Papertrade, an on-chain derivatives trading platform, is one of the hottest projects recently.
According to DefiLlama data, Papertrade's protocol revenue over the past 7 days reached $42.24 million, ranking third among all protocols by revenue, behind only Tether and Circle. This even surpassed Pump.fun's $16.06 million and Hyperliquid's $14.56 million.
Image source: https://defillama.com/protocol/papertrade
A trading platform that launched not long ago, relying on a very aggressive incentive model, has in just a few days generated revenue exceeding that of many leading protocols that have been operating maturely for years.
But behind Papertrade's explosive growth is the hard fact that traders have lost money one trade at a time.
On October 13, according to Lookonchain monitoring, a trader on Papertrade accumulated a notional trading volume of as high as $2.84 billion, with trading losses of about $1.71 million. Although he received about $1.07 million in USDC staking rewards and 136 million PAPER tokens, after deducting staking rewards, his net cash loss still reached about $637,000.
These two seemingly contradictory phenomena stem from Papertrade's very aggressive and special economic model, which directly links traders' actual losses to PAPER token issuance, and then through a staking dividend mechanism allows token holders to share in the revenue generated by the protocol. In other words, participants can obtain the right to future protocol revenue distribution by bearing real trading losses.
So why are some people willing to actively bear trading losses to obtain PAPER? Why can some early whales obtain substantial returns, while later participants may lose more the more they trade? How exactly is Papertrade's astonishing protocol revenue generated, and can this model be sustained over the long term?
This article will start with Papertrade's trading mechanism, token economic model, and the interest distribution relationships among participants, attempting to break down this special economic model and analyze the operating logic and economic game behind it.
Papertrade is a synthetic perpetual contract trading platform built on HyperEVM within the Hyperliquid ecosystem. It officially opened trading on October 10, 2026. Initially it mainly supports two assets, BTC and ETH, with leverage of up to 1000x.
From the product form perspective, it is actually not much different from Perp DEXs such as Hyperliquid and Aster. Users can likewise go long or short and profit by judging the direction of asset price movements.
But in terms of underlying operating mechanism, Papertrade adopts a completely different design.
The core of a traditional order book exchange is matching buy and sell orders.
For example, if you open a long position on BTC, the exchange needs to find a counterparty willing to take the corresponding short exposure.
Papertrade does not rely on its own order book for matching. It adopts a liquidity pool trading mechanism similar to GMX's early model.
The user's counterparty is the protocol's own liquidity pool, and all profits and losses generated by trades must be settled against this liquidity pool.
For example, if you open a BTC long position worth $100,000 on Papertrade: if BTC rises, your position generates a profit, and the protocol liquidity pool needs to pay you the corresponding profit; conversely, if BTC falls, your position generates a loss, and the funds flow into the protocol liquidity pool.
Papertrade itself directly reads the midpoint of the best bid and ask quotes in Hyperliquid's order book as the trading reference price, so it actually does not need to execute a corresponding perpetual contract trade on Hyperliquid.
This enables it to provide a trading experience with zero order book slippage and no traditional funding rates.
However, Papertrade is not entirely without trading costs.
It introduces a special asymmetric profit and loss mechanism, which they call Asymmetric PnL. That is, when traders are profitable, they must bear a certain percentage deduction from profits, while when they are at a loss, they do not receive an equivalent reduction.
For example, if a trade originally generates $100 in profit, after asymmetric PnL adjustment and related settlement fees, the trader may actually receive noticeably less than $100 in profit.
But if the trade loses $100, the trader still needs to bear the corresponding loss. Of course, the specific deduction ratio is determined by the magnitude of price changes, the trading asset, and protocol parameters.
In other words, even if the profits and losses of longs and shorts in terms of price direction can offset each other, traders as a whole may still incur net attrition due to profit deductions, and this attrition also constitutes an important source of income for the protocol liquidity pool.
PAPER issuance mechanism: trading losses can also earn tokens
PAPER is Papertrade's native token, and it adopts a so-called fair launch model, meaning the initial supply is zero, with no team pre-mining, VC allocation, or traditional airdrop share.
The main way to obtain PAPER tokens is for users to trade on the platform and generate actual losses that meet the minting conditions.
According to the project's publicly disclosed issuance rules, when the LP scale tracked by the protocol is below $2 million, every $1 of qualifying loss can mint 100 PAPER.
