Follow Up on Huobi’s HUSD v2.0 Solution

We want to give special thanks to TRM for helping co-review and co-research this analysis. TRM is rolling out analytical solutions to monitor blockchain activity for the purpose of identifying what channels and user groups drive growth.


Given the multitude of available stablecoins, it’s becoming more apparent that a basket of stablecoins serves an important use case for facilitating crypto-market activity. In our perspective, a successful stablecoin basket needs to have the following interrelated properties:

⬜ Incentivized pricing to properly balance reserves

⬜ Evenly distributed flows amongst supported assets

⬜ Balance sheet preservation and avoidance of capital losses

⬜ Promoting liquidity onto broader exchange pairs in the market

Huobi’s stablecoin solution, HUSD, advertised that it would become a unique service for users’ needs by interchanging supported stablecoins — an announcement which was well-received. However, overlooked design flaws prevented a beneficial situation, and in less than 4 months Huobi was losing millions of dollars, as we pointed out in our analysis. As a result, Huobi needed to go back to the proverbial drawing board in order to design and figure out what a meaningful stablecoin solution would look like.

Since then, several recent events occured:

  • Gemini’s original strategy was to bootstrap their new token by offering discounts to traders, who would then disseminate their token onto various exchanges in order to drive liquidity. Unfortunately, these desks merely arbitraged profits through the HUSD solution and left Gemini Dollar illiquid and locked up in Huobi’s balances. We expect an emerging conflict looming in the future when reported news surfaced that Gemini shut down several accounts attempting to redeem their stablecoin. This in mind, Huobi is unlikely to lower their GUSD reserves by exiting to fiat because Gemini is pushing back, which means their only way to lower exposure is to figure out a way to distribute GUSD effectively into the markets while promoting the adoption of other stablecoins onto their platform.
  • Huobi upgraded HUSD to “v2.0” where the supported stablecoin exchange rates were considered based on an internal pricing methodology. This trial period recently ended so we were curious to see whether or not certain issues were resolved. While we didn’t expect this to be the case, based on our initial analysis, we can now take a more granular look since this phase of the operation has ended.
  • Huobi also announced ERC20 support for Tether. Whether this will be supported in HUSD on interchangeability is an open question at this time. We will touch lightly on the implications near the end of this article.

We were quick to realize that not all aggregate stablecoin basket products are created equal — and each unique design can carry implications if overlooked. In order to service the needs of participants in a constructive fashion, we need a solution which can lead to less friction and complication in the markets — all while providing traceable price discovery mechanics. By digging deeper and acknowledging the pitfalls of HUSD’s implementations, we hope to prove the need for a product like the Neutral Dollar.


Based on Huobi’s announcement, the purpose of the HUSD solution was to “better meet users’ demand for stablecoins”. It’s unknown whether this is to promote stablecoin trading on their exchange platform or to serve as a use case for traders utilizing stablecoin strategies. Ideally, Huobi would want to bring more inflows and drive trading activity on their exchange (as that is their main revenue driver), but we can interpret their success on this through our research. Author’s opinion: we don’t think it’s working effectively due to some short-sightedness.

The data we used was sourced directly from etherscan, where name tags were used to identify Huobi wallets. We queried all TUSD, USDC, PAX, and GUSD transactions sent and received from these wallets since the start of Huobi’s HUSD solution.

Below is the aggregated activity of inflows and outflows of HUSD thus far, which gives us a current snapshot of what stablecoins are in their reserves.

From the table, most of HUSD consists majorly of GUSD, some PAX, TUSD, and a bit of USDC. Across their wallets, Huobi currently holds roughly 66 million tokens of these stablecoins.

We also would like to break down further the total amount of flows between the two operations conducted by Huobi, the first being the 1:1 exchange rate (v1.0) and the second with the modified methodology (v2.0).

During version 1.0 we saw huge arbitrage plays that lead large inflows into Huobi, primarily from GUSD. Traders exploited the exchange rate and diverted their profits to cash out through PAX due to its easier redeemability into fiat. From our initial case study, we found that the overall use case for HUSD at this time was to exploit and profit off of design flaws. At the time of our initial case study, over 70% of GUSD was in Huobi reserves due to this arbitrage play. We feel that a few of the version 2.0 goals were to 1) not be taken advantage of and 2) lower GUSD exposure while facilitating more transactions from other stablecoins.

During version 2.0, Huobi experienced net outflows and lowered their supported stablecoin balance by ~13 million. The amount of inflows during this entire month was roughly 20 million, and most of the stablecoin flows were around GUSD. It’s still an open question as to how much Huobi wants to actually lower their exposure to GUSD; while they did experience almost ~29 million tokens of GUSD outflows, they still saw more inflows comparatively. It’s reasonable to assume that the pricing methodology they used to establish exchange rates may still not be completely representative of all market activity.

