Arthur Cheong：大清洗後的主動管理新時代

0:00加密市場大清洗後的主動管理契機

0:00大家早安。今天我要談的是加密貨幣市場中主動管理的理由。將近一年前，我們經歷了加密史上最大規模的清算日之一。一天之內有超過 190 億美元的部位被清算。那麼，市場在這 12 個月後教會了我們什麼？我的論點很簡單：這是近年來主動管理最有力的時刻，不只是口號，而是有資料支持。這並不是因為像我們這樣的主動管理人比較聰明，而是因為加密貨幣的機會組合已經改變，成長已經不再集中在最大市值的資產和主流幣種上。Good morning, everyone. So today I'm going to talk about the case for active management in crypto. Almost one year ago, we had one of the biggest liquidation day in the crypto history. More than 19 billions of position were liquidated in one day. So, what the market have taught us after that 12 months? And my argument is simple. This is the strongest case in years for active management, not as a slogan, but backed by data. And this is not because active managers like us are smarter. But because the opportunity set in crypto has changed, the growth is no longer in all the biggest assets and major.

0:58我會先給大家看資料，接著帶大家看三個案例，最後說明我認為現今主動管理的意義。剛剛提到，一天內有超過 190 億美元被清算。一般的反應可能會是：這是一個重置，把槓桿洗掉，市場會再創新高，持有主流幣就好。但 12 個月後，比特幣仍比一年前下跌約 24%，以太幣跌近 30%，Solana 跌 36%。事實上，今年六月，比特幣曾跌破六萬美元，最大回撤超過 50%。如果策略只是買主流幣等市場反彈，大多數人到現在還沒回本。I'll show you the numbers first, then walk through three case studies and close with what I think what active management means today. So I mentioned just now, more than 19 billion were liquidated in one day. The normal reaction might be, okay, this is a reset, flush the leverage, we will come back higher again. Own the majors. And 12 months later, Bitcoin is still down around 24% compared to one year ago. Ether close to 30%, Solana 36%. And actually, in June this year, painful memory, Bitcoin traded below $60,000, roughly like 50%, more than 50% max drawdown. So if the strategy is just to, okay, I'm buy the major and wait for the market to rebound, most people still are not back to par yet.

1:57這就是加密貨幣的被動基準，大家通常會拿來比較的對象。現在，我要給大家看同一時期、同一市場，其他資產類別的表現。藍色的是被動投資、主流幣種，比特幣跌 24%，以太幣和 Solana 更多。今天我要用三個代幣作為案例：Zcash 漲了 483%，Hyperliquid 漲 145%，NEAR 漲 114%。這三個在大清洗當天市值都超過 10 億美元，不是那種你買不到的小幣或垃圾幣。Hyperliquid 當時其實已經是 100 億美元資產。以比特幣為基準，Zcash 表現超越 500%，Hyperliquid 169%，NEAR 139%。這是非常大的超額報酬。And that is the passive benchmark of crypto, what people usually benchmark to. And now, I'll show you what have happened to the same market in the same period for the other asset class. So here's the same 12 months, same market side by side. In the blue, you have the passive, the majors. Bitcoin down 24%, yeah, and Ether and so more. And the three tokens I will use as a case study today, Zcash up 483%, Hyperliquid up 145%, and Near up 114%. And all three of them had a market cap of above a billion dollar at the day of the washout. So these are not micro caps or shitters, as I like to call it, that you couldn't buy. Hyperliquid was actually already a $10 billion asset. And measured against Bitcoin, you look at it, Zcash have outperformed by 500%, Hyperliquid 169%, and Near 139%. This is a huge amount of outperformance.

3:09如果把這三個代幣等權重放在一起，漲幅大約 250%，而主流幣等權重則是負 30% 左右。我不是要大家去買這三個代幣，而是要說明主流幣和產業中最佳機會的落差從未這麼大。只持有最大資產的策略會錯過這些機會。這不只是結果，而是過程。圖表上所有資料都是以大清洗當天為基準 100，對數刻度。要注意兩件事：第一，贏家不是一開始就領先。事實上，清洗後前三個月大家都一起下跌。NEAR 在二月跌破一美元，HYPE 一月跌到約 21 美元，Zcash（他們喜歡叫 Zed）在三月初把去年十一月的漲幅全吐回去。If you put all three of them in an equal weight basket, they're up around 250% against roughly minus 30% for an equal weight basket of the majors. Now, I'm not showing this to say buy these three tokens. I'm using this to explain that the gap between the majors and the best opportunities in the industry has never been wider. A strategy that only owns the largest assets misses every one of these. So this is the path, not just the endpoint. So everything on this chart is indexed to 100 on the day of the washout, on a log scale. So two things to notice. First. The winner did not win from day one. In fact, from the first three months of the washout, they all fell together. NEAR traded below a dollar in February, HYPE fell to about $21 in January, and Zcash, or they like to call Zed for the token, gave back its entire November rally by early March.

