XRP Reserves on Binance and Upbit Drop 228 Million to Multi-Month Low
XRP/USDT
$908,896,932.23
$1.0972 / $1.0367
Change: $0.0605 (5.84%)
+0.0028%
Longs pay
AI SummaryAI
- Combined XRP reserves on Binance and Upbit fell roughly 228 million tokens to multi-month lows, per on-chain data.
- Binance's XRP balance dropped about 6% from 2.78 billion on May 12 to 2.61 billion, its lowest since March.
- Ripple CTO Emeritus David Schwartz proposed a transaction reservation scheme to counter XRP Ledger sandwich attacks.
- COINOTAG's composite engine scores $1.2157 resistance at 77/100 while derivatives show a 2.98 long/short ratio, 74.9% long.
This summary was AI-generated, AI-reviewed and published under COINOTAG editorial oversight.
XRP News
Exchange reserves of XRP fell sharply this week, with the combined holdings of Binance and Upbit dropping roughly 228 million tokens to their lowest level in several months, according to on-chain data. The decline stands out because it spans two of the deepest venues for the fourth-largest altcoin by market value, suggesting coordinated withdrawal rather than a single-desk event. Falling exchange balances are frequently read as supply leaving liquid order books, a setup that can tighten sell-side pressure. As of 13:00 UTC, XRP trades near $1.09, up more than 5% on the day even as the broader tape remains gripped by extreme fear.
The Binance leg of the outflow is the larger of the two. On-chain data shows the exchange's XRP balance slipped from about 2.78 billion tokens on May 12 to roughly 2.61 billion, a decline of close to 6% and the lowest reserve reading since March. A shrinking exchange float typically means holders are moving coins into self-custody or off-venue storage, reducing the immediately sellable supply. Whether that reflects conviction accumulation or simply a rotation of custody arrangements is not confirmed by the flow data alone. What the numbers do establish is that Binance, the single largest spot market for the token, has thinner XRP inventory than at any point in roughly four months.
South Korea's Upbit, the country's dominant exchange, contributed the second half of the withdrawal. On-chain data shows Upbit's XRP reserves eased from approximately 6.515 billion tokens on May 30 to 6.457 billion on July 2, a reduction of about 58 million XRP. The move is proportionally smaller than Binance's, but it matters because Korean retail flow has historically been an outsized driver of XRP demand. When both a global venue and the largest Korean venue bleed reserves at the same time, the signal is harder to dismiss as venue-specific plumbing. The token's Korean premium and local order-book depth remain worth watching as this drawdown develops.
The reserve shift landed against a recovering but nervous macro backdrop. Bitcoin reclaimed the $60,000 handle after briefly trading below $58,000, a bounce analysts tied to central-bank commentary at a European policy forum. Even so, structural demand looks soft: on-chain data shows retail Bitcoin inflows to Binance from wallets holding under one BTC fell to about 329 BTC a day, a record low for the exchange. That figure sits far beneath the roughly 2,690 BTC daily peaks seen at the 2021 cycle top. Thin retail participation across majors helps explain why altcoins like XRP are trading on flows and mechanics rather than broad speculative froth.
On the protocol side, Ripple CTO Emeritus David Schwartz spent the start of the week addressing front-running and transaction sandwich attacks on the XRP Ledger. Sandwich attacks, where an actor places orders around a pending trade to profit from the price move, are a known risk for on-chain payments and offer crossing. Schwartz floated a transaction reservation scheme that would reserve execution slots so a submitted transaction settles before any later transactions created after it are disclosed. The design aims to neutralize the ordering games that make automated market maker and offer-crossing flows exploitable, a concern that also shadows cross-chain mechanisms like the atomic swap.
The proposal drew a caveat from Schwartz himself. Asked whether transactions could be timestamped to the second and ordered chronologically, Ripple software engineer Mayukha Vadari pushed back, noting that nodes receive each transaction at slightly different times as it propagates across the peer network, so a strict clock-based order is not reliable. Schwartz said the closest workable alternative is consensus-based ordering, with validators voting on transaction sequence as part of the consensus round. The catch he flagged is meaningful: baking ordering into consensus would materially slow the process, because validators would need to reach agreement on far more data. It is a genuine engineering trade-off between fairness and throughput, not a settled upgrade.
Reading our own signals, COINOTAG's proprietary 42-indicator composite S/R scoring engine rates the $1.2157 resistance at 77/100, the strongest overhead level, driven by the confluence of the volume point of control and the 50-period moving average; the $1.1367 band scores 69/100 on the Fibonacci 0.236 retracement and R1 pivot. On the downside, the $1.0708 support carries a 71/100 score from Fibo 0.114 and the Ichimoku Tenkan. Derivatives lean crowded-long: our aggregate open interest reads $650.9 million with a 2.98 long/short account ratio (74.9% long) and a barely positive 0.0028% funding rate, while the Fear & Greed Index sits at 19 (Extreme Fear). A daily close above $1.14 opens the $1.22 test; losing $1.07 with 74.9% of accounts long risks a long-squeeze back toward $1.03. RSI at 44.34 in a downtrend keeps the bounce unconfirmed until buyers reclaim resistance.
COINOTAG does not provide financial advisory services. This content is for informational purposes only and should not be considered investment advice. Cryptocurrency investments involve high risk.
Add COINOTAG as a Preferred Source
Add COINOTAG to your preferred sources in Google News and Search to see our coverage first.
Add on GoogleRelated Tags
AI-generated, AI-reviewed, under COINOTAG editorial oversight.
