refund time lock is about 3 days. Whether the testnet can run through the process is one thing; whether a large number of mainnet users can recover smoothly under failure conditions is another.

So I won’t simply understand BTCVault as “having no custody risk.” More accurately, it swaps custody risk for key-management risk, file-backup risk, and user-operational risk. Self-custody isn’t risk-free—it just moves the risk from “trust someone else” to “manage your own stuff.” And managing your own stuff is far more complicated than you might think.

Next, I’ll wait for the publicly available recovery drill data—failure rates and average redemption time—before judging just how practical this self-service channel really is.
#baby $BABY @BabylonLabs_io

The US stock market feels a bit like a knockoff—there are especially huge fluctuations even before the market opens. Mainly, it likes to trade sideways to wear people down and exhaust their patience. But I found that during the day, buying high and selling low still has a decent win rate. Since the swings are so large, it can drop back within a short time. The main reason is that I feel there will probably be another round of the drop afterward—it’s not possible for it to quickly return to 1500!
Selling everything before the market opens is a good habit…
#TradFi晒单

I rarely play contracts. I’m slowly learning how to play them, but I still don’t understand this multiplier. I opened at 50x, but it dropped by five moves—why am I only making a bit more than one dollar? I’m really confused. Spot trading is easier: I buy how much, and when it rises by ten points, that’s ten percent. I feel that Sandisk will pick up these days—the market is gradually coming back! #TradFi晒单

Babylon’s value proposition has always been very clear: BTC stays on the base layer, with self-custody—no bridging, no wrapping. It sounds hardcore, and for that reason I originally wanted to toss part of my BTC in and try it.

But as I read further through the technical documentation, I found it’s not that simple. BTC really hasn’t left the mainnet, but you have to hand over “finality” to a person—Finality Provider (FP). And currently, a single BTC can only be staked to one FP; it can’t be distributed across multiple FPs. Choosing one FP means betting on its security. If it goes wrong, your BTC faces the same risk of being slashed.

An FP must continuously submit randomness and finality signatures. If, at the same block height, it signs two conflicting blocks (a double-signing), the EOTS private key will be exposed, and the staked BTC goes directly into the slashing path. This mechanism is designed to prevent malice, but the technical risk is: a single software bug could cause a node to accidentally double-sign, which would also trigger private key exposure and result in slashing of BTC. Once slashed, the FP’s voting power is permanently zeroed out (tombstoning)—it can never be restored.

Choosing an FP isn’t just about yield. Babylon’s documentation explicitly states that the EOTS key and the Genesis key cannot be rotated. If key management goes wrong, the consequences are severe. Zellic’s audit report also pointed out that Babylon’s BLS keystore implementation had previously stored the key password in plaintext in files on the node machine—just imagining that is chilling.

There’s another, more hidden layer of risk: Babylon’s slashing mechanism uses entirely new cryptographic primitives, with no historical precedents. Running an untested mechanism at a scale of billions of dollars means any edge case could be catastrophic.

The phrase “the coins are still in your own hands” masks another risk: who you are ultimately putting your security trust in. In Babylon’s current state—before it truly supports delegating a single BTC to multiple independent FPs, and before the dashboard can clearly display each FP’s real-time signing rate, historical downtime, and slashing records—I won’t put BTC in. At this stage, the cost of choosing the wrong FP is no different from being hacked.
#baby $BABY @BabylonLabs_io

When discussing Babylon’s governance, I noticed an interesting misalignment.

BTC stakers lock their assets, bear slashing risk, and provide economic security to the network, but when it comes to on-chain governance, voting power belongs only to the people holding $BABY . The documentation states this very clearly—BABY is responsible for Gas, staking, and governance, while BTC provides additional economic security, with roles clearly divided. Logically, that makes sense, but the people who bear the asset risk don’t have direct voting rights over the rules—it just feels like something’s off.

One detail is worth paying attention to: for a regular proposal, you need 50,000 BABY just to enter the vote (voting period: 3 days); for an expedited proposal, you need 200,000 BABY (voting period: 1 day). The threshold itself isn’t the main issue. The problem arises when proposals involve staking parameters, reward allocation, or Finality Provider rules—those who bear the risk of BTC being slashed can only enter the on-chain process by holding BABY, or they can express opinions off-chain. There may be little overlap between the people who actually bear the asset-slashing risk and the people who obtain voting power.

For stakers to participate in governance, the cost isn’t low. Regular proposals require 50,000 BABY to get onto the chain for voting, while expedited proposals require 200,000 BABY. The voting window is only 1 day. If users don’t actively vote, their voting power will be inherited from the validator’s choice. Even though users can vote themselves to override, during a one-day expedited window, many may not even see the proposal.

