Gauntlet's Real Time Dashboard Explained

Summary

Gauntlet is excited to provide a deep dive into our Dashboard and highlight some key features and use cases. Our public front end is a key pillar in serving the Jupiter community with real time data about positions and risks affecting the protocol. We expose the same data we use in our analysis internally, allowing the community to work in parallel with us.

Components

Topline Metrics

At the protocol level, key metrics we track to evaluate protocol health includes volume, open interest, and liquidity. Our data is streamed via RPC calls in virtually real time (leaving a small buffer to ensure blocks are finalized). Freshness is important as positions can change rapidly. Some aggregations such as returns and volatility will update on a longer cadence, as they derive meaning not from a single point in time but describe performance over time. ‘Last updated’ timestamps provide this context around the data.

On Jupiter protocol, stablecoins USDC and USDT supply liquidity to short positions, while volatile assets SOL, ETH, and BTC back long positions. Short liquidity is provided in the summary section, while liquidity for specific assets is provided in the market summaries.

Market summaries

On the protocol summary page we include high level metrics for each market, namely 24 hr volume, long open interest, short open interest, available long liquidity, and leverage.

24 hr Volume represents the total notional amount traded in the last 24 hours, calculated using the volume in the current hour, updated in real time, plus the preceding 23 hours.

Long/Short OI represents the notional amount in long and short positions currently open on the protocol.

Long liquidity represents the LP provided funds that are currently unmatched to open positions.

Leverage is a function of notional trader position value and the underlying asset price. As asset prices fluctuate, so does the leverage of current positions since collateral provided as margin is denominated in stable coins. The red/green color coded numbers (’trend’) represent how a given metric has changed over the last 24 hours. The trend helps us derive useful insights. For example, if leverage is increasing along with available long liquidity, this indicates that LPs are more willing to provide liquidity to the market even as asset prices are falling. This bodes well for traders continuing to be able to open positions.

Market utilization represents the portion of JLP liquidity matched against trader net long positions for a given asset; this is a measure of capital efficiency. Higher utilization results in higher fees accruing to JLP holders, while leaving less opportunity for trading to open or increase their positions.

JLP Pool Information

We describe, in real time, the current composition of the Jupiter Liquidity Pool. We monitor the pool composition, as well as the tracking error against target weights. An asset being far from target indicates that arbitrageurs are not willing to rebalance the pool, signaling that incentives are misaligned (liquidity for a given asset may be too low on the Solana ecosystem for example, resulting in slippage higher than the arbitrage opportunity). If tracking error remains high for an extended period, the core protocol team may need to come in and rebalance the pool.

In addition, we monitor whether JLP is trading at a premium or discount on DEXes. We generally want to keep JLP trading at a premium, which would indicate the market has an expectation of future fee revenues accruing and outpacing trader profits. One goal of our trading fee methodology is to keep JLP at a premium such that liquidity providers are encouraged to add more liquidity to the pool, increasing liquidity available to traders. When JLP trades at a discount, LPs are encouraged to purchase JLP on a DEX which does not increase the size of the pool.

Market Specific Pages

At the market level, we monitor long and short open interest, borrow rates (which represent the cost of carry similar to a funding rate on other protocols), utilization, leverage and volume. Just as on the protocol summary pages, the ‘trend’ numbers provide useful insights to how risk is changing on a given market.

Trader Behavior

Below the summary statistics we have a number of charts that visualize activity on the protocol such as volume and distribution of leverage, and open interest. A healthy protocol sees higher volume (turnover of positions) when asset prices move as traders use perps to hedge their positions. Interest and OI Cap usage charts are provided as additional context.

JLP Activity

The JLP activity charts are specific to a given volatile asset (in the example above, ETH). Utilization numbers indicate whether trader demand is being effectively met by ETH liquidity providers, while the fee revenue chart displays how profitable this liquidity is for JLP holders.

Taker Activity

Taker fees summarize the amount paid by traders for opening (position fee) and maintaining (interest fee) positions in ETH derivatives, long or short, on the protocol. We always seek to balance fees to remain attractive for both sides of the market.

Features and Navigation

Tooltips - we’ve added tooltips on every chart and data tile to provide definitions and context around how each datapoint is calculated and why it matters. Hover over the circle with the “i” next each title to view.

Timestamps - each datapoint displays a timestamp that defines when the data was fetched. Note: the data will not currently update automatically while you have the dashboard displayed. You must hit refresh if you want the latest data (auto refresh is in our backlog for a future release). Upon hover, the date and time of the dataset will appear, like so:

Filters - charts with more than one data series have filter functionality, represented by a funnel icon at the top right. Click on the icon to expand the menu to select / deselect any series. You may also click on the series names in the legend to filter them out. Charts are filtered individually (applying one filter on a chart will not affect any other charts).