This means that in the initial stage, if you lose $100 in trading, you can theoretically obtain about 10,000 PAPER. If you lose $1,000, you can obtain about 100,000 PAPER.
The issuance efficiency of PAPER tokens is not always fixed; it adopts a dynamic decay curve.
When the LP scale exceeds $2 million, as the protocol's cumulative revenue grows, the number of PAPER that can be minted per dollar of qualifying loss will gradually decrease.
Its issuance curve can be simplified as:
Mint quantity per dollar of loss = 100 × [120 million / (120 million + H)]²
Among them, H is the cumulative growth indicator recorded by the protocol issuance curve, not the current LP balance. For example, when H reaches $120 million, the PAPER obtained per dollar of loss will drop to 25, equivalent to only one quarter of the initial issuance efficiency.
Such a mechanism actually creates a very obvious first-mover advantage.
For the same $10,000 trading loss, early participants and later participants may ultimately receive PAPER quantities that differ by several times.
So the question is, since PAPER can only be obtained through losses, why are so many people still willing to participate?
The answer lies in PAPER's revenue distribution mechanism.
Papertrade adopts a special bootstrap liquidity model.
Unlike traditional DeFi protocols that rely on external LP deposits to establish liquidity, Papertrade's protocol trading liquidity pool starts from zero and gradually accumulates funds mainly through traders' actual losses.
If traders are profitable and the liquidity pool temporarily does not have enough funds to pay profits, the protocol also has a queued payment mechanism, waiting for subsequent funds to flow in before settlement.
The PAPER that traders obtain through losses can be staked into the protocol to share in future USDC revenue.
Currently, staking yield mainly comes from two parts.
The first is trading settlement fees. The protocol charges a 2% fee on qualifying realized profit and loss and liquidation losses, half of which is distributed to PAPER stakers and the other half to the development fee recipient.
The second is excess returns from the LP liquidity pool. When the LP funds tracked by the protocol exceed the $5 million dividend threshold, qualifying excess funds can be further distributed to PAPER stakers.
Therefore, PAPER actually has the function of sharing in the protocol's future cash flows.
We can summarize this set of economic relationships as:
Traders generate actual losses → the protocol LP receives funds → loss-making traders obtain PAPER → PAPER holders stake tokens → share in the distributable revenue generated by the protocol in the future.
From the perspective of the economic model, Papertrade actually converts part of the real losses borne by traders into a cost of obtaining the right to distribute the protocol's future cash flows.In traditional Perp DEX trading incentive campaigns, long-short hedging is a fairly common way to farm volume. For example, we can establish equal-sized BTC long and short positions respectively, use the profit and loss of the two positions to offset each other, minimize directional risk as much as possible, and then obtain points or token rewards through continuous trading.
But on Papertrade, this approach does not seem to enable lossless volume farming.
As mentioned earlier, Papertrade adopts an asymmetric PnL mechanism, where the profitable side has to bear a profit deduction, while the losing side has to bear the full trading loss.
We can understand this through a simple example.
Suppose a trader simultaneously opens two BTC positions, one long and one short, each with a notional value of $100,000, using low enough leverage to avoid liquidation in this example.
When the BTC price rises by 0.1%:
The long position has a theoretical profit of $100.
The short position has a theoretical loss of $100.
Under ideal conditions without considering trading fees, the PnL of the two positions exactly offsets, and the trader neither makes money nor loses money.
But on Papertrade, the situation is different.
According to the BTC asymmetric PnL parameters announced at the project's launch, after accounting for the 0.2 basis point adjustment on the profitable side, the profit discount, and the 2% profit settlement fee, the long position that originally had a profit of $100 can actually receive only about $72.4.
Meanwhile, the losing short position still has to bear the full loss of $100.
In the end, the two positions combined lose about $27.6.
That is to say, without considering other execution costs, even if BTC price directional risk is fully offset, the trader still has to bear an actual loss of about $27.6.
On the other hand, Papertrade's profit deduction ratio is actually not fixed, but related to the magnitude of price changes.
When the price changes only very slightly, the profit deduction ratio tends to be higher; as the profit magnitude expands, the deduction ratio gradually decreases, but does not completely disappear.
This means that through its asymmetric PnL mechanism, Papertrade effectively imposes a cost on long-short wash trading that cannot be simply eliminated through hedging.
