In-depth analysis of the "harvesting technique" of leveraged tokens

In 2020, leveraged tokens gradually became strategic products for exchanges such as Binance, Huobi, and MXC. However, at the same time, an increasing number of complaints about leveraged tokens have emerged, with many investors stating that the price fluctuations of the leveraged tokens they purchased do not align with the product's stated 3x leverage, resulting in losses far exceeding expectations.

Chain Catcher believes, based on research, that this loss risk is similar to the "impermanent loss" in DeFi staking, primarily caused by the rebalancing mechanism of leveraged tokens, which effectively avoids the risk of liquidation in the product's strategy design, as there are always trade-offs.

However, amidst widespread complaints from investors, the negligence of exchanges cannot be overlooked. There is a general lack of market education and excessively low entry thresholds, coupled with many investors' strong speculative emotions and weak risk awareness, which have led to repeated "harvesting" scenarios in the leveraged token market.

Author/Gong Quanyu, Echo

Leveraged tokens are now gradually occupying a significant position in major cryptocurrency exchanges, with trading volumes continuously rising, but the underlying rules still leave many investors puzzled.

Recently, several secondary market investors have reported to Chain Catcher that the price fluctuations of the leveraged tokens they purchased differ from the multiples claimed by the products. "When the spot price was rising before, the leveraged token I bought for shorting dropped significantly, but recently when the spot price plummeted, the leveraged token didn't rise much," said an investor named Hehe (pseudonym).

According to Hehe, she purchased approximately $3,100 worth of 3x short LINK leveraged tokens at an average price of $4.6 on a certain exchange in early November, when the LINK price was around $10.33. During this period, it experienced a significant increase and was quoted at $11.7 on December 23, totaling an increase of about 13.2%. However, the price of the leveraged tokens she held dropped by as much as 81.97%, which she found very hard to accept.

Feedback from investors regarding leveraged tokens in a WeChat group

According to common explanations, leveraged tokens are products issued by exchanges that track the price fluctuations of target assets at specific multiples. In August 2019, FTX was the first to launch leveraged tokens, allowing investors to use 3x leverage to invest in specific cryptocurrencies without margin or borrowing. This product attracted a large number of investors with its "low cost, high return" features and the claim of no liquidation risk, prompting other exchanges to want a piece of the pie.

In the following months, exchanges such as Binance, Gate.io, Poloniex, and MXC launched similar products, making leveraged tokens almost standard for first- and second-tier exchanges, indicating that leveraged tokens have become an important strategic product for exchanges. "Currently, we see many users trading leveraged tokens with 10% to 20% of their assets," said Aaron, Vice President of Futures at Binance, to Chain Catcher.

"Our ETF product's trading volume has now reached 10-20 times that of March, making it one of the most important strategic products. We are also trying 4/5x leveraged tokens," said a relevant product manager at MXC.

Today, there are various designs and names for leveraged tokens on the market. Exchanges like Bittrex, Bitmax, and Gate.io have adopted the FTX model, with their leveraged token products supported by tokens issued by FTX, allowing for transfers between exchanges. Binance issues its leveraged tokens separately, while Huobi and MXC's similar products do not anchor underlying tokens but are merely internal database products, named ETP and ETF products respectively. However, the basic operational principles of the latter three products are consistent with the FTX model. For convenience, this article will uniformly refer to such products as leveraged tokens.

Given that this product has rapidly developed since its emergence in 2019, its extensive coverage and significant industry impact have exceeded many people's expectations. However, the operational mechanism of this product and its bizarre candlestick patterns have left many investors confused, leading to substantial losses. To this end, Chain Catcher conducted a detailed investigation into various design schemes of this product and its actual candlestick patterns, attempting to unveil the "harvesting techniques" behind exchanges' promotion of leveraged tokens.

1. Identifying Problems from Data

According to descriptions from exchanges like Huobi, FTX, and MXC regarding their leveraged token products, if the price of the underlying asset rises by 1%, the net value of the corresponding 3x long product will rise by 3%. Conversely, if the price of the underlying asset falls by 1%, the net value of the corresponding 3x short product will fall by -3%.

