Robinhood charges fees on multiple chains through Uniswap API, but in the end, they don't want to keep a bunch of blue-chip tokens, long-tail assets, and small balances. Instead, they aim to convert them into USDC and USDG. In the past, applications had to set up their own exchange programs, find routes, and handle multiple chains. On October 29th, Uniswap Labs launched TJAR, turning this process into a public market: fees go into Token Jar, and buyers bid to clear out the assets, with the application receiving the pre-selected tokens.
TJAR was initially designed to serve for the conversion of transaction fees according to the Uniswap protocol and for the destruction of certain assets using UNI. Now, it has been developed into an infrastructure that can be used by any blockchain system. Uniswap states that it processed over 15.3 million US dollars in conversions last month, and it has been deployed on seven different chains. Robinhood also uses it to handle fees collected through Uniswap API.
A jar entrusts the exchange work to the seeker.
Each operator can deploy their own Token Jar and Releaser through TJAR Factory. Token Jar is responsible for accumulating native tokens and ERC-20 assets. Releaser specifies which target token these assets should be exchanged for, how much needs to be paid, and where the target tokens will ultimately be sent.
The operator sets a release threshold first. The searcher continuously monitors the “jar.” When the market value of the assets inside it is high enough to cover the cost of the target tokens and still leave a profit, the conversion is triggered. The searcher pays the target tokens required by Releaser, and the “jar” releases the selected assets to the searcher. The target tokens then go into the treasury designated by the operator, a cross-chain destination, or are destroyed.
This is not the traditional approach where protocols actively seek the best quotes. TJAR is more like a public opportunity with a fixed price: the threshold remains unchanged, while the value of the assets in the “jar” changes with the entry of transaction fees and market prices. Searchers decide when it is worth taking on a order and compress their profit margins through competition.
Low threshold, quick clearance, and it may also leave more of the price difference to the searchers.
The release threshold determines how long an asset remains in the “tank.” If the threshold is low, the tank is cleaned more frequently, and the asset is exposed to price fluctuations for a shorter period of time. However, if a larger transaction fee suddenly enters, the value of the assets in the tank may exceed the fixed threshold, resulting in a larger profit margin for those searching for such opportunities.
The example given by Uniswap is: the release threshold for the jar is 1000 USDC; the original asset value was 950 US dollars; after a sudden transaction fee of 500 US dollars, the value of the jar rises to 1450 US dollars. The searcher still only needs to provide 1000 USDC; after deducting the execution costs, they can obtain a larger remaining value. For the operator, this means that faster clearance does not necessarily equate to higher conversion efficiency.
A higher threshold can reduce the impact of large single deposits, but it will cause assets to remain in that "jar" for a longer period, exposing them to more price fluctuations. There is no fixed value that suits all applications. The frequency of transaction fees, the size of each transaction, and the price risks that operators can afford all affect the configuration. Owners of Releaser or the designated configurators can adjust the target tokens, thresholds, and destinations; they can also choose to relinquish control in exchange for a guarantee that the configuration remains unchanged.
Robinhood First, convert multi-chain fees into two types of stablecoins
Robinhood has already started using TJAR on the supported chains, converting the transaction fees collected through Uniswap API into USDC and USDG. For a cross-chain application, having the same Factory and deterministic deployment address can reduce the list of addresses that need to be maintained across different chains. Additionally, Token Jar will also be able to receive fees from more sources in the future without having to recreate the conversion mechanism.
There are still limits to this convenience. TJAR outsources the execution of exchanges to economic incentives, without relieving operators of their pricing responsibilities. Setting the threshold too low or too high, choosing the wrong target tokens, or making mistakes in configuring permission management can all affect actual revenue. Whether searchers are willing to take action also depends on whether the value within the “jar” (in this context, likely referring to the platform’s asset pool) is sufficient to cover costs and generate a profit.
What truly changes with TJAR is the “last mile” of on-chain transaction fees. Applications no longer have to maintain a set of robots that convert countless assets into a single revenue currency manually; instead, they can simply initiate a transaction and let searchers handle it. Operators receive a more consistent stream of assets to be recorded in their accounts, searchers get the opportunity to profit from price differences, and configurators take on the responsibility for setting thresholds and granting permissions.
When the source of transaction fees spans multiple tokens and chains, collecting fees itself becomes an asset management issue. Token Jar breaks down this problem into observable, competitive markets, but it does not eliminate the costs. It merely shifts these costs from within the exchange system to a choice between threshold settings, price exposure, and profits for searchers.