It’s unknown what Huobi’s intentions are with this service and it’s difficult to define what their actual goals are for their product. **While we do not have substantial evidence, this is our guess as to what has been going on recently **(and we are happy to be wrong, so long as there is proof):

  • Huobi cannot directly liquidate their GUSD exposure because of rumored Gemini pushback, so they need to figure out how to lean down their exposure through pricing. Their offset used to calculate the premium/discount is unknown so we are unsure either if Huobi is taking a loss or if they are succeeding in facilitating an organic outflow of GUSD. We think the latter is unlikely to happen quickly, as several OTC desks find GUSD to be difficult to work with.
  • GUSD, like all other stablecoins, can still issue rebates or discounts to traders. Because GUSD is already illiquid and locked up in Huobi reserves, traders may not feel that distributing GUSD onto other exchanges is a worthwhile operation. They can still see some form of profits by transferring into Huobi after acquiring it somehow, which is why you see more inflows into Huobi from GUSD.
  • PAX, TUSD, and USDC should see more favorable exchange rates offered by HUSD because of the lower balance. This theory is supported by less outflows on each of these stablecoins, meaning that the offset on these stablecoin pricings are at least working in the right direction. Why there aren’t as much inflows into HUSD is the bigger question.
  • Less stablecoin balances on HUSD means there is likely less stablecoin trading activity taking place on Huobi. If the intention is to work as a clearing house, then we can argue that version 2.0 may be working better than version 1.0. If it’s to promote more stablecoin trading activity on their platform to access BTC, ETH, amongst others — it’s a harder sell. Huobi’s trading using HUSD as a quote currency does not procure any distinctive advantage over any exchange, so there’s no natural incentive for people to want to deposit funds into Huobi, which can snowball into a liquidity problem.

Additionally, we can break down our findings by seeing which transactions were in-between Huobi wallets and which weren’t. The former likely being organic behavior to promote activity on Huobi’s platform, while the latter conveys only using HUSD as a service to exchange amongst stablecoins for arbitrage or liquidity purposes to avoid slippage elsewhere.

To date, it looks like much of the activity was utilized by traders to avoid using Huobi’s exchange platform directly since most of the sending and receiving was not directly between Huobi-owned wallets.

When we look at the activity during version 1.0’s duration, we notice the arbitrage taking place — a ton of activity is seen, especially between GUSD and PAX, with Huobi eating whatever loss to prop up the 1:1 exchange rate.

Under version 2.0, we do see some operational transfers on PAX and USDC. However, this total amount of ~1.3 million pales in comparison to the ~$32 million that was moved outside of Huobi wallets. As a clearing house service Huobi does seem to be making slight progress on improving exposure, but to the net outflows show that there isn’t much promotion for activity on their exchange platform.

We can also show Huobi’s reserves over time with respect to each stablecoin. Below is a visual below, with the red line marking when version 2.0 was taken into effect.

A quick inspection shows that when version 2.0 became online, HUSD slowly started decreasing their GUSD exposure and increasing a bit of PAX and USDC.

We can observe this taking effect when we look at the weighting composition of HUSD — while still predominantly consisting of GUSD, Huobi does seem to be lowering their exposure in favor of other stablecoins. Still, there is a lot of work to be done to actually balance out their stablecoins reserves. The offset from their stablecoin positions are unknown, so we don’t know the ideal weighting. If it’s even possible, it will definitely take time for Huobi to reach their targets.

Next, we want to take another perspective on Huobi’s reserves by extending their balances in proportion to the circulating supply of each supported stablecoin through version 2.0. We can measure HUSD’s success if they see more flows to balance their asset exposures properly. The purpose of doing so is to see how much of the stablecoin market Huobi gets to interact with in a relative sense, and if there are any concerns to identify.

This chart indicates that Huobi has begun to acquire a larger proportion of PAX relative to the circulating supply. We view this positively in Huobi’s attempt to balance collateral as a first step but need to reason why this is specifically only happening on PAX at this degree.

There’s a similar trend on TUSD, though to a lesser degree. Interestingly, PAX currently has less presence than TUSD, which we believe could also explain why there would be more deposits from PAX into HUSD. There’s less opportunity to work through various exchanges compared to TUSD.

Huobi owns the least amount of USDC (at the time of writing, the largest market cap of all the supported stablecoins) and doesn’t seem to be acquiring more at scale in a relative measure. It’s interesting to note that USDC experienced the most drawdown in circulating supply than all supported stablecoins. Likely, pricing is not attractive enough for traders to use the HUSD service, despite already being favorable (given the current reserve amounts on Huobi). Since we don’t see any complementary flows on HUSD, we can conclude most activity on USDC is happening outside of Huobi’s reach which presents an opportunity for improvement.

Looking at the outstanding tokens owned by Huobi relative to the circulating supply, we observe that Huobi still owns ~70% of GUSD. While exposure is slowly getting lower, this is currently as a very prevalent liquidity issue which means that other GUSD markets are likely to be less favorable than other stablecoins pairs. The idea that ‘liquidity begets liquidity’ makes this a troubling issue to resolve naturally.