The 66.7% approval threshold can block malicious proposals, but it doesn’t solve the asymmetry in information and response time—validators are online long-term and familiar with the parameters, while ordinary users may not even receive a reminder.

Babylon’s governance documentation emphasizes “time-locked and accountable”—all voting results and token allocations will be recorded. It’s not simply one vote after another, but it makes me think of a question: who truly has the time, funding, and organizational capability to influence outcomes?

Governance being written on-chain doesn’t automatically mean it’s naturally fair. BTC stakers provide security, but they lack rule-making voice. When staking rules and slashing parameters are adjusted, can the people bearing the risk speak up in time?

Next, I’ll keep an eye on the voter turnout, the ratio of voting power carried by validators, the frequency of expedited proposals, and whether, when BTC staking rules are involved, there is a dedicated process to solicit input from BTC users.
#baby $BABY @BabylonLabs_io

Talking about Babylon nodes: Many people think that running a Vigilante (sentinel) requires pairing it with a Bitcoin full node, otherwise you can’t play properly. I thought so at first too. Later, I read the official docs and found that this “vigilante” layer can actually be run in a light-client mode—saving disk space, but with reduced verification capability.

Babylon’s BTC Light Client module does not store the full blocks; it only syncs the block header chain. The light client uses SPV (Simplified Payment Verification) to verify Merkle branches, allowing it to confirm the current mainnet height and whether the header chain is continuous. The vigilante does two things: (1) monitors whether the Finality Provider has produced any conflicting double-signature blocks, and if it detects wrongdoing, it uses EOTS’s two signatures to recover the private key, then broadcasts the slashing transaction. Recovering the private key is purely cryptographic computation and does not rely on full-node history—so the light client can do it too. However, it cannot independently verify whether a specific UTXO truly was locked into Babylon’s scripts; that requires a full node or an indexer in cooperation. The light client only trusts the header chain.

In the official documentation’s installation guide for Vigilante, it requests “a synced Bitcoin full node.” But in practice, running the sentinel script using Bitcoin Core in light mode (prune=1) passed the tests: after locking 0.05 testnet BTC, the sentinel reported an FP skip signature once, and nothing was missed.

If the mainnet experiences a deep reorg (more than 6 blocks), the light client may temporarily misjudge the timestamp position, whereas the full node would detect it first. What you save with the light client is disk space (about 80MB/year). The trade-off is that if your header source gets polluted, you could be led astray. For personal sentinel operation, using a light client is enough—provided you trust the header source you chose. If you’re more paranoid, run a full node plus an indexer and verify everything yourself.

It’s not either/or; it’s a trade-off between cost and how much you self-host and verify.
#baby $BABY @BabylonLabs_io

Bitcoin’s applications in DeFi—Babylon has recently been one of the most aggressive players.

As of July 2026, its staked TVL has surpassed $4 billion. For a protocol that has been live for a little over a year, this scale is certainly eye-catching.

But what really makes it feel different isn’t the TVL number itself—it’s what it’s been doing lately: putting native BTC into Aave V4.

Previously, for BTC holders who wanted to borrow money, they either wrapped BTC into wBTC and handed it to a custodian, or used a cross-chain bridge. Both require trusting someone. Babylon’s Trustless Bitcoin Vaults (TBV) logic flips this around: BTC never leaves the Bitcoin network; it’s locked inside Taproot UTXOs. On the Ethereum side, a “verifiable collateral status” is generated so that Aave knows you have BTC available to borrow.

This isn’t just pie in the sky. On May 26, 2026, Babylon Labs already launched a temperature-check proposal in the Aave community, planning to deploy two Spokes: the Core Lending Spoke will handle borrowing and lending, while the BTC Vault Swap Spoke will manage liquidation and settlement. Aave founder Stani Kulechov said this could unlock more than $4 billion in liquidity. The collateral exists in the form of vaultBTC—an ERC-20 token with transfer restrictions, limited to whitelisted addresses only. During the liquidation process, liquidators can immediately settle with WBTC; later, arbitrageurs complete the Bitcoin-side redemption. In the testnet phase, users can deposit real BTC and borrow USDC or USDT on Ethereum—throughout the entire process, the private key remains under the user’s control.

No packaging, no cross-chain, no custody—BTC’s DeFi utilization has long hovered at extremely low levels. This playbook is indeed aimed at solving that.