Although the losing position also receives corresponding PAPER tokens, this is essentially converting the actual loss into the cost of acquiring PAPER.
1000x leverage further amplifies traders' loss risk
In addition to the asymmetric PnL mechanism, Papertrade also offers up to 1000x leverage, which partly explains why the platform can rapidly accumulate such a huge notional trading volume.
Suppose you use $100 in margin to open a BTC position worth $100,000 with 1000x leverage.
Theoretically, a BTC price move of just 0.1% in the unfavorable direction is enough to wipe out the entire initial margin.
And according to Papertrade's liquidation mechanism, a 1000x leveraged position may be liquidated after an unfavorable price move of about 0.05%.
This means that even a very ordinary short-term BTC price fluctuation is enough to cause high-leverage traders to lose their entire margin.
According to on-chain statistics by TradingBeats during Papertrade's early launch period, among 5,836 addresses that used 1000x leverage, 4,395 addresses experienced at least one liquidation, accounting for 75.3%.
During the same period, 1000x leveraged positions were liquidated 12,657 times in total, involving about $20.08 million in isolated margin losses.
Although these data cannot directly prove that all participants ultimately lost money, I think they are enough to reflect the risks faced by high-leverage traders.
And for participants whose goal is to farm PAPER, there is another contradiction here.
Increasing leverage can create huge notional trading volume with less margin, but it cannot reduce the actual loss caused by asymmetric PnL, and may instead make positions more likely to be liquidated.
In addition, although Papertrade does not have traditional order book slippage and funding rates, on-chain transactions still involve gas costs, transaction confirmation delays, and the risk that long and short positions cannot be executed fully synchronously.
Therefore, when high leverage, high-frequency trading, and the asymmetric PnL mechanism are combined, strategies that rely purely on mechanical volume farming to obtain PAPER can easily continue to generate real losses.
The actual cost of acquiring PAPER is the key to determining profitability
Volume farming on Papertrade itself has a cost. The reason some whales are still willing to spend hundreds of thousands or even millions of dollars to acquire PAPER is that they are not entirely pursuing profit from trading itself, but are calculating the cost of acquiring PAPER and whether future staking returns can cover these costs.
Here it is necessary to establish a return calculation method different from traditional trading.
Comprehensive return from participating in Papertrade = realized trading PnL + USDC staking returns already obtained + the value of PAPER that can actually be realized − other costs.
We still take the whale mentioned in the introduction as an example.
According to Lookonchain's monitoring data on October 13, this trader generated a cumulative notional trading volume of $2.84 billion, incurred trading losses of about $1.71 million, and at the same time obtained 136 million PAPER and about $1.07 million in USDC staking rewards.
Of this, about $943,000 in rewards had been claimed, and about $130,000 remained unclaimed.
If claimed and pending rewards are counted together, his net loss is about $637,000.
But in fact, there is still a very important potential asset that has not been included, which is the 136 million PAPER he holds.
If PAPER can be freely traded in the future, then without considering market impact, additional fees, and subsequent distributions, as long as these tokens can all be realized at an average price of about $0.00468, there is an opportunity to cover the above net loss of $637,000.
That is to say, he is actually using part of the trading losses that have already occurred in exchange for a large amount of PAPER whose final value has not yet been determined, as well as the corresponding future return distribution rights.
For ordinary traders, the most direct goal is to make money by judging the price direction of BTC and ETH; but for some participants whose goal is to obtain PAPER, trading losses themselves have become a kind of input for acquiring tokens. They actually need to calculate how much actual cost is required for each PAPER obtained, and how much return can be recovered from each token in the future.
However, as PAPER issuance efficiency continues to change, participants entering the market at different times may need to bear completely different costs to obtain the same number of tokens.
Papertrade's token issuance mechanism inherently has a certain first-mover advantage.
As mentioned earlier, when the LP scale tracked by the protocol is below $2 million, every $1 of qualifying trading loss generated can mint 100 PAPER.
But as the protocol's cumulative revenue grows, token issuance efficiency will gradually decline according to a predetermined curve.
We can make a simple comparison.
Suppose two traders enter Papertrade at different stages, each bearing $10,000 in qualifying trading losses.
Early participant A, under the initial minting rate, can obtain 1 million PAPER.