4:21如果你在今年二月看這張圖，會發現 Zcash 大致回到起點，其他線都還在水下。但從四月開始，差距才開始拉開，驅動因素對每個資產都很明確：有的是隱私 ETF，有的是衍生品平台搶市佔，有的是跨鏈產品加上 AI 題材。這就是為什麼我一直強調要逐一檢視這些資產。這裡用不同方式呈現同一組資料，中間欄是每個資產的最大回撤，右邊是兩天前的最終表現。你可以看到它們的交錯情況。Zcash 最大回撤高達-72%，但報酬也是最高的。其他像 Solana 最慘時跌近 70%，最後還是跌 36%。If you had looked at this chart in February this year, you have seen Zcash roughly back at where it started and every other line underwater. But the gap started to open from April onwards, and the drivers were very specific to each asset: a privacy ETF, a derivative platforms taking market share. And a cross-chain product with a real volume plus the AI narrative. That's why I will keep coming back to the idea that you have to look at these assets one at a time. Here's the same data in a different way. The middle column is the max drawdown of each asset. The right column. It's where it ended up as of two days ago. Look at the interleaf, how they interleaf. Zcash had the deepest drawdown of all the six, minus 72%, and the best returns. The rest didn't done as well, like Solana fell almost 70% at the worst and ended up down 36%.

5:34如果你看所有最大回撤，其實主流幣並沒有比我強調的這三個幣好多少，但主流幣年對年還是負報酬，而這三個代幣即使經歷深度修正，現在已經大幅上漲。大家都經歷了大修正，問題不是你持有什麼，而是你持有這些資產有沒有得到報酬。這對主動管理的做法很重要。我認為選擇是報酬來源，但部位大小和風險管理才讓你能撐得夠久去收割這些報酬。So in fact, if you look at all the max drawdowns, the major did not really do much better than the three coins I highlighted. And yet they are still negative year-on-year basis, while the three tokens I have selected, they are already up massively even after suffering a very deep drawdown. So they all suffered a huge correction. The only question is. Not what you own, but are we being paid for owning these assets? Because this matters for how active management should be done. I would argue that selection is the source of return, but the sizing and risk management are what let you stay in the trade long enough for you to collect it.

6:23當然會有人質疑：這不是事後馬後炮嗎？我挑了三個表現最好的資產。但我要強調，我不是叫大家去買山寨幣。事實上，前 50 大山寨幣只有 9 個一年內表現超越比特幣，41 個都更差，而且差很多。被動持有山寨幣組合不會比比特幣好。我也不是說這三個幣或加密貨幣很明顯，而是說它們是可以分析的，每個都有公開的證據，很多時候在價格起漲前就能看到。這就是接下來案例要說明的。有這麼大的分歧，市場曝險不會帶來報酬，選對資產才會。Now, there will be some obvious objection. Cherry-picking, right? I picked the three best assets after the fact. And I'm not arguing that you should go and buy altcoins, okay? In fact, only nine of the top 50 altcoins have outperformed Bitcoin on a one-year basis. 41 did worse, and actually way worse. A passive basket of altcoins would not have done better than Bitcoin. And I'm not claiming these three tokens or crypto are obvious. I'm just claiming that they were analyzable. Each one had an evidence you could look at in public, often before the price moved. And that's what the case study will show. With this much dispersion, market exposure doesn't pay you. Picking the right asset does.