Partnerships are also moving forward. GoMining plans to activate up to 1,000 BTC (about $82 million) via TBV: users lock BTC, borrow stablecoins to fund mining rigs, and earn BTC rewards. Aegis plans to combine TBV with fixed-rate lending in Q4 2026. In January 2026, a16z invested $15 million to support TBV development.

But it’s not that simple. The annualized yield from staking BTC is only around 0.05%, and choosing the wrong Finality Provider comes with Slashing risk. In January 2026, a vulnerability was exposed in Babylon’s BLS signature code—malicious validators could slow down block production. On the TBV testnet, the per-vault cap is only 0.4 BTC, so it still has a long way to go before it can support large-scale capacity.

I agree with the direction—turning Bitcoin from “digital gold” into a reusable security resource.
#baby $BABY @BabylonLabs_io

I was just wondering why a company as good as Hynix is seeing its stock price keep falling now—doesn’t anyone believe in it?
I’ve been wanting to buy it. The price used to be too high, but it has finally come down, so I can go bargain hunting. I won’t just look at short-term fluctuations; I’m going to buy spot and hold it for the long term—so this should be a sure thing, right? I plan to keep making regular investments in Hynix. This time I’m definitely going to win! #TradFi晒单

Talking about Bitcoin staking, you can’t get around Babylon.

Open DeFiLlama: Babylon’s TVL figure is eye-catching—its peak hit about $7.2 billion, and it’s since stabilized around $4 billion. But there’s a point that’s easy to misunderstand behind that number: staking TVL doesn’t equal borrowing demand.

Stakers are aiming for self-custody plus yield—the locked-in BTC is used to serve PoS chains as “security guards.” But people who actually want to pledge BTC to borrow stablecoins need to consider the collateral ratio, interest rate, liquidation risk, and redemption time—their motivations are completely different.

If you simply treat the staking scale as borrowing demand, it’s like counting all time deposits as borrowing customers: the numbers look huge, but the behavior hasn’t really happened.

TBV (Trustless Bitcoin Vaults) is the line that truly points to lending/borrowing. It’s currently running on the Bitcoin signet and Ethereum Sepolia testnets. The per-vault and per-address cap is 0.4 test BTC, and the overall Aave app cap is 10 BTC. Testnet limits are good for validating the flow, but they can’t prove real demand.

The partner list is indeed getting longer. GoMining plans to activate up to 1,000 BTC through TBV. Aegis plans to provide fixed-rate lending/borrowing in Q4 2026. But the official wording is all “plans” and “depending on testing”—partnerships may be treated as demand assumptions, but you can’t book them as revenue in advance.

My current view is to break it into two tables: staking scale looks at TVL and Finality Provider performance; borrowing demand looks at the mainnet TBV’s real collateral amounts, outstanding borrow balances, and liquidation data. The first one already has volume; the second is still being built from zero.

With BABY, to obtain a new valuation logic, the key isn’t to keep citing staking TVL—it’s to make the second table actually grow. Once you see real BTC being lent out consistently for stablecoins, applications willing to pay, and fees flowing back to BABY holders, then it won’t just be the same batch of BTC changing statistical definitions.

The above is my personal observation. DYOR.
#baby $BABY @BabylonLabs_io

There’s an old joke in the crypto world: besides just holding Bitcoin, what else can it do?

Babylon has a new answer—staking. Not the kind where you package BTC into wBTC and toss it into some pool to mine; this is genuinely “hands-off” with your BTC. Your BTC is locked in Bitcoin’s own Taproot UTXO on-chain, stays fully offline (never goes out), and then you use the staking weight to act as a “security guard” for other PoS chains.

Sounds pretty convoluted, right? I didn’t really get it the first time either.

The core logic is like this: what does a PoS chain fear the most? It fears its token price dropping—security budget shrinks, and validators may abandon the network. Babylon’s idea is to let BTC holders delegate Bitcoin’s “voting power” to PoS chains so they can produce finality signatures. If someone misbehaves? EOTS cryptography ensures the private key is automatically exposed, and penalties are triggered automatically.

As of mid-May 2026, Babylon’s TVL has reached $5.6 billion, exceeding 56,000 BTC. The largest Bitcoin-native yield protocol in the world—no contest. Aave, dYdX, and Polygon have already integrated it.

But it’s not that simple.

BABY’s total supply is 10 billion coins. Its circulating market cap is currently between $52 million and $55 million. The TVL is nearly 100 times that market cap. This ratio is extremely rare in DeFi. Either the TVL is overstated, or BABY is undervalued. I lean toward the former: the amount of BTC staked is indeed large, but BABY, as the token of a new chain, hasn’t fully completed the economic flywheel yet. The rewards validators receive are BABY, and BTC stakers also get BABY—so in a chain’s economic loop, three tokens flow in: the BSN token, BABY, and BTC. The conversion costs and slippage ultimately get passed on to users.

