In the past week, the battle between Binance and FTX ensued, one that will enter the history of the crypto industry, garnering attention globally from all industries. The aftereffects of this battle spread throughout the entire industry, consequently triggering a “crisis of trust” in the industry. Users and their level of trust in exchanges have fallen to an all-time low.
This fall in trust levels prompted various major exchanges to brainstorm ways to prove that they are “safe” in order to avoid a bank run and becoming the next FTX.
The first to release their proof is Binance, situated at the center of this saga. Binance’s founder, CZ announced on 9 September that they are beginning to do their merkle tree proof-of-reserves to prove full transparency of their reserves. Thereafter, exchanges like OKX, BitMEX, Gate.io, Kraken, Huobi, Bitget, Kucoin, Poloniex and Bybit all announced that they will soon release their proof-of-reserves.
When these exchanges are all prepping to do their proofs, the term “merkle tree proof-of-reserves" also started gaining traction. Basically, the merkle tree is a method of proving that the assets an exchange possess has 100% reserves. When the exchanges release their merkle root, there are 3 ways to prove if it has 100% reserves:
1) Any user can verify their balance and UID on the tree;
2) A third party audit company to audit the aggregation of all the users’ balances on the merkle tree;
3) A third party audit company to audit all the wallet addresses and the sum of their balances from the exchanges.
For more details, please refer to https://www.coinlive.com/news/detail/?id=18700
Since the merkle tree can proof 100% of an exchange’s reserve, does that mean all exchanges which proved their merkle tree are safe? The answer is no.
Although the merkle tree is able to proof 100% of an exchange’s reserves, but due to its limitations, it is not appropriate to validate the safety of an exchange. Exchanges flock to prove their merkle tree because it provides a façade of safety for the industry. It does not actually solve the actual problem.
Through an interview with GeniiData’s analyst, we collated a few pointers for discussion’s sake on why the merkle tree is not suitable to prove the safety of an exchange’s short term reserve assets.
1.The merkle tree only proves the current state of the assets, whereas in order to prove its entire assets, there is the possibility of borrowing.
Due to a mechanical flaw in the merkle tree, it can only prove that the originator possesses this amount of reserve assets. However, it is neither able to prove the origin of these assets, nor fully prove if it is the originator. This therefore results in the possibility of the originator “borrowing” assets to gather enough in order to get approval, right before preparing its “proof”.
At the same time, due to the volatility of coins like Bitcoin and Ethereum, it is unable to prove an asset’s subsequent value. Hence, pumping of altcoins to maintain its value will occur. The pumping will only stop after the proof has ended, in which it will return to its original value. FTX’s FTT is a case in point.
These loopholes therefore exist:
1)The “merkle tree proof-of-reserve" is unable to validate the origin of the assets, and naturally unable to prove whether or not the assets are “borrowed”;
2)The “merkle tree proof-of-reserve" is unable to prove if an asset belongs to the originator, and hence unable to prove its assets;
3)The “merkle tree proof-of-reserve" is unable to proof an asset’s subsequent value.
2.Does the time the merkle root is updated mean that the assets during this period can still be misappropriated?
The merkle root is not a regulatory mechanism. Therefore, its mechanism is only able to prove the assets the originator possesses at the time of “proving”, but not where the assets go after. Hence, despite the originator completing its “proof”, they can still engage in misappopriation of assets before the next “proof”.
Notwithstanding, we can understand from GeniiData’s analyst that some exchanges will collate their assets once daily. During its collation, the assets will flow within the liquidity pool. As a result, all the assets within the liquidity pool will be shuffled. There will be a considerable amount of movement for the assets during this period, which makes it impossible to confirm the general trajectory of the assets.
Thus, these loopholes exist:
1)The “merkle tree proof-of-reserve" is hence invalid the moment the “proof” is complete;
2)Before organising the next “merkle tree proof-of-reserve", the originator can still misappropriate assets.
3)The “merkle tree proof-of-reserve" is unable to prove the entire trajectory of users’ assets.
3. The prover’s number of users and deposited fund balance remains a “black box”.
As most of the exchanges in the industry currently is stuck in a “black box” situation, every exchange’s number of users and their assets remain unknown. One party who released the “merkle tree proof-of-reserve" is consequently unable to be clear about their data source. As a result, all of the supposed assets are unable to be proved as entirely the originator’s assets.
There are therefore these issues:
1)The “merkle tree proof-of-reserve" is unable to verify if the originator’s deposited funds are correct.
2)The “merkle tree proof-of-reserve" is unable to verify if the number of users provided by the originator is correct.
4. Are there any other methods to verify all assets in an exchange?
Other than the “merkle tree proof-of-reserve", there are many other ways to prove one’s proprietary assets, including a simple full disclosure principle.
Nonetheless, this method is not simply settled by releasing a wallet address. It requires not only a signature to verify that the wallet belongs to the exchange, but also a signed private key and timestamp to verify that the private key indeed is in the hands of the exchange.
On top of that, it is also vital to check the health of an exchange’s asset structure in order to avoid situations like FTX using FTT to act as its assets. When a user is checking on an exchange’s assets, they should take note of the ratio of funds like Bitcoin, Ethereum and USDT in their wallet. If the ratio of these assets is significantly low, they should be cautious.
5. Can a centralised exchange truly be regulated?
We can reference the logic behind traditional financial institutions. Securities in traditional finance require steps like trading, settlements and delivery. In the example of trading and settlements, it already involves two departments. Exchanges are the platform in which users trade on, so settlement institutions need to have sufficient settlement assets to serve as the counterparty in the trade, ensuring that funds are settled after trading.
However, once the exchange acts as both the “exchange” and “clearinghouse” roles, it is akin to being both the referee and the player. If we wish to truly realise regulations, we can do so by learning from the management of traditional financial industries. For instance, by letting third parties like banks or organisations to manage, we can leverage on the perfect system of traditional finance to ensure the safety of assets.
Such a method is merely ideal. To an exchange, no one is able to give up users’ “deposit funds”, a big slice of the cake. Naturally, there will not be self-regulated exchanges who will leave their fund to a third party to manage.
Currently, the only way to properly manage exchanges are the policies and regulations imposed by the various countries and regions.
6. Are decentralised exchanges the saviour of exchanges?
Centralised exchanges (CEX) are being constantly under fire for misappropriating assets. Will decentralised exchanges (DEX) solve the issue of asset misappropriation?
The answer is yes, it is possible. However, as GeniiData’s analyst mentioned, besides solving the issue of “misappropriation”, other new problems arise.
Currently, there are two types of models in the market:
1)The AMM model represented by Uniswap
There is insufficient depth from the exchange, large transactions will greatly influence Uniswap’s liquidity pool.
2)The Orderbook model represented by 0x
Opensea operates by an Orderbook model as well, but it mainly trades NFTs. Conversely, 0x does ERC20 transactions. They have a few issues as well, such as requiring wallet authorisation, having high transaction fees, and it mainly relies on the Ethereum blockchain.
Both models also possess the security issue of being attacked by hackers, as well as having a high barrier to entry due to its complicated operating manner. DEXes are consequently unable to replace CEXes as the main exchange in a relatively short amount of time.
Bernice
JinseFinance
Cheng Yuan
Weiliang
Joy
Coinlive 
cryptopotato
Siew Meng
Cointelegraph