While the US slept, Asian traders sold everything that wasn't nailed down. BTC absorbed $592.9M in sell-side volume with essentially zero buying. ETH fared worse at $808.5M in sells. No pumps. No dumps either — just an orderly, one-directional liquidation of risk across all major pairs. Boris has seen quieter sessions. Not many.
Good morning. Pull up your charts, pour your coffee, and prepare yourself for a briefing that is — how do I put this gently — not particularly encouraging. While America slept, the Asian session delivered one of the more lopsided order flow readings I have tracked in recent memory. Total sell pressure across monitored exchanges came in at $1,757.4 million against a buy-side tally of exactly $1.7 million. That is not a typo. The ratio is roughly 1,034-to-1 in favor of sellers. If you are a bull, you may want to sit down.
To be fair — and Boris is always fair, even when the data is not — this session registered zero major pump events and zero major dump events in the traditional sense. No single coin collapsed 20% overnight. There were no cascading liquidations that lit up the feeds. What we had instead was something arguably more unsettling: quiet, sustained, professional-grade selling across BTC, ETH, SOL, and 44 separate order flow imbalance events. This is not panic. This is distribution. And that distinction matters a great deal for how US traders should approach the open.
The 17 arbitrage windows that opened during the session tell their own story: when spreads blow out to 9–12% between major venues, it signals fragmented liquidity and thin bid-side depth. Market makers were not stepping in aggressively to close those gaps. That is additional confirmation that the smart money posture overnight was defensive at best, outright bearish at worst. US traders waking up to this tape should treat the first hour of their session as information-gathering, not action-taking.
Let us start with Bitcoin, because Bitcoin is where the story begins and — based on overnight flows — possibly where it ends for the near term. BTC recorded $592.9 million in sell-side volume during the 00:00–08:00 UTC window against a buy volume of $0.0 million. The average buy ratio across BTC-tracked pairs landed at 9.5%. To contextualize that number: a neutral market runs around 50%. A mildly bearish session might print 35–40%. A 9.5% buy ratio means that for every ten dollars traded in BTC during Asian hours, roughly ninety-five cents came from the buy side. The rest was sellers finding whoever would take the other side, and finding fewer and fewer willing participants as the session wore on.
The most severe BTC readings came from two separate clusters. The first: 93% sell pressure ratio on $260.8 million volume spread across Binance Futures, Hyperliquid, and OKX Spot. The second: 86% sell ratio on $283.1 million across Hyperliquid and OKX Spot. Both readings are extreme. The concentration across Hyperliquid in both clusters is worth noting — perpetual futures platforms tend to amplify directional moves when funding rates and spot pressure align, which they clearly did here.
Ethereum's numbers are even more pronounced in absolute terms, if not percentage. ETH absorbed $808.5 million in sell volume against zero buy-side volume, with an average buy ratio of 10.9%. The two major ETH flow events mirrored BTC's structure: one at 86% sell pressure for $664.9 million on OKX and Hyperliquid combined, another at 93% sell ratio for $143.6 million across Hyperliquid and Coinbase. The Coinbase appearance in the second cluster is notable — US retail accounts are not the primary driver of Asian session flows, but institutional desks running 24-hour books absolutely are. Someone with a Coinbase Prime account was selling ETH at 93% ratio intensity at 4 AM UTC. That is not a retail trader who forgot to set an alarm.
The altcoin picture during this session was defined less by heroic individual movers and more by the absence of them. With zero pump events registered across 61 total tracked events, the narrative writes itself: alts were not the story last night. That said, the order flow data gives us texture on where pressure was concentrated and where it was not.
Solana emerged as the most notable altcoin in the flow data, though not for reasons bulls would celebrate. SOL registered a 92% sell pressure ratio on $131.4 million in volume across KuCoin, Hyperliquid, and Bybit Spot. The KuCoin presence here matters — KuCoin has historically skewed toward Asian retail and smaller market participants. When KuCoin is appearing in a 92% sell-side cluster alongside Hyperliquid perps, it suggests the selling was not purely institutional. Retail was participating in the distribution. That typically means the move has further to run before it exhausts itself, because retail exits tend to be slower and more spread out than institutional ones.
The arbitrage data offers a few glimpses into where specific coins were seeing unusual price dislocations. ATOM registered the widest spread of the session at 11.88% — the Cosmos ecosystem token was pricing at $1.9360 on one Coinbase venue and $2.1660 on another simultaneously. That gap does not exist in efficient markets. It exists when one side of a trade has dried up and market makers cannot profitably close the spread. BEAM, the gaming and privacy blockchain, showed an 11.49% gap between Bybit Spot at $0.0018 and Coinbase at $0.0020. TST printed an 11.29% spread. These are not coins at the top of most US traders' watchlists, but the spread data signals that thin-liquidity altcoins were experiencing acute liquidity fragmentation overnight — a condition that tends to resolve violently once a major venue moves.
From a regional preference standpoint: the KuCoin and Bybit dominance in sell-side flow, combined with the near-absence of buy-side activity on any venue, suggests that Asian retail — which typically drives speculative buying in gaming tokens, DeFi plays, and AI narratives — was either sitting on its hands or actively reducing exposure. AIGENSYN's appearance in the arb data (9.47% spread, KuCoin vs Bybit) and AI token's 9.14% gap (Binance vs Coinbase) both point to the AI-themed altcoin sector experiencing liquidity stress. These were some of the hotter narrative plays in recent weeks. The overnight data suggests the narrative is cooling.