Babylon also has another shared security architecture called BSN (Bitcoin Security Network). The mainnet’s first phase has already started, with roughly 250 Finality Providers active. In 2026 Q3–Q4, it plans to move into a multi-staking phase, where a single BTC stake can secure multiple PoS chains at the same time. The direction is definitely right—turn Bitcoin from a “digital gold” asset into reusable security infrastructure. But fast TVL growth doesn’t mean the ecosystem is already mature. Issues like integration thresholds, operational costs, and cross-chain coordination still haven’t been truly tested.

I’ll keep watching the TVL growth rate and the actual number of BSN-connected chains. It’s too early to draw conclusions now.
#baby $BABY @BabylonLabs_io

On the 10th of every month, when I see 136 million BABY unlocked, I open a short—this trade might have been wrong in two ways: first, I undercounted the total supply; second, I treated “can be sold” as if it were “will be sold.”

Let’s do the math using official data: every month, insiders unlock about 136 million tokens (4.9 billion insider allocation ÷ 36 months). For the ecosystem and R&D allocation, 25% unlock in the first year, and the remainder releases linearly over 36 months starting from the one-year mark—about 112.5 million tokens. A 5.5% annual inflation rate translates to roughly 45.83 million tokens per month. Summing the three, the net new BABY entering a potentially liquid state is about 294 million per month.

But these 294 million are only “how many could theoretically be sold.”

Different people who receive coins have different plans. Early investors might keep holding; the ecosystem payees need to pay salaries and audit fees, so they’re more inclined to swap into stablecoins; BTC stakers who receive BABY rewards might sell immediately to get back to BTC. The urge to sell varies completely.

So what really matters is whether the order book can actually absorb it. You can’t simply use the 24-hour trading volume as buy-side demand—those same coins can trade back and forth many times. Within a 2% range below the current price, how many truly standing buy orders are there, and can proactive buying overwhelm the exchange’s net inflow? Those reflect more closely the real ability to absorb.

On-chain transfers for BABY can be checked, but what’s missing from the official side is a continuously updated address list: which coins came from the foundation’s distribution, which came from staking rewards, and finally how many end up flowing into exchanges. Without that data, the calculated sell pressure is always guesswork.

The unlock schedule answers “the maximum that could be sold,” address flow answers “how much is likely to be sold,” and order-book absorption determines “how much price needs to drop to get it sold.” Looking at these three metrics together is more reliable than focusing on just one unlock date.
#baby $BABY @BabylonLabs_io

Let’s talk about the Babylon whitepaper, starting from its most core question: the security of PoS chains is too expensive.

The whitepaper opens by running the numbers. In the Cosmos ecosystem, many new chains offer initial annualized inflation rates of anywhere from 20% to 100% just to attract stakers. Meanwhile, Bitcoin is an asset with a market cap of $600 billion or even more, and for most of the time it just sits there. On one side, it’s desperately short on security; on the other, there’s money with nowhere to be used. What Babylon wants to do is connect this supply and demand.

The most essential selling point is “Bitcoin stays in custody.” No cross-chain bridge, no wrapping into wBTC—BTC remains on the Bitcoin network, under your own control. How does it achieve this? The whitepaper mentions two key designs.

One is EOTS (Extractable One-Time Signatures). Put simply, it’s a mathematical lock. If you stake honestly, nothing happens. But if you try to cheat—for example, performing double-spends on another chain—this mechanism will automatically expose your private key, and the staked BTC will be slashed and forfeited. The other is a timestamping mechanism that anchors critical events on the PoS chain to Bitcoin blocks. Since nobody can change Bitcoin’s ledger, it effectively provides a Bitcoin-level insurance layer for the security of the PoS chain.

The mainnet launched in August 2024. To date, the locked value (TVL) has already exceeded $5.6 billion (over 56,000 BTC). dYdX and Polygon have already integrated it. Aave is also pushing a proposal to use native Bitcoin as collateral. The founder is Stanford professor David Tse, a winner of the IEEE Shannon Award. In terms of funding, Paradigm led an investment of $70 million, and later a16z also participated with $15 million.