For later participant B, if the H metric in the issuance curve has already reached $120 million when they enter, and each dollar of loss can only obtain 25 PAPER, then with the same $10,000 loss, they can ultimately obtain only 250,000 PAPER.
The two bear exactly the same loss, but receive a fourfold difference in the number of tokens.
But in the normal issuance decay stage, early participants may indeed be able to obtain large amounts of PAPER at a lower cost.
PAPER itself has the function of participating in the protocol's revenue distribution, which means that early participants, by obtaining large amounts of PAPER through lower trading losses, actually obtain a larger share of future revenue distribution rights in advance.
And later entrants who want to obtain revenue rights of the same scale may need to bear higher actual losses.
In addition to acquisition cost, Papertrade also faces another problem, which is the continuous issuance of PAPER.PAPER does not have a fixed, hard supply cap in the traditional sense; as long as qualifying trading losses continue to exist, the protocol may continue to mint new tokens.
However, the returns shared by PAPER stakers do not increase proportionally with the growth in token supply.
Suppose that on a given day, the protocol can distribute $100,000 in USDC to PAPER stakers.
When the total network staking amount is 5 billion PAPER, a participant holding 1 million PAPER can receive $20.
But if the total network staking amount grows to 10 billion tokens while the distributable returns for that day remain only $100,000, then the same holder of 1 million PAPER can receive only $10.
In other words, the continuous issuance of PAPER dilutes existing holders' share of the total staked amount, and the growth of protocol revenue may not necessarily keep pace with this dilution.
This is especially unfavorable for later participants.
They may need to pay a higher cost to obtain the same amount of PAPER, but how much return these tokens can generate in the future still depends on the overall revenue scale of the protocol and how many new tokens subsequently participate in distribution.
Therefore, Papertrade actually has two competitions taking place simultaneously:
One is the competition among participants for PAPER issuance share, and the other is the competition among PAPER holders for the protocol's future cash flows.
Participants who accumulated a large amount of PAPER at a lower cost early on are generally better able to withstand subsequent fluctuations in returns.
Looking at real data, to whom did Papertrade's early returns actually flow?
According to on-chain retrospective data cited by Odaily on October 13, in the first 8 hours after Papertrade launched, approximately $23.9 million in returns were distributed cumulatively to PAPER stakers.
Image source: https://www.lookonchain.com/feeds/76698
In the following 37 hours, stakers received only about $2.4 million in returns.
Converted into average hourly distributions, the first 8 hours were about $2.99 million per hour, while the subsequent 37 hours were only about $65,000 per hour.
In just a few dozen hours, the average hourly return distribution scale differed by about 46 times.
This at least shows that in the initial period after Papertrade launched, there was a very brief window with highly concentrated returns.
Also within its statistical period, traders accumulated net losses of about $32.4 million, of which about $26.3 million flowed to PAPER stakers, $4.3 million remained in the LP pool, and the remainder belonged to developer fees and other items.
This means that competition among early participants actually revolved around two timing points:
First, whether one could obtain a large amount of PAPER earlier at a higher issuance efficiency.
Second, whether one could complete staking earlier and participate in distributing the huge returns in the initial period after the protocol launched.
We can also look at a real case.
On October 12, according to Lookonchain monitoring, an address starting with 0x35b2 completed 1,012 transactions on Papertrade, cumulatively generating about $10.1 billion in notional trading volume.
Image source: https://lookonchain.com/feeds/76439
The address incurred trading losses of about $979,000, but through trading obtained about 252 million PAPER and also cumulatively received about $997,000 in USDC staking returns.
Calculating only trading losses and USDC staking rewards, the address achieved a net cash gain of about $17,800.
In other words, through staking rewards, this account basically covered the trading losses previously incurred to obtain PAPER, while still holding about 252 million PAPER.
For such early participants, they initially needed to bear real losses through trading to obtain PAPER, but once the staking returns already obtained covered the previous input, the remaining economic value of this batch of PAPER no longer needed to bear historical costs of the same scale.
If the protocol can continue to generate distributable returns in the future, then this portion of PAPER can continue to bring cash flow.
Conversely, if later participants need to pay hundreds of thousands of dollars to obtain the same amount of PAPER, they will need to rely on future dividends to gradually recover costs, and they will also face the risk of return dilution from newly issued tokens.