7:19加密貨幣不再是一筆交易，產業分化明顯

7:19我的主要論點是：加密貨幣已經不再只是一筆交易。過去大部分時間，你可以把它當作一個整體：比特幣漲，其他幣漲更多；比特幣跌，其他幣跌更慘。這是過去七、八年一直如此，讓很多人覺得只要持有主流幣就等於持有整個加密資產類別。但我認為這已經不成立了。現在產業有明確的分工：隱私、鏈上金融市場、跨鏈基礎建設、AI。我再強調一次，AI、AI、AI。這些領域成長的原因和時機都不同，最大資產已經無法捕捉所有成長，有時甚至完全沒參與到。所以像我們這樣的投資人，必須逐一評估加密機會，就像股票投資人分析個別公司，而不是只買指數。So the main argument is crypto is not one trade anymore. For most of this industry history, you can kind of treat it as one. Bitcoin went up, everything went up more. Bitcoin went down, everything went down more. That has been the case. For the past seven, eight years. And that led a lot of people to think that, hey, owning the majors was a reasonable approximation of owning crypto, the asset class. I would argue that that's no longer true. The industry right now has distinct businesses: privacy, on-chain financial market. Cross-chain infrastructure and AI. And I would like to use the word AI again. AI, AI, AI. So they grow for different reasons, different times. The largest asset don't capture all the growth anymore. And sometimes they don't even capture any of it. So investors like us, we need to judge the crypto opportunities. One by one. The same way like an equity investor, look at individual companies and not just buy the index.

8:35那要怎麼逐一評估機會？你要針對每個資產問對問題，再去找證據。接下來我會用三個案例來說明。第一個是 Zcash：是什麼讓投資人重新評價私密數位貨幣的價值？第二個是 HYPE：作為加密貨幣最成功的去中心化交易平台，什麼是合理的估值倍數？第三個是 NEAR：加密貨幣真的能從 AI 應用中捕捉經濟價值，還是只是個故事？So how do we judge the opportunities one by one? You start by asking the right question for each asset. Then you go and find the evidence. Three case studies. One. Zcash. What made investors reassess the value of private digital money? Second question, second asset, HYPE. What is the right valuation multiple for the most successful decentralized trading platform in crypto? Third, NEAR. Can crypto actually capture economic value from AI adoption, or is it just a story?

9:23Zcash 案例：公開資料與敘事驅動重估

9:23我們先從 Zcash 開始，漲幅高達 483%。市值從約 37 億美元漲到現在的 220 億美元。問題是，什麼讓市場重新評價私密數位貨幣的價值？首先，有些證據在大漲前就已經公開。左圖可以看到，2024 年初，Zcash 在隱私池（shielded pool）的供應比例約 8%；到 2025 年 1 月升到 11%，2025 年 10 月約 18%。也就是說，價格還沒漲之前，已經有越來越多人選擇用隱私功能。現在這個比例大約 29%。So let's start with Zcash, up 483%. Market cap went from about $3.7 billion to $22 billion today. The question is, what made the market assess what private digital money is worth? So the first thing is some of the evidence was public before the big move. On the left, you can see that the total share of Zcash supply held in the shielded, which means the private pools, are around 8% in early 2024. They went to around 11% at January 2025, and 18% by around October 2025. So people were actually already choosing to use the privacy before the price told them to. And today, it's around 29%.

10:19右圖可以看到，十月第一週，Zashi 這個主要的 Zcash 錢包推出了透過 NEAR Intents 的私密兌換功能。你可以直接從比特幣、Solana 或 USDC 換到 Zcash 的隱私池，不需要中心化交易所。知名天使投資人 Naval Ravikant 曾說，比特幣是對抗法幣的保險，Zcash 則是對抗比特幣的保險。這個敘事真的引爆了。Winklevoss Capital 領投了一家募資 5900 萬美元的金庫公司，最後證實就是 Zcash 的金庫。Grayscale 早在 2017 年就有 Zcash 信託，這讓它有現成的 ETF 通路。事實上，Zcash 在 2025 年 10 月就已經漲了 4 倍，這些訊號早就公開，並不是什麼祕密。On the right, you can see that on the first week of October, Zashi. The main Zcash wallet shipped private swaps through Near Intents. You can go from Bitcoin, Solana, or USDC straight into shielded Zcash without a centralized exchange. Naval Ravikant, a very famous angel investor, had posted, Bitcoin is insurance against fiat. And Zcash is insurance against Bitcoin. And boy, that narrative really took off. Winklevoss Capital led a $59 million raise for a treasury company that turned out to be a Zcash treasury. And Grayscale had a Zcash trust dating back to 2017, which that made it a ready route to an ETF. And to be fair, Zcash already rallied like 4x even in October 2025. So the signals were already there and not secrets.