In its present state, the HUSD solution as a clearing house seems to be more beneficial for the traders. For us to view HUSD as a successful endeavor, we would like to see more growth in activity and total reserves their various supported stablecoins. While Huobi is making slight progress on lowering GUSD exposure, they aren’t exactly doing the best in relative terms. During version 1.0, PAX was the primary choice for traders to redeem because it was easier. Since that opportunity has elapsed, the likely situation unfolding sees PAX become less attractive for traders to hold on their balance sheet. Therefore, instead of the traders moving PAX off their balance sheet through trading activities, they clear transactions through Huobi. This would explain why there is less proportion of TUSD and USDC in Huobi despite being larger stablecoins: because traders have more options.

It seems Huobi is also finding trouble in growing their reserves on these stablecoins with their methodology on pricing. However, we can’t say for sure if Huobi is taking a loss for leaning or skewing their quotes to manage their balance sheet. This shows that Huobi is paling in comparison to the growth of the supported stablecoins while managing their own internal complications, which is something they will need to figure out in the longer term.

We also want to make note of Huobi’s announcement of ERC20 Tether support. It’s currently unclear if this is going to be included on HUSD. Despite Tether being supposedly fiat-backed, there can be some implications. ERC20 Tether support is not very widespread at the moment and can result in liquidity or operational issues, as many other exchanges only support the Omni version. It will be interesting to see if Huobi wants to integrate the other variant and if they can do so without friction. Because HUSD is not directly transparent about it’s positioning or collateral, incorporating Tether needs to be justified by strategic reasoning. We hope that people acquire healthy skepticism based on version 1.0’s failure and can avoid missteps in the future. Whether Huobi also wants to be more open about their goals for the HUSD solution, is an open question to raise.


2018 was coined the ‘Year of the Stablecoin’, which is materializing even more in 2019. Even J.P. Morgan announced their stablecoin for cross border payments, and Facebook is also entering the ring. We view the stablecoin competition heating up with no signs of slowing down. However, we feel strongly about the opacity with many current stablecoin models. Many stablecoin projects (despite broadcasting that they are transparent) employ several opaque practices such as fractional reserves and rebates to special partners. Holder of such stablecoins really see no benefit outside of convenience, and it’s hard to outline preferences of one versus the other. Simply put, owning a stablecoin allows for their originating business to profit off of your standing balances. We expect changes to occur to distinguish and actually benefit users in the nearing months. Still, the impact of offline agreements for discounts can be argued on who actually benefits the most (probably not the retail users). With this all occurring in the backdrop, many projects are looking to provide services on top of the existing industry activity to address different stablecoin needs such as Huobi, XBTO, and ourselves.

We believe that Huobi’s version 2.0 of HUSD is avoiding large losses like what we saw in version 1.0, but it’s hard to estimate success on whether it’s profitable as the overall change has been marginal. In light of servicing Huobi’s exchange platform, we do not see this as beneficial for the company itself with the net outflows seen in version 2.0’s trial operation and less operational usage on other pairs. The HUSD service could be an avenue for traders to deposit less attractive stablecoins which, at times, simple pricing methodologies cannot fully reveal. There is more to estimating a stablecoin’s worth than from its redeemable value, which can be a difficult problem to assess. From our investigation, we’ve found that inventory management can be hard to implement in practice, and diversifying asset exposure can be tricky to navigate while avoiding losses. For a stablecoin basket to be effective, there must be careful review of the collective position targets and pricing mechanics in order to rebalance stablecoin reserves effectively. Going back to our criteria on what makes a successful basket, we can see whether HUSD v2.0 did match all these requirements:

✘ Incentivized pricing to properly balance reserves

✘ Evenly distributed flows amongst supported assets

? Balance sheet preservation and avoidance of capital losses

✘ Promoting liquidity onto broader exchange pairs in the market

Upon review, we do not find that Huobi is popularizing the stablecoin basket concept. In fact, HUSD as an endeavor is a struggling one. We can only wait and see how Huobi plans to iterate on their solution in the future, as we feel that their operation is not functioning appropriately.

With more stablecoins on the horizon, it’s become more apparent to us that a metastable solution is necessary. HUSD’s closed-form service is being viewed more as a clearing house and not a direct, tradable instrument to foster liquidity and growth directly in the markets for all types of users. The current design causes limitations on its offerings, and major alterations might be required for more successes to materialize on Huobi’s businesses. We believe an open-end construct by design can align incentives and offer better transparency along with liquidity. We also conclude that not all metastable solutions are equal, and design flaws can limit the ability to have meaningful impact. Neutral believes wholeheartedly that our metastable basket can match all these checkpoints. Having options available is one part of helping consumers, but enabling people to use them in beneficial ways is really the solution that the Neutral Dollar is going to hopefully become.

— — — —

If you’re interested in learning more about our team and products, check out our website or start a conversation with us on Telegram or Twitter.