Although these two types of participants hold the same tokens, the economic risks they actually bear are completely different.
Looking at the entire model, Papertrade's wealth distribution actually depends on when they entered the market, the actual cost of obtaining PAPER, and how large a share they occupied in the protocol's return distribution.
According to DefiLlama data on October 15, Papertrade's protocol revenue in the past 7 days reached $42.24 million, of which PAPER holders' revenue reached $39.37 million, accounting for about 93% of protocol revenue.
Image source: https://defillama.com/protocol/papertrade?fees=true&holdersRevenue=true&tvl=false
This ratio is quite astonishing.
But if we further break down DefiLlama's revenue statistics method, we will find that Papertrade's revenue structure is significantly different from that of traditional Perp DEXs.
Taking the cumulative protocol revenue of about $48.04 million at that time as an example:
Trading settlement fees were about $5.69 million.
Liquidation fees were about $1.02 million.
LP excess return distribution was about $41.33 million.
In other words, about 86% of cumulative revenue came from excess returns distributed from the LP pool to PAPER stakers, rather than fees charged on notional trading volume in the traditional sense.
For trading platforms such as Hyperliquid that use an order book matching model, profits and losses between traders are mainly transferred between counterparties, and the platform earns revenue through methods such as fees.
In contrast, Papertrade's protocol pool directly assumes the role of counterparty. When traders profit, it means the pool needs to pay profits; when traders lose, it means funds flow into the pool.
Under the influence of the asymmetric profit-and-loss mechanism, the protocol can also obtain a certain statistical advantage through profit deductions.
When the funds accumulated in the pool exceed the specified dividend threshold, the excess portion can be distributed to PAPER stakers.
Therefore, an important source of Papertrade's current revenue is actually the net returns obtained by the pool after profit-and-loss settlement between traders and the protocol pool and distributed to stakers.
This type of return is real USDC cash flow, and it is clearly different in nature from the operating revenue formed by traditional exchanges relying on stable fees.
As long as traders as a whole generate relatively large profits, the LP pool may likewise experience fund outflows.
In other words, the huge revenue currently created by Papertrade actually cannot be directly projected long term at the existing level.
What happens to this model when everyone is wash trading for PAPER?
From an economic perspective, Papertrade's model actually contains a rather peculiar contradiction.
On the one hand, the protocol hopes to use PAPER incentives to attract more people to participate in trading.
On the other hand, PAPER issuance is tied to traders' actual losses, and part of traders' losses is precisely an important source supporting LP returns and staking dividends.
This means that the trading losses participants actively bear in order to obtain tokens, while driving protocol revenue growth, are also continuously creating new rights to return distribution.
We can assume an extreme scenario.
Suppose the vast majority of participants on Papertrade are not traders who genuinely hope to make money by judging market movements, but rather high-frequency wash traders seeking to obtain PAPER.
In the trading process, they continuously bear losses, obtain PAPER, and then share protocol returns through staking.
Then, from the perspective of the entire system, part of the losses borne by traders will enter the LP or developer fee recipients, while another part may be redistributed to PAPER holders through dividends.
If these traders and stakers overlap heavily, then to a large extent the entire system becomes a redistribution of funds among participants.For individual participants, it may still be possible to achieve profitability through earlier entry, lower token acquisition costs, and a higher staking share.
But overall, participants cannot rely solely on wash trading against each other to allow everyone to obtain cash returns exceeding their actual input.
Because internal trading itself does not create new funds out of thin air, and the continuous issuance of tokens does not automatically increase the USDC the protocol can distribute.
This model may form a mutually reinforcing cycle.
In the early stage, high staking dividends attract traders to participate, and traders bear losses in order to obtain PAPER, thereby helping the protocol accumulate funds and generate returns.
But as token issuance efficiency declines and the total staking amount increases, the payback period for new participants may gradually lengthen.
When more and more people find that the expected return from farming PAPER is insufficient to cover costs, trading activity may decrease.
A decrease in trading activity may in turn lead to a decline in the trading fees and net returns generated by the protocol, further weakening the dividend appeal of PAPER.
In this way, the incentive mechanism that initially drove the protocol's rapid growth may face the problem of continuously diminishing marginal effects.
Of course, if it can attract enough users with real trading demand, even if these users do not primarily aim for PAPER incentives, the protocol can still generate returns through the trade settlement mechanism.