11:23Zcash：波動與機會

11:23但要承擔這個投資論點並不容易。如果你仔細看，這段期間經歷了很多波動和回撤。細節非常多，我就不一一細講，但最值得注意的是，如果你看第三點，核心開發團隊之間曾發生治理上的爭執和衝突，這導致價格在一月大幅修正。當然，二月時主流幣也大跌，市場也被拖累下來。但 Zcash 價格慢慢回穩，到了第四點，一家大型加密基金 Multicoin 公開持有 Zcash 部位，帶動幣價上漲。但僅僅一個月內，Zcash 就發生了最可怕的事件之一：在 Orchard 隱私池中發現偽造漏洞，這個漏洞自 2022 年就存在。雖然很快修復，但沒有人能證明它未曾被利用。Zcash 兩天內下跌約 50%，但之後又有回升。七月有 Ironwood 升級，漏洞被修補，沒有證據顯示隱私池曾被利用，市場也恢復了元氣。But underwriting this thesis is not easy. If you look at it, they went through a lot of volatility and drawdowns throughout the period. So there's a lot of details there. I'm not going to go through all of it one by one, but the most notable one is, if you look at point three, there was a dispute and clash in governance between the core dev team. So that actually led the price to correct significantly in around January. Obviously, in February, the majors also dropped significantly, so that dragged the market down as well. But the market, the Zcash price slowly recovered, and in the point for a major crypto fund, Multicoin disclosed a position in Zcash that led the coin to rally. But in just one month, one of the scariest events for Zcash had happened. A counterfeiting bug was found in the Orchard shielded pool. And the bug that had been live since 2022. It was fixed very quickly, but nobody could prove that it had not been exploited. And Zcash fell about 50% in two days. But we have recovery afterwards. There was an Ironwood upgrade in July, and the exploit were fixed. There was no evidence that the shooter pool of S was ever exploited, and the market recovered.

12:56八月，美國首檔 Zcash 現貨 ETF 開始交易，短短幾週吸引超過兩億美元資金流入。這些事件並非同時發生，先有使用者需求，接著才有投資管道，最後才是機構持有。這一切發生在兩次重大崩跌之中，其中一次是真正的系統漏洞。And in August, the first U.S. spot Zcash ETF started trading, and that ETF attracted 200 plus millions of inflows in the first few weeks. So there was a lot of events, and they did not come in together. The usage came first, and then came access, and then came the institutional ownership. And it happened despite two crashes. One was a real flaw in the systems.

13:31Zcash 的啟示與持有成本

13:31那麼 Zcash 給我們什麼啟示？首先是使用情況，目前有 29% 的 Zcash 存放在隱私池，這也是 Zcash 與其他加密貨幣最大的不同。再來是持有結構，有一家在 Nasdaq 上市的公司持有約 1.8% 的供應量，機構也透過 ETF 大量進場，這些都是實質的變化。至於持有成本，最大回撤曾達 72%，兩天內也曾跌掉 50%。明年還有歐盟反洗錢規則等監管變化，可能影響資產流動性。So what's the takeaway from Zcash? First, the usage. 29% of all Zcash right now sits in the shielded pool, which is what makes Zcash different from all the other cryptocurrency out there. And ownership. You have a Nasdaq-listed company who owns about 1.8% of the supply, and institutional access. The ETF, which took in a lot of inflow in the first few weeks. Those are real changes. And how the private money get used and who owns it. And what does it cost to hold? A 72% drawdown at worst, and 50% in two days. And next year, we have some regulatory change that might affect the liquidity of the asset, the EU AML rules.

14:21所以結論是，投資論點雖然正確，但你必須撐過回撤才能獲利，這就是主動管理的精髓。主動管理不是每個波動都要交易，而是要有自己的投資論點，並且部位大小要能承受短期錯誤。像我們這種主動管理者，有時會自以為太聰明，想要頻繁操作，結果反而賺得更少。So the lesson is the thesis was right. But you still have to survive the drawdown to get paid. So this is active management. It's not about trading every single move, but it's about having a thesis and size it so you can still hold it through while being wrong for a while. Obviously that for active manager like us, sometimes we like to think we are too smart, we try to trade it more, but actually sometimes we end up. Making less.