To investigate the accuracy of the aforementioned descriptions, Chain Catcher took BTC, ETH, LINK, and SUSHI as examples and analyzed the candlestick data from the four exchanges: Huobi, Binance, MXC, and FTX.

Specifically, Chain Catcher captured the time periods in which the spot price of the underlying asset returned to a specific price after significant rises or falls (selecting dates on the candlestick chart with similar daily price peaks or troughs, with a price difference not exceeding 1%). From this, three time points were selected: the initial date, the date with the largest price change, and the date of price return, recording the corresponding prices of each cryptocurrency's leveraged token products at these time points across the four exchanges, and calculating the performance of price fluctuations for similar leveraged token products at the same time across different exchanges.

To visually present the aforementioned data, Chain Catcher created the following four visual charts, each reflecting the price fluctuation performance of two types of leveraged tokens for a single cryptocurrency across the four exchanges, specifically including the price increase/decrease of leveraged tokens at the highest/lowest points of the underlying asset price during the statistical period, as well as the price increase/decrease when the underlying asset price returned to the initial date price.

Due to the share consolidation mechanism of MXC leveraged tokens, the price candlestick of its XRP short token is distorted, so it will not be compared here.

Based on the data in the above charts, Chain Catcher reporters identified four issues:

First, almost all leveraged token products, after experiencing a period of volatility, find it difficult to return to their original price even when the price of the underlying asset returns to its original position, and may even experience losses exceeding 70%.

Second, the higher the price fluctuation of the underlying asset, the greater the depreciation of the leveraged token products at the return point. For example, XRP experienced a doubling in price during the statistical period, but as it subsequently fell back to its original price, both the long and short leveraged tokens significantly depreciated, with declines exceeding 60% across major exchanges.

Taking FTX's 3x short XRP token as an example, this token was priced at $1,746 when it was issued in August 2019, while the spot price of XRP was $0.32. By December 29 of this year, after significant fluctuations, the spot price of XRP had dropped to $0.17, but the price of the short token had fallen to $3.

Third, in a unidirectional upward or downward market, most leveraged token products that follow the correct direction can achieve gains exceeding 3x. For instance, Huobi's 3x long BTC token rose nearly 15 times compared to early October, while the spot BTC only increased by 2.8 times during the same period, but leveraged token products that followed the wrong direction also generally experienced declines exceeding theoretical values.

Fourth, there are significant differences in the price fluctuations of similar leveraged token products across different exchanges. For example, the decline of MXC leveraged tokens is often greater than that of other exchanges, while Binance's declines are relatively smaller.

Additionally, according to the aforementioned exchange data, as of December 28, Huobi had issued a total of 26 leveraged token products tracking 13 cryptocurrencies, of which only 7 products had prices above the issuance price; Binance had issued a total of 36 leveraged token products tracking 18 cryptocurrencies, with only 7 products priced above the issuance price. However, based on general assumptions, it is expected that the number of profitable and losing products in such bidirectional hedging products should be roughly equal.

The strange phenomena observed in leveraged tokens are primarily due to the rebalancing mechanism of the leveraged tokens themselves and the differences in rebalancing schemes among major exchanges.

2. The Working Principle of the Rebalancing Mechanism

According to Chain Catcher, the reason leveraged tokens can achieve leverage is that exchanges implement it through spot leverage (e.g., Huobi) or futures contracts (e.g., Binance, FTX, etc.). The leveraged tokens purchased by users represent a certain amount of leverage or contract shares. However, whenever spot leverage or futures contracts are involved, any product will carry liquidation risk, and exchanges claim that leveraged tokens do not have liquidation risk because they have introduced a so-called rebalancing mechanism for leveraged tokens.

