Business Model

Fees and Trade Duration

One of the most common experiences for an inexperienced trader at a cryptocurrency exchange is making a trade that appears to be profitable, but which in fact is not when the fees are taken into account.

In a random market movement, it takes on average four times longer for the price to go twice as far.

Although exchange fees might appear to be small on absolute numbers, their effect is very significant, especially for users trading with high leverage.

For instance, the Bybit charges a taker fee of 0.06%. At the same time, users are allowed to trade at up to 100x leverage. This means that a player opening and then closing a position at the same price, executing as a taker in both cases, actually pays 12% of its margin as fees for this trade.

Regardless of the leverage, a user only begins to win at a trade when the price moves at least 0.12% in the right direction. Assuming a 2.0% daily volatility for bitcoin that is constant over the day, the bitcoin price is expected to move approximately 0.12% one way or another every 5 minutes. This means that it is quite likely that, after 5 minutes, your trade is still a loss even if you pick the direction right.

If you want your trades to give you a profit that is at least 5x or 10x the fee you pay when you pick the direction right, you need your trades to last for around 2 or 8 hours, respectively. High fixed fees therefore encourage users not to do short-term trading.

Players do react to these incentives, and as a result, most players in the cryptocurrency exchanges today do not perform short-term trades lasting less than an hour. This despite the fact that they often do such trades at the beginning, before they understand how the fees work, and despite the fact that trades of such duration are common in traditional futures markets in which fees are significantly lower.

Now consider that 2 to 8 hours is a long time for an user to be logged in and paying attention to price movements, only waiting to see if the trade will pay off. Most traders can therefore be expected to wander the internet or do some other stuff, checking only periodically to see how the trade is going.

Our Proposal for Fees

We are exploring a completely novel fee structure: charging fees only on profitable trades. For profitable trades, fees are set to 0.1% of volume, or to 10% of trade profits, whatever is is smaller.

The idea is to allow users to make small trades without having to worry whether their gains are enough to offset the fees that they need to pay. Even very small trades, lasting a few minutes or even a few seconds could be profitable for the user. As a result, a user would be able to make dozens of trades in a single trading session, which hopefully will be much more interesting for him and increase retention.

It is important to point out that, under this structure, fees can be dramatically lower for short trades. This, however, is actually a good thing for us! As we will see later, encouraging small trades, even when accompanied by proportionally smaller fees, may generate an increase in volume that actually increases user monetization. And because other exchanges have to pay rebates to marker makers in order to have low spreads, they cannot afford to set very low fees for very short-term trades. If they decide not to pay maker rebates anymore, then their spreads will be much higher, which discourages small, short-time trades in the same way that high fees do. In either way, our model will be very difficult to duplicate.

We believe a fee structure proportional to trade profits, by encouraging small trades, is ideal for many users that are looking for short-term speculation or entertainment. It does, however, offer problems for other types of users. Professional and long-term traders might be unwilling to pay a significant portion of their profits for a trade, especially when they can pay a fixed fee in a competing exchange. Hedgers, which negotiate futures to protect themselves against losses in the value of their assets elsewhere, also generally perform longer-term trades and would rather pay a fixed fee. Finally, high-frequency traders, including snipers and market-makers, may find that optimizing for such a fee structure adds complexity to their bots.

That is why we decided to cap fees on profitable trades to a maximum of 0.10% of the transaction. This makes sure that even more sophisticated trades doing long-term speculation or hedging will be pleased with our fee structure. Adding such a cap means that fees paid on our exchange will always be strictly better than those at any other exchange.

A Simplified Monetization Model

We believe that trading volumes may increase dramatically when users are encouraged to make small trades instead of big ones, and as a result that monetization may increase even if the average transaction fee is smaller.

Let us create a model in which users start with a balance of 100$, and leave the exchange if their balance goes to zero (bankruptcy) or if their balance doubles (cashing out). Of course, exchanges want users to deposit more and refrain from cashing out in order to continue trading! However, it is important for our model to incorporate both churn scenarios, as they do occur in practice.

Further, let us consider in our model that trades are like bets that have a 50% chance of paying off. We then have two scenarios: one with big (longer) trades and another with small (shorter) trades, which we model respectively as bets in which $50 and $10 are at stake.

In the scenario for big trades, what we see is that users either go bankrupt or cash out after only a few trades. The average number of trades they make before leaving is only four.

In comparison, in the scenario for short (small) trades, they make many more trades. They are risking 5 times less in each trade, but they end up trading 25 times more! Now the average user will make 100 trades before leaving.

We can model a fee (10% of profits from a trade) by considering that the bets have a chance of paying off that is slightly less than 50%. This changes the simulation only slightly; users make an average of 3.99 long trades or 91.82 short trades before leaving.

The results show a dramatic increase in monetization: an average user that makes long trades will pay $5.25 on fees before leaving, compared to $24.15 on fees for an user that makes short ones.

While this is of course a very simplified model, we do believe that it is possible to significantly increase user monetization in an exchange by encouraging them to make many small trades at a reduced fee.

Our Business Model

As a result of keeping 100% of user funds in a smart contract, Quiver is prevented from investing, lending or otherwise “gaining yield” from stored user funds.

Our business model is therefore completely dependent on trading fees, as exchanges are supposed to!

Some may question: is it viable for Quiver to offer the lowest effective fees of any futures exchanges? The answer is yes, and comes from (1) our special matching technology, that allows us to have liquidity with no need for continuous subsidies, and from (2) an innovative fee structure that encourages users to trade much more, increasing monetization.