Solana is one of the fastest, cheapest blockchains for trading. High throughput, near-zero fees, dozens of active DEXs. On paper, it looks like a trader's paradise.
But there's a number that should make you pause before applying your favorite strategy to any Solana token:
In 2025, arbitrage bots accounted for roughly 50% of Solana's average DEX volume.
Over 90 million arbitrage transactions were recorded in a single year — generating $142.8 million in combined profits at an average of just $1.58 per trade.
Half of the volume you see on a Solana chart isn't humans making trading decisions. It's automated systems correcting price discrepancies between DEXs — often within the same block. And that has consequences for almost every indicator you might use.
How Solana arbitrage actually works
When a large trade shifts a token's price on one DEX, it temporarily creates a price gap versus other DEXs. Raydium shows $1.00. Orca shows $1.02. That gap is worth capturing.
Arbitrage bots — written in Rust, co-located near validators, submitting transactions via Jito bundles — detect this gap and close it within milliseconds. The transaction is atomic: both legs execute simultaneously or neither does. No partial fills, no stuck positions.
The bot captures the spread. The price gap closes. The volume is recorded on both sides of the trade.
This happens thousands of times per day, on every active token, across every DEX. Solana's speed and low fees make it uniquely efficient for this kind of high-frequency, low-margin operation. The result: a baseline layer of bot-generated volume running permanently underneath all visible price action.
Why this makes volume-based indicators unreliable
Volume-based indicators assume that volume reflects genuine human conviction — that a high-volume candle means real buyers or sellers were active, and that you can read intent from it.
That assumption holds reasonably well on Bitcoin (primarily traded on centralized exchanges) or traditional stocks (with regulated market makers). It breaks down on Solana DEX tokens.
OBV (On-Balance Volume) adds up volume on up-days and subtracts on down-days to track cumulative flow. When a large trade triggers cascading arbitrage across five DEXs simultaneously, OBV records all of that volume — but it carries zero information about human sentiment. The "accumulation signal" you think you're reading might just be bots correcting a price imbalance.
VWAP and Volume Profile face the same problem. If 50% of volume is arbitrage, your volume-weighted average price is meaningfully distorted. Support and resistance levels derived from volume profiles may not reflect where real market participants actually traded.
Volume spikes — often used as confirmation signals — lose their meaning. A 3x volume spike on a Solana token might indicate institutional accumulation. Or it might mean a large trade temporarily widened a spread and triggered a wave of arbitrage responses.
There is no easy way to separate bot volume from human volume on a DEX. The data you receive is the combined total.
What is less affected
Price-based indicators (RSI, moving averages, Golden Cross, MACD on price) use only the price — which reflects the actual consensus between buyers and sellers regardless of how much bot activity surrounds the trades. These are more reliable on Solana than volume-based equivalents.
Higher timeframes help. A weekly candle on SOLUSDT averages out a lot of the arbitrage noise. The net direction of weekly price movement reflects real market forces more accurately than a 5-minute candle where bots may have executed thousands of corrections.
Relative volume — comparing today's volume to its own 30-day moving average — is more meaningful than absolute volume figures, because at least it captures unusual deviations from the established noise baseline. If volume is 5x the average, something genuinely unusual is happening even if half of that volume is still bots.
Liquidity-adjusted analysis: for smaller Solana tokens with thin liquidity, even price-based signals become unreliable — a single large trade can move price dramatically with no real market consensus behind it. Stick to tokens with deep, consistent liquidity.
The broader lesson: always question your data source
Before applying any indicator to any asset, there are two questions worth asking:
1. What percentage of this asset's volume reflects real human decision-making?
Bitcoin on Binance or Coinbase: high.
Apple stock on NYSE: high.
Solana token on a DEX: often below 50%.
2. Does the indicator you're using depend on that volume being real?
If yes — and you're on a Solana DEX — you're building your analysis on a compromised foundation.
This isn't unique to Solana. The same issue exists to varying degrees on any DEX with significant MEV activity. But Solana's architecture makes it particularly efficient for arbitrage, which means the problem is more concentrated here than almost anywhere else.
The indicator isn't broken. The data it's reading is.
Practical takeaway
If you're backtesting strategies on Solana tokens:
Prefer price-based indicators over volume-based ones
Use weekly or daily timeframes rather than intraday
Be skeptical of volume confirmation signals
OBV-based strategies that work well on Bitcoin will likely underperform on Solana tokens — not because the logic is wrong, but because the input data is fundamentally different
The strategy needs to match the market structure. On Solana, that structure includes a permanent, invisible layer of automated trading that never shows up in your backtest assumptions.