The rebalancing mechanism is a method for exchanges to rebalance the positions of leveraged tokens to ensure that the actual leverage ratio of the product does not deviate significantly from the agreed 3x ratio. Taking a 3x long token product as an example, if the price of the underlying asset drops significantly, it will cause the value of the margin to shrink, resulting in an actual leverage ratio exceeding 3x, thereby increasing the risk of liquidation.

At this point, the exchange will adjust the contract positions of its products, selling a certain amount of contracts to maintain the leverage ratio at 3x. However, if the price of the underlying asset subsequently rises significantly, due to the reduction in the product's position, its actual yield will deviate from the theoretical yield.

For example, if a user uses $100 as capital to purchase a 3x long token product, then under 3x leverage, it is equivalent to holding a position of $300 with $100 margin. If the spot price of the underlying asset then drops by 10%, the actual position value will fall to $270, resulting in a loss of $30 for the user, reducing the capital to $70. At this point, the actual leverage ratio of the position will increase to 3.85x.

Assuming this multiple reaches the rebalancing conditions set by the exchange, to reduce position risk, the exchange will reduce the user's position to 3 times the user's capital, which is $210.

If after rebalancing, the price of the underlying asset rises by 11.1% back to the original price, the actual value of the user's position will reach $233.3, an increase of $23.3 compared to before the reduction, while the margin will also increase to $93.3, resulting in a loss of $6.7 compared to the user's initial capital of $100. This is the logic behind the general losses shown in the previous charts when the spot price returns to the point.

Based on the same analytical logic, if the user uses $100 to purchase a 3x long token product and the spot price of the underlying asset rises twice by 10% (a cumulative increase of 22.1%), and the user increases leverage/positions once during this period, the user's actual asset return rate will reach 69.00%, exceeding the theoretical value of 66.3%. This is the logic behind some leveraged tokens showing actual returns exceeding theoretical values in a unidirectional market.

3. Differences in Schemes Among Different Exchanges

Of course, the aforementioned analysis is based on a simplified model of market trends. In reality, the return situation for users purchasing leveraged tokens also needs to consider various factors such as the actual rebalancing mechanism of the exchange, the actual leverage ratio at the time of purchase, fee mechanisms, market depth, etc.

The most critical factor is the exchange's rebalancing mechanism, which can generally be divided into scheduled rebalancing and unscheduled rebalancing. The former refers to exchanges conducting position rebalancing management at a fixed time each day, usually around midnight, to ensure that leveraged tokens are always at the agreed leverage ratio. The latter refers to temporary rebalancing by the exchange when the actual leverage ratio of leveraged tokens exceeds a certain level, to ensure that leveraged tokens do not face liquidation risk during extreme market conditions.

From the previous analysis, it can be seen that when the underlying asset of leveraged tokens experiences a reverse market trend, each rebalancing by the exchange will result in net value erosion for the user's position, and the more frequent the rebalancing, the greater the erosion. In extreme market conditions, this could lead to a situation where the value approaches zero, and user assets may be "harvested" invisibly. It can be considered that position net value erosion is a side effect of the strategy design of leveraged token products to avoid liquidation risk, as there are always trade-offs.

Although erosion cannot be avoided, exchanges can minimize it through detailed design of the rebalancing process. Most exchanges adopt a mechanism of parallel scheduled and unscheduled rebalancing, such as FTX, Huobi, and MXC, but the actual operations vary slightly. According to more public information, FTX will automatically trigger the rebalancing mechanism as long as the actual leverage ratio of the leveraged token exceeds the target leverage ratio by 33%; MXC stipulates rebalancing when the price of the underlying asset fluctuates by more than 15%; Huobi generally conducts rebalancing when the actual leverage ratio reaches 4x (for long products) or -5x (for short products).

Binance, on the other hand, has eliminated the scheduled rebalancing mechanism, maintaining the target leverage ratio between 1.25x and 4x, and only rebalances when the actual leverage ratio approaches the threshold. This means that Binance's leveraged tokens do not fix the leverage ratio at 3x like FTX and other exchanges, which can reduce the frequency of rebalancing. This is also the main reason why Binance's leveraged tokens have relatively lower gains in upward markets and smaller declines in downward markets. Binance also stated, "The outside world cannot predict the target leverage ratio and rebalancing time that Binance's leveraged tokens aim to maintain, which helps reduce preemptive trading and manipulation."