The key to long-term sustainability
From the current economic model, I believe the long-term development of Papertrade depends on at least three factors.
1/ Whether it can form real trading demand that does not rely on PAPER incentives.
Papertrade offers up to 1000x leverage, zero order book slippage, and no traditional funding rates, and indeed has certain product differentiation.
But whether these advantages are enough to attract long-term traders still needs time to verify.
In particular, the asymmetric profit and loss mechanism essentially increases traders' cost of profitability. Without PAPER incentives, are users still willing to bear this cost and trade on Papertrade over the long term?
This will directly determine how much of the protocol's future revenue comes from sustainable real trading activity.
2/ The stability of the LP fund pool and its ability to pay out profits.
Papertrade's fund pool mainly accumulates funds through trader losses while also bearing the payout responsibility for profitable traders.
Once the fund pool is insufficient to pay profitable positions, the protocol may place unpaid profits into a waiting queue and gradually complete payments relying on subsequent capital inflows.
This means that the protocol's actual payout capacity is directly linked to future trading results and capital flows.
It is worth noting that the project team has announced plans to raise the LP excess return dividend threshold from $5 million to $10 million on October 17.
This means the protocol needs to retain a larger fund pool, and only the portion exceeding the new threshold can continue to be distributed to PAPER stakers according to the rules.
From a risk control perspective, retaining more funds helps strengthen the LP's risk tolerance; but for PAPER holders, it also means the trigger condition for excess return dividends has been raised.
The protocol needs to find a balance between ensuring traders' payout capacity and maintaining token holders' returns.
In addition, there was a price manipulation incident worth watching during Papertrade's early launch period.
On October 10, a trader profited by influencing the reference quote of ETH on Hyperliquid and using Papertrade's high-leverage synthetic positions. According to The Block's on-chain analysis, the relevant wallet once net withdrew about $1.28 million from Papertrade.
Image source: https://www.theblock.co/news/
After the incident, the project team stated that it had begun tightening position limits to reduce the profit space for similar operations.
This also shows that when the protocol itself bears counterparty risk, in addition to normal market fluctuations, it also needs to face risks such as price manipulation, extreme market conditions, and concentrated trader profits.
3/ Whether PAPER's issuance speed can remain balanced with actual return growth.
As more and more PAPER is minted, the returns created by the protocol in the future will need to be distributed among more tokens.
If the growth rate of distributable returns remains below the growth rate of staked token quantity for a long time, then the dividend capacity corresponding to each unit of PAPER will decline.
Conversely, if the protocol can continue to expand the scale of real trading and increase long-term distributable returns, even if PAPER supply increases, the return capacity per token may not necessarily decline.
Therefore, to judge PAPER's long-term value, more attention should be paid to the actual returns corresponding to each unit of token, and whether such returns can continue to grow.
So Papertrade's model actually has the possibility of working.
By linking trading losses to token issuance, it quickly accumulates liquidity and attracts users in the early stage, and allows participants to directly share in protocol revenue. This design is indeed somewhat innovative.
But in the long run, token incentives can actually only solve the problem of how the protocol attracts participants in the early stage; they cannot replace real market demand.
When the initial wash trading boom gradually fades and traders are no longer willing to actively bear losses in order to obtain PAPER, whether Papertrade can continue to maintain active trading, stably pay out profits, and continuously generate distributable returns is, I think, the key test of whether this economic model is viable.
From early liquidity mining and trading mining to later points incentives across various Perp DEXs, the crypto industry has always been trying to attract users, capital, and trading volume through token economic models.
Papertrade is a rather special attempt built on this foundation — it turns traditional trading mining into a competition over the right to distribute future cash flows.
For participants who obtained large amounts of PAPER at a lower cost early on, this may be a good investment; but for latecomers, as the cost of acquiring tokens rises and the right to return distribution is continuously diluted, it is becoming increasingly difficult to replicate the returns of early whales.
This design is indeed somewhat innovative, but traders' losses can become income for the protocol and other participants, while token issuance itself does not create new value out of thin air.
When PAPER is no longer the main reason to attract users, is Papertrade itself still a trading platform worth using?
So compared with generating tens of millions of dollars in protocol revenue in just a few days, we should pay more attention to the durability of these revenues and whether the protocol can gradually reduce its dependence on incentive-driven trading activity.