14:49Hyperliquid：共識交易與估值重評

14:49接下來我們談 Hyperliquid。我會講快一點，因為時間有限。這大概是加密圈所有聰明資金最共識的交易之一。我認為這個投資論點大家都很清楚，Hyperliquid 是目前加密貨幣圈最成功的去中心化交易平台。去年十月要問的問題是：HYPE 的合理估值是多少？And next, we move on to Hyperliquid. I'll go a little bit faster because we're running a bit short of time. But okay, this is probably one of the most consensus trade for all the smart money in crypto. I think the question is, I think the thesis is very well understood. Hyperliquid is the most successful. Decentralized trading platform in crypto. The question to ask in October last year is, what is the right valuation for Hype?

15:22Hyperliquid 過去兩年表現極佳，不斷贏得市場，執行力很強，各項指標都很亮眼。全市場永續合約未平倉市佔率（包含去中心化交易所）目前約 12%。未平倉量創新高，HYPE 代幣這幾年也大量回購，但營收維持平盤，因為熊市大家交易量變少。以任何標準來看，這都是加密圈最好的生意之一，但問題不在這裡，重點是 HYPE 的合理估值倍數是多少？So, Hyperliquid probably have one of the best run in the past two years in crypto. They keep winning. Really good executions. By all metrics, they've done very well. The market share of open interest across all perp venues, including decentralized exchange, have climbed to around 12% as of now. Open interest is at record high, and a lot of hype were being bought back over the years. But the revenue have stayed flat because we're in a crypto bear market. And naturally, people trade a lot less. So by any measure, this is one of the best business you can find in crypto. But that's not the question. The question to ask is, what is the right valuation multiple for Hype?

16:17去年 HYPE 的市值對營收比大約是 9 倍，現在則約 32 倍。我認為，當市場真正意識到 HYPE 有持續競爭力時，估值倍數會不斷被重評。它從原本比 Coinbase、Robinhood 等中心化平台還低的折價，現在已經追上來，這過程大約花了一年。Coinbase 大約是 6 倍營收、50 倍獲利，Robinhood 則是 21 倍營收、50 到 70 倍獲利。HYPE 還有很大成長空間，重評的關鍵在於市場對其執行力的認知改變，不能只看營收數字。Last year, it was trading at around 9x on the market cap to revenue. And now it's trading around 32 times. And I would argue that when people really start to see that HYPE have staying power, they continue to win, you are seeing the valuation multiple keep getting re-rated. So it went from trading in the discount compared to all the other centralized... Crypto platform like Coinbase or Robinhood, to trading at, I would say, still on par with them. And that took around one year. If you look at it, like Coinbase trade at around six times revenue, adding around like 50 times profit, and Robinhood is around 21 times revenue and probably around like 50 to 70 times profit. So. Hype is still trading at, like, I would say there's still a lot of room for growth compared to two of them. But the re-rating process was really on how the market perception of them have changed, given the continuous execution, and you need to look beyond the metrics of revenue only.

17:30NEAR：AI 敘事與價值捕捉

17:30第三個案例是 NEAR。NEAR 定位為 AI 區塊鏈，其共同創辦人 Illia 是 Transformer 論文的共同作者，這篇論文讓 LLM 成為可能。我們要問的問題是：加密貨幣真的能從 AI 捕捉到經濟價值嗎？對 NEAR 來說，有三個證據面向：可量化部署、有部署但無法量化的資料，以及敘事。The third case study, NEAR. NEAR has positioned itself as the blockchain for AI, and its co-founder Illia co-wrote the Transformer paper, which made LLM possible. The question we ask is, can crypto actually capture economic value from AI? So for NEAR, there's three bucket of evidence: deploy. And measurable, deploy, but no usage data you can measure, and the narrative.