At the same time, the fee mechanism can also erode the value of leveraged token positions. Major exchanges deduct management fees from the net value of leveraged token positions daily, usually ranging from 0.01% to 0.1%. However, MXC charged a funding rate of 0.2% to 0.4% per leverage for this product before the end of December last year. Binance also has a funding rate setting, charging three times daily, but claims that this rate is paid by both long and short sides, not collected by Binance.

Thus, exchanges can also generate considerable income. For example, as of around 5 PM on December 24, the 24-hour trading volume of all ETP products on Huobi reached 2.78 billion yuan. Assuming a transaction fee of 0.2% for both parties, the daily fee income could be 5.56 million yuan (considering that some users may enjoy fee discounts, the actual income would be slightly less). At the same time, the total fund scale of all ETP products managed by the exchange is 780 million yuan, allowing for a daily deduction of approximately 270,000 yuan in management fees.

In addition to these apparent fees, leveraged tokens also have some hidden costs, such as the daily interest on the underlying spot leverage or the daily funding rate of futures contracts. The fees collected by exchanges and the transaction fees for each rebalancing will also erode the net value of leveraged token positions.

4. Market Education Has a Long Way to Go

It is precisely due to the aforementioned reasons, as well as issues related to market depth and speculative emotions, that leveraged tokens have exhibited the various problems presented in the previous charts, while most investors remain unaware, leading to significant losses.

In fact, much of the information and logic regarding leveraged tokens is presented on the product introduction pages of major exchanges. However, among the many investors Chain Catcher inquired about, most were not familiar with the specific operational mechanisms. This may indicate that many investors exhibit a lax attitude and are overly casual when purchasing leveraged tokens, and on the other hand, it may also suggest that major exchanges have significant issues in market education and entry thresholds, particularly with a tendency to downplay risks and obscure important information.

According to tests conducted by Chain Catcher on various exchange apps, currently, users of MXC and FTX do not receive any pop-up prompts or mandatory test questions when they first enter the leveraged token trading page, allowing them to trade directly.

Although Binance and Huobi will pop up product recognition test pages, the system will also provide correct answers under certain conditions, and these product explanations and tests do not present the principle of net value erosion that could lead to significant losses. Users can only find related information on the exchanges' dedicated product explanation pages, including warnings against long-term investment and the significant net value erosion in volatile markets.

Initial purchase test for leveraged tokens on an exchange

These issues indicate that exchanges have not fulfilled their responsibilities in market education and entry thresholds. Under the temptation of enormous profits and the influence of a restless industry environment, exchanges have long neglected and become complacent in this regard. However, as leveraged tokens become increasingly known and purchased by more investors, if exchanges do not strengthen their efforts in this area, they may ultimately face backlash from the impressive data they currently showcase.

In interviews with several exchanges, Binance, Huobi, and others stated that they would strengthen user education. "We have developed a detailed user education plan that will be gradually implemented early this year, while continuously optimizing products with a focus on risk control and user experience," said Huobi.

In reality, leveraged tokens, as a type of cryptocurrency derivative product, have their own value, similar to futures contracts, such as for intraday arbitrage or hedging, or for investing in cryptocurrencies that have relatively unidirectional trends over a longer period. However, the problem lies in the significant volatility of most cryptocurrencies, coupled with the superficial market education provided by exchanges, leading a large number of investors to enter the market without understanding the truth, ultimately suffering from "harvesting" in an invisible manner.

Although this "harvesting" is primarily caused by the product mechanism and cannot be directly attributed to the subjective intentions of exchanges, it remains a product that has harmed a large number of investors, necessitating more robust reforms and improvements from exchanges, and it warrants greater vigilance from investors.