Seventeen arbitrage windows opened during the Asian session, which is elevated. In a well-functioning market with healthy liquidity, these spreads would close within seconds. The fact that they persisted long enough to be captured and logged tells you something about the state of cross-exchange liquidity during these hours. When market makers pull back — either because volatility is uncertain or because they have directional inventory concerns — spreads widen and arb windows stay open longer. That is the environment we had overnight.
The practical takeaway for US traders: most of these windows will have partially or fully closed by the time New York desks come online. However, the coins involved — ATOM, AI, TST — are worth watching in the first trading hour. Arb windows that close via price convergence can move the lagging venue sharply. If Binance-side AI pricing comes up to meet Coinbase, that is a quick move on modest volume. Keep these tickers on a watchlist.
The order flow data from the Asian session is, frankly, one of the more striking whale-behavior profiles I have documented. Let me be direct about what the numbers suggest: someone — or more likely several coordinated someones — spent eight hours methodically selling Bitcoin and Ethereum into whatever buy-side liquidity existed, and the buy-side had essentially nothing to offer. This is not a crash scenario. Crashes involve panic on both sides. This is controlled, patient distribution.
Consider the math. BTC: $592.9 million sold, $0.0 million bought in tracked flow. ETH: $808.5 million sold, $0.0 million bought. Total sell pressure: $1,757.4 million. Total buy pressure: $1.7 million. These figures come from the highest-volume, highest-liquidity trading pairs on the planet's largest exchanges. The buy side did not just underperform. It was essentially absent. In eight hours of trading, smart money or at minimum large-money was a one-way seller and found no comparable buyer willing to take the other side at volume.
The venue distribution across the two BTC clusters is informative. Hyperliquid appeared in both — this is a perpetuals platform popular with sophisticated traders and algorithmic systems. OKX Spot appeared in both clusters as well. Binance Futures led the first cluster. What you are seeing is coordination not in the illegal sense, but in the market-structure sense: large players using the highest-liquidity derivatives venues alongside spot markets to distribute risk simultaneously. This is not a retail behavior pattern. This is institutional position reduction.
SOL's whale picture adds another data point: $131.4 million at 92% sell ratio across KuCoin, Hyperliquid, and Bybit Spot. The KuCoin component here is interesting because it muddies the pure-institutional read — KuCoin serves a different client profile than Hyperliquid. This suggests the SOL selling had both institutional and retail layers, which typically means the move is more durable than a pure institutional flush. Retail takes longer to fully exit a position. Their tail is longer.
One final observation on the whale data: the total buy pressure of $1.7 million against $1,757.4 million in sells is not zero, but it is close enough to zero to be remarkable. At some point during those eight hours, someone was buying. They just were not buying at meaningful scale relative to the sellers. Whether those buyers are long-term accumulators who will eventually support price, or simply market makers mechanically providing liquidity while hedging elsewhere, is unknowable from flow data alone. But they exist. The question for the US session is whether they scale up.
US traders are waking up to a market that has been sold aggressively for eight straight hours with essentially no buy-side resistance. What they do with that information in the first sixty minutes will set the tone for the remainder of the trading day. Here is how Boris would frame the key considerations.
The most important question at the US open is whether domestic institutional desks absorb the overnight selling or join it. Given that Coinbase appeared in at least one of the ETH sell clusters — suggesting some US institutional flow was active overnight on the sell side — the base case should lean toward continuation rather than reversal. Markets do not typically snap back aggressively from sessions with 93% sell ratios on half-a-billion dollars of volume. They tend to find a level, consolidate, then resolve in the direction of the prior trend.
The 17 open arbitrage windows represent the first technical event to watch. As US liquidity comes online, these spreads should begin to close. The direction of closure — does the low-priced venue come up, or does the high-priced venue come down — will be a real-time signal of US session demand. If ATOM's Coinbase spread closes with the $1.9360 price coming up toward $2.1660, that is bullish price discovery. If the $2.1660 comes down, bears remain in control.
SOL deserves specific attention given its size and venue profile. The $131.4 million in Asian-session SOL selling across KuCoin and Bybit will need to be digested by US buyers if the coin is to hold current levels. Watch Coinbase SOL volume in the first thirty minutes — if US buyers show up at scale, you have a potential reversal candidate. If volume is thin or sell-side, the Asian trend extends.
The AI token sector — flagged by both the AIGENSYN and AI arbitrage windows — is worth a close look from a narrative perspective. These coins have carried momentum recently, but overnight data suggests the Asian session saw liquidity fragmentation and potential distribution in this category. If the US session brings no fresh catalyst for AI-adjacent tokens, the overnight spread data may be a leading indicator of broader AI-narrative fading in the near term.
Key levels to watch: BTC's overnight flow profile, with $592.9 million in pure sell pressure and sub-10% buy ratios, creates a technical situation where any meaningful bid at current prices represents dip-buying against trend. That can work as a trade, but the overnight data does not give you structural support for it. ETH's situation is similar, possibly more acute given the larger absolute sell volume. US traders looking for entries should demand higher conviction — strong volume, clear reversal candles, or a macro catalyst — before stepping in front of what was a very organized Asian session downside push.
There you have it. Asia sold. America is about to find out if they want to buy. Based on what I have just walked you through, I would not be in a hurry to find out the answer by being on the wrong side of it. The data from the overnight session is about as clear as data gets: sellers were organized, buyers were absent, and the spread windows tell you liquidity was thin in all the wrong places. Be patient. Let the US open give you information before you give it your capital. The market will still be there in an hour. It has a habit of that.
— Boring Boris | Asian Session Wrap — May 16, 2026