Full screen views - click on the four corners icon at the top right of any chart to expand to full screen. This view will also provide you a unique URL that allows you to reference a specific chart directly, such as in a tweet or hyperlink.

Navigation Menu - the dashboard has two level of information: summaries for the protocol as a whole and market level data (ie. SOL, ETH, etc.). You can navigate in two ways.

  1. Using the Top Menu - click on the “+” next to the protocol name and select the desired market. This works on any page.

  1. Using the “Details” button in the Market Summary Table

We’ve seen many of you visit the site and welcome any feedback you have!

Links

6 Likes

Excellent work. Would be great to have:

-Performance of JLP vs a benchmark like a 50/50 SOL/USDC pool
-Traders win/loss amounts per day.

GMX has these graphs in their dashboard and its quite helpful for LPs

2 Likes

Why does the TVL of each pool not match the jup.ag dashboard pool sizes?



for SOL: 209.10M <> 238.3M

Some remarks and open questions:

  • How to explain?
    In the case of SOL
    long OI + available long liquidity <> SOL TVL
    114.1 + 119.7 <> 206.1M
    233.8M <> 206.1M
    leverage 4.9x

Numbers are constantly changing, but there is always a significant gap.
long OI includes 1 out of 4.9 capital provided by the trader?
114.1*3.9/4.9 + 119.7 = 210.5M <> 206.1M

  • formatting issue

  • I would recommend consistency in naming. Chart names differ from the names used in the description. Does the position fee include liquidation fees?
    image

1 Like

Hi @GCB, thanks for the questions.

At a high level, the reason for most of the noted mismatches is that our dashboard has two different kinds of metrics definitions:

  1. Metrics that represent actual smart contract state, such as utilization. These metrics give you information about what would actually occur when you try to take action on the protocol. For example, utilization informs the borrowing rate that open positions are currently accruing interest against. Thus, Gauntlet faithfully matches the smart contract definition for such metrics.
  2. Metrics that inform us about protocol health, such as open interest. These metrics only abstractly affect outcomes for traders and LPs. For example, open interest represents a portion of the exposure that JLP holders have to the returns of the underlying assets. So, Gauntlet defines them in the way that we believe provides the most insight to the community. In some cases, Gauntlet’s definitions may not exactly match comparable metrics displayed on the Jupiter frontend. In other cases, Gauntlet’s definitions may appear not to be internally consistent, when trying to validate calculations using a mix of smart contract state metrics and protocol health metrics.

Why does the TVL of each pool not match the jup.ag dashboard pool sizes?

Pool TVL is an example of a protocol health metric whose definition differs between Gauntlet’s dashboard and the Jupiter frontend.

Jupiter Frontend Pool Size

Jupiter frontend defines Pool Size differently between risky and stable assets.

  • Risky assets: Pool Size is the market price of all the assets in the pool, including trader collateral, idle JLP liquidity, and liquidity that is currently reserved for PnL for open trader positions.
  • Stable assets: Same as risky assets, but does not include trader collateral.

Gauntlet Dashboard Pool TVL

Gauntlet defines Pool TVL as the total market value of the assets that would remain in the pool if, hypothetically, all traders were to instantaneously settle and close all their open positions, incurring zero fees.

Please note the following properties of Gauntlet’s Pool TVL, which differ from Jupiter’s Pool Size:

  • Trader collateral is never included
  • Unsettled trader PnL of open positions is taken into account
  • The sum of the five Pool TVLs is equal to the JLP TVL
1 Like

Please see the replies below:

Leverage

How to explain?

In the case of SOL

long OI + available long liquidity <> SOL TVL

The reason why the above relationship does not quite hold, even after accounting for Leverage, is because of a slight difference in how fees are incorporated. If not for those fee differences, then it would indeed be true that long OI * (leverage - 1) / leverage + available long liquidity = pool TVL.

For the purposes of Pool TVL, any trader fees that have been accrued but not yet paid (specifically, borrow fees accrued, and position fees needed to close the position) are ignored. This is because those fees technically don’t belong to JLP holders yet, and any JLP redemptions will not have access to their pro-rata portion of those unpaid fees. Similarly for Long OI and Available Long Liquidity, accrued fees do not come into play.

On the other hand, for the purposes of Gauntlet’s definition of Leverage, the collateral in the denominator does account for accrued but unpaid fees. This is due to Jupiter’s liquidation mechanism, which considers a position liquidatable if its collateral would not meet the maintenance requirement after paying the costs to close the position.

Thus, the persistent gap between long OI * (leverage - 1) / leverage + available long liquidity and pool TVL is exactly equal to the fees accrued by traders but not yet paid to the pool.

Liquidation Fees

Does the position fee include liquidation fees?

No it does not, but we are working on an update to add liquidation fees to the chart. We expect to roll it out next week.

1 Like