Robot wars

Bot-led fake trades account for nearly half of crypto trading. Can more bots thwart them?

The Blockchain Transparency Institute has some alarming news: of the $13bn-worth of daily cryptocurrency trading volume, a whopping $6bn is fake. Bots, deployed by scammers, simultaneously buy and sell large volumes of cryptocurrency. This burst of activity creates the illusion of high demand, ramping up purchase prices unexpectedly for innocent buyers. 

Sometimes, the purchase doesn’t even have to be made in the first place. “Spoofy,” a mysterious bot lurking on the bitcoin network, regularly manipulates prices by placing tens of millions of dollars worth of prospective  orders, only to sharply revoke them. As the price subsequently spikes, Spoofy performs a bait-and-switch, selling a wad of bitcoin for the higher mark-up.

Bots on bots

The problem is that traders, when assessing the liquidity of an asset, cannot distinguish between real trades and robo-trades. All they have to go on is the asset’s nominal “trade volume,” which, as we’ve seen, can be easily tampered with. The exchanges themselves have little incentive to provide better transparency—the illusion of high liquidity raises commission fees and brings in revenue.

Bancor, the developer of the eponymous Bancor Protocol, wrote a blog post on Thursday claiming the solution is to introduce an array of “Smart Tokens” that create liquidity through a fixed algorithm, giving traders an accurate sense of demand. The Smart Token, Bancor says, facilitates “fair, stable and low-cost trading.” Of course, Bancor would say that, given that it invented the damn thing. Still, it’s pause for thought.

In essence, Smart Tokens (a misnomer, they’re more like Smart Contracts) create fully-automated markets that maintain currencies against “reserves” of themselves. A market with ten dogecoin and ten shitcoin, for instance, will have ten more of each token lying in reserve.

As regular trading takes place, the protocol adjusts these reserves so that they accurately reflect the demand of the tokens they represent. Buy three dogecoin with three shitcoin, and its reserve will decrease to seven to account for the main supply’s jump to thirteen. Simultaneously, the reserve for the shitcoin you just blew on the dogecoin purchase will grow, to account for the loss in its own main supply. In this way, the market makes itself , continually re-calibrating and adjusting price according to each trade.

Thus, Bancor rids us of human fallibility and creates a self-sustained liquidity that’s far harder to game.

But why should we trust Bancor’s bots over malevolent bots like Spoofy? “Bancor is not asking people to trust its algorithm, we’re asking people to trust on-chain math,” says Nate Hindman, Bancor’s communications director. That is, the calculations take place on the blockchain itself, not in the heads of human, fallible interlocutors. Moreover, Bancor is a foundation, so takes no profit.

This method could pick up, and fast. According to Hindman, there have already been “$1.5bn” in cumulative trades on Bancor’s platform since it launched last year, a figure that we’ve yet been able to corroborate. Yet if the figures hold true, the bots should soon be battling it out among themselves. Not quite the clash of chrome and steel we were hoping for, but it’s a start.


The Blockchain Transparency Institute has some alarming news: of the $13bn-worth of daily cryptocurrency trading volume, a whopping $6bn is fake. Bots, deployed by scammers, simultaneously buy and sell large volumes of cryptocurrency. This burst of activity creates the illusion of high demand, ramping up purchase prices unexpectedly for innocent buyers. 

Sometimes, the purchase doesn’t even have to be made in the first place. “Spoofy,” a mysterious bot lurking on the bitcoin network, regularly manipulates prices by placing tens of millions of dollars worth of prospective  orders, only to sharply revoke them. As the price subsequently spikes, Spoofy performs a bait-and-switch, selling a wad of bitcoin for the higher mark-up.

Bots on bots

The problem is that traders, when assessing the liquidity of an asset, cannot distinguish between real trades and robo-trades. All they have to go on is the asset’s nominal “trade volume,” which, as we’ve seen, can be easily tampered with. The exchanges themselves have little incentive to provide better transparency—the illusion of high liquidity raises commission fees and brings in revenue.

Bancor, the developer of the eponymous Bancor Protocol, wrote a blog post on Thursday claiming the solution is to introduce an array of “Smart Tokens” that create liquidity through a fixed algorithm, giving traders an accurate sense of demand. The Smart Token, Bancor says, facilitates “fair, stable and low-cost trading.” Of course, Bancor would say that, given that it invented the damn thing. Still, it’s pause for thought.

In essence, Smart Tokens (a misnomer, they’re more like Smart Contracts) create fully-automated markets that maintain currencies against “reserves” of themselves. A market with ten dogecoin and ten shitcoin, for instance, will have ten more of each token lying in reserve.

As regular trading takes place, the protocol adjusts these reserves so that they accurately reflect the demand of the tokens they represent. Buy three dogecoin with three shitcoin, and its reserve will decrease to seven to account for the main supply’s jump to thirteen. Simultaneously, the reserve for the shitcoin you just blew on the dogecoin purchase will grow, to account for the loss in its own main supply. In this way, the market makes itself , continually re-calibrating and adjusting price according to each trade.

Thus, Bancor rids us of human fallibility and creates a self-sustained liquidity that’s far harder to game.

But why should we trust Bancor’s bots over malevolent bots like Spoofy? “Bancor is not asking people to trust its algorithm, we’re asking people to trust on-chain math,” says Nate Hindman, Bancor’s communications director. That is, the calculations take place on the blockchain itself, not in the heads of human, fallible interlocutors. Moreover, Bancor is a foundation, so takes no profit.

This method could pick up, and fast. According to Hindman, there have already been “$1.5bn” in cumulative trades on Bancor’s platform since it launched last year, a figure that we’ve yet been able to corroborate. Yet if the figures hold true, the bots should soon be battling it out among themselves. Not quite the clash of chrome and steel we were hoping for, but it’s a start.


Get The Daily Debrief In Your Inbox

2019 © Decrypt Media, Inc. All Rights Reserved.

2019 © Decrypt Media, Inc. All Rights Reserved.

Get The Daily Debrief In Your Inbox