The whitepaper lays out the framework very clearly, and I’m aligned with the direction—put dormant Bitcoin to work, while helping PoS chains solve the security cost problem. But the risks are also there: the slashing mechanism currently mainly targets double-signing and doesn’t cover all malicious scenarios; a 15-month staking period is not very friendly to users who need higher liquidity. Whether this narrative can ultimately hold water depends on whether the ecosystem keeps adopting it and whether TVL growth can sustain the model. I’ll keep watching—no rush for now.
#baby $BABY @BabylonLabs_io

I recently looked into a project called Babylon, and honestly, it’s pretty interesting. What it wants to do is enable Bitcoin to be staked.

You know, Bitcoin has the largest market cap in the world, but on-chain it’s basically just sitting there. Besides holding, selling, or packaging it into wBTC to mess around in DeFi, it doesn’t seem like there’s much else you can do. Wrapping it isn’t reliable either—you have to trust the custodian, and if anything goes wrong, wBTC can collapse overnight. What Babylon wants to do is this—keep your Bitcoin in your own wallet, lock it in a self-custodied vault, and then use it to act as “security” for other chains to earn yield. All the way through you don’t need a cross-chain bridge, you don’t need to swap assets, and you don’t have to hand over your private keys.

Sounds pretty great, right? My first reaction back then was: can this actually work?

Later I read its whitepaper and found the technical path is indeed quite hardcore. The core is something called EOTS (Extractable One-Time Signature). In simple terms: if you stake honestly, nothing happens. If you try to cheat—for example, double-spend or do malicious stuff—then the mathematical mechanism will automatically expose your private key, and the staked Bitcoin will be forfeited. This is harsher than the slashing mechanisms of many PoS chains—most only slash your staked coins. Babylon’s design is to directly hand over your private key, leaving you totally exposed.

This made me feel it’s not some makeshift operation. On the funding side, Paradigm led a $70 million round, and later a16z also invested $15 million for buying tokens. The total funding is close to $100 million.

TVL has already exceeded $5.6 billion, making it the largest Bitcoin native yield protocol in the world right now. Aave, dYdX, and Polygon are all integrated.

But I still have two concerns in my mind. First is the staking period—currently fixed at 15 months. That’s too long. If the market suddenly turns, you can’t even run. Second, the slashing mechanism currently mainly protects against double-signing. If someone commits other types of malicious acts, I haven’t seen a complete fallback plan.

I think the direction is solid, the team is solid, and the data is beautiful too. But I’ll keep watching for a while longer before deciding.
#baby $BABY @BabylonLabs_io

A couple of days ago, I chatted with a market-making friend. He asked me how the maker strategies I’ve been running on GRVT are doing lately. I showed him the backend data, and his first reaction was: “Is the rebate really real?” I chuckled and told him that I didn’t believe it at first either—I only confirmed after running for a week that I wasn’t seeing things.

First, let’s talk about the order book. In normal trading hours, the first-tier spread for BTC perpetuals is basically stable at around 0.5 to 1.5 bps. The first-tier depth is roughly 100k USDT, and there’s another 200k to 300k USDT sitting in the second tier. On-chain, this is already pretty strong for Perps. Compared with top CEXs it’s still some distance away, but compared to DEXs whose order books are as thin as paper, the advantage is very clear.

What really made me seriously run as a maker is the fee structure. GRVT’s maker fee is negative: as low as -0.01 bps, and as high as -0.3 bps. In plain terms, when your orders get filled, you don’t just avoid paying fees—the platform actually gives you money back. For my current account, the maker fee is about -0.0008%, while the taker fee is 0.039%. This fee difference is extremely friendly for spread-capture strategies. As long as your orders can keep getting picked off, you can cover inventory risk and still make money.

I’m running a fairly simple two-sided quoting strategy: in the BTC first-tier spread, I place bid and ask orders each for a bit over 1000 USDT, relying on takers to actively consume. Over the past week, the filled volume from my resting orders was around a few hundred thousand USDT. When you add up the rebates plus spread income, it comes out to dozens of dollars. After subtracting slippage and the cost of inventory hedging, my net profit is a bit over $20. Not mind-blowing, but steady. Converted to an annualized figure, it comes to just over 20%, which is already a decent number in this market.

The main risk is inventory accumulation. The on-chain matching rhythm isn’t as even as on CEXs. Sometimes one side gets eaten through consecutively, and once your position skews toward one side, you have to actively hedge. My current approach is to open the opposite position on another platform for hedging—this is standard practice for quantitative market makers, but it adds an extra layer of cost.

For individual market-making players, GRVT is currently at a stage of “you can come in and test.” Liquidity is sufficient, the fee structure is favorable, and the zk-proof mechanism hasn’t slowed things down. There are still some rough edges, but the direction is right. I’ll keep running, and after the TGE, I’ll see whether more policies come out.
#grvt @grvt_io