18:04如果看可量化資料，像 intents（你可以把它想成一種跨鏈移動資產的方式，省去技術複雜度），甚至可以讓沒有智慧合約功能的資產（如 Zcash）跨鏈。現在很多 Zcash 隱私需求都透過 NEAR 路由，過去六到十二個月成長很多。NEAR AI Cloud 則提供私有推論，讓你和 LLM 互動時沒人能看到內容。他們也推出 Ironclad，這是他們自家的開源 AI 代理人框架。這些都是實際產品，敘事則是 AI 代理人在 NEAR 鏈上執行交易。價值捕捉也開始跟上，所有 NEAR Intents 已經帶來 2300 萬 NRISE 回購，且每月雙位數成長。雖然波動大，但還有減通膨提案正在治理流程中，預計會通過，減少代幣供給。這些敘事讓 NEAR 迎來強勁催化劑，九月 NEAR.com 上有私有永續合約路由到 Hyperliquid，美國也有首檔 NEAR 現貨 ETF。這次是產品和資產帶動 NEAR 重評。So if you look at the measurable data, the intents, which you can think of as a way of moving your asset cross-chain without all the technical complexity. And you can even do that for assets. With no smart contracts capability like Zcash, than any of the other chains that have no smart contract capabilities. A lot of the on-chain Zcash privacy demand are routed via Near right now, and it has grew a lot over the past six to twelve months. You also have Near AI Cloud, which provide a private inference. So that nobody can see what are you talking to your favorite LLM. They also launched Ironclad, which is a private version and their own version of the OpenCL, like open source AI agent framework. These are the real products, and the narrative and what they are really working on is AI agent conducting transactions. On Near blockchain, yeah. And the value capture have also started to caught up. We have all the Near intents generating 23 millions of NRISE buyback to the Near blockchain, and actually it's still increasing on a double-digit percentage on a month-on-month basis. Granted, it is. A little bit volatile. There's also the inflation reduction. There was a governance proposal being proposed right now, which is, I think, likely to pass, that reduce the token supply. So all these narratives combined have set up NEAR for a very good catalyst rally, and In September this year, which was last month, we are private perpetuals routed to Hyperliquid on NEAR.com and the first spot NEAR ETF on the U.S. market. So product and assets are the one that re-rated NEAR this time.

20:18主動管理的四大原則

20:18總結一下，今天在加密貨幣的主動管理到底是什麼？有四點：第一，每個資產都要獨立評估，像隱私、鏈上金融、AI 都是完全不同的生意，每個都需要不同的問題和證據。如果你只持有單一加密 beta，這三種機會你都會錯過。第二，要在價格反應前找到證據，每個案例都能在公開儀表板上看到基本面領先於價格，這些都不是秘密。第三，你要知道持幣人實際能捕捉什麼價值，以及你實際付出什麼代價。第四，你要根據回撤調整部位，並隨新事實重新評估投資論點。So let me put it all together. What does active management mean in crypto today? So four things. One, we need to judge each asset on its own. Privacy, on-chain finance, and AI are all different businesses. Each needs its own questions and its own evidence. If you just own a single crypto beta, you will miss all three. Two, we should find the evidence before the price, and in every case, you have start to see some public fundamentals on the public dashboards ahead of the price actions. And none of it was secret. Three, you should know what the token holders actually capture and what you actually pay. And the fourth is you should size for the drawdowns and re-underwrite on the new facts.

21:12這就是加密主動管理的本質，因為總會有新事實和情境出現，你必須不斷重審投資論點。這三個大贏家在一年內都曾下跌 56% 到 72%，還發生很多壞事。如果你的部位無法撐過回撤，就拿不到報酬。當事實或價格改變時，你的看法也要跟著調整。這些都不稀奇，但現在加密市場開始真正獎勵主動管理。And I think this is really the essence of crypto management, active management in crypto, because there's always new facts and circumstances that emerge. And you need to keep re-underwriting the thesis. And all three of them, the big winners, they all fell between 56 to 72% inside the year, and a lot of bad events. So if your position size can't survive the drawdown, you can't collect the return. And when the fact or price change, your view should too. None of this is exotic, but what is new is that crypto now rewards this active management.

21:53結論：選擇比持有更重要

21:53總結我剛剛說的，我認為現在在加密市場，單純持有市場 beta 已經不像以前那麼有利，選擇才是關鍵。如果你只持有主流幣，年對年其實還是虧損，但如果你選對資產，報酬可以高達 150% 到 480%。所以現在每個機會都要獨立評估，這就是我對加密主動管理的論點。謝謝大家。So I'll conclude that what I've said so far, I believe that right now in crypto, owning the market beta alone doesn't pay as well as it used to, but selection did. If you just own the major, you're still down on a year-on-year basis. But if you actually own the right asset, you're up between 150% to 480%. So we should always judge each opportunity on its own in crypto right now. And that is my case for active management in crypto. Thank you.