# TCR Skew-Farming

Tracer Perpetual Pools allow users to go long or short assets with no risk of liquidation. For this reason, Umami Finance is using them to launch our new vaults which generate substantial delta-neutral returns (15+%) by minting GLP, and then hedging the exposure to the underlying BTC & ETH within the GLP using Tracer Pools. This involves going short 3x on the BTC & ETH pools. These pools are also free of any fees!

So what is the catch? How can the pools be free to use? The answer is they are relatively inefficient in hedging if there is not exactly as much TVL on the short-side as there is on the long. As the TVL on any one side increases proportionally to the other, the effective leverage decreases from the target. This can be thought of a "funding fee" that the overweight side is paying to the underweight side. This relative differences in TVL is described as skew.

With Umami expecting considerable TVL for a product which offers 15+% on USDC in sustainable yield, this could present a serious problem for maintaining that delta-neutrality due to the low TVL that Tracer pools currently have. Thus, if Umami cannot find a counterparty to go long with as much as TVL as they will be putting on the Short side (roughly 13% of the total vault TVL is deposited into the 3x short BTC + ETH pools), TVL would have to be significantly capped to ensure Delta-Neutrality. Thankfully, this large TVL inflow onto the short side presents a very profitable opportunity:

## Mechanism

This "funding fee" that is paid by the overweight side to the underweigh side presents an opportunity known as Skew-Farming (or Basis-Farming). To illustrate this, lets show a real example from the 10x leveraged pools:


In this extreame example, The TVL on the Short-side ($79,982) serverly outweighs the TVL of the Long-side ($39,798). As a result of this, the effective gains of the short-side winning in this epoch would be only 4x, notice that the effecive losses are always 10x in this pool however. On the other hand, effective gains of going long is 24x but with only 10x losses. It should be immediately obvious how this assymetric payoff on going long can lead to a very profitable trade.

The basic idea would be to go long in this pool, and then to hedge you 10x exposure to ETH from this pool elsewhere (such as on a CEX). If the price of the underlying ETH went up, the wins from the long would be 24x the movement in the underlying whereas the hedged short would only lose 10x. This would lead to a completely delta-hedged payoff.

It should be noted, since Tracer only recently released their v2 pools, the data on the returns of these types of strategies is very limited. However, if we define:

Skew = Long_TVL / Short_TVL

Then we can see the skew history as:


This shows that, more often than not, the pools are extreamly skewed which provides insane opportunities for a skew-farmer. You can imagine that at Skew > 1, Umami will be able to increase their TVL which will again shift the skew in favour of the long-skew-farmer (profitable when Skew < 1).

Two additional important mechanisms:
- Volatility decay - This is a problem for the v1 pools but the v2 pools have heavily reduced this by taking simple moving averages of price moments over time, thus favouring the skew-farmer who takes a positions and maintains if over a few epochs (8 hour intervals). This would be the case for a partnership with Umami, since Umami would be able to notify partners when they were about to increase TVL caps on their vaults, and thus be adding significant sticky TVL to the short side of the Tracer pools
- P&L Compression - This is a complex mechanism which is designed to make skew-farming more profitable and follows a max-sigmoid function. The idea behind it is that, under high volatility within an epoch, the gains from the winning side and the losses of the losing side are dampened. This means that under high volatilty, even if the skew farmer loses on the Tracer pool position, the losses will be damped and thus the overall position will still generate return. In the above example, the losses could get dampened from 10x to only 8x on the long tracer pool, meaning if the price goes down the tracer pool losses only 8x but the external hedge will win 10x!

## Strategy

There are multiple strategies to take advantage of skew farming, but i'll focus on the delta-minimised one here. The step-by-step strategy would be as follows:

**Step 1**: Get the information
Go to https://pools.tracer.finance/

**Step 2**: Find the best yield
Umami will help with this by directly communicating when they will increase their TVL caps.

**Step 3**: Acquire the token benefitting from skew
Immediately mint the leveraged token which benefits from the skew. There is an 8-hour wait before you receive the new tokens, so ensure you consider the time until your exposure begins. This should all be done while in contact with Umami to make sure the timelines match up from the TVL Cap increase (Umami faces this same 8-hour restiction).

Note: Exposure begins when your collateral is processed (at the end of the 8 hour commit window). Do not hedge your position until your tokens are created and show up on the interface in the Portfolio tab.

**Step 4**: Hedge after receiving tokens
Calculate the trade needed to offset the Perpetual Pools position. The hedging trade will be the value of the position multiplied by the market’s leverage e.g. $1000 of 3S-BTC requires a $3000 long position to hedge.

**Step 5**: Rebalance regularly
Your position size will change over time. Repeat step 4 to offset any additional exposure you take on. Rebalancing frequently is advised (once a day in most market conditions) but is up to the trader. Higher leveraged markets may require more observation.

**Step 6**: Be aware of skew changes
The partnership with Umami will help in this since they will want to increase their TVL Caps if the skew ever benefits them to do so. You can also be almost guaranteed that the sticky TVL in the vaults will mean that the short position from Umami will rarely change much. However, this is still an important thing to keep in mind.

## Additional Resources

Great overview of Tracer:
https://twitter.com/Riley_gmi/status/1534976626206707712

Full explanation on Skew-Farming as a concept, also explains p&l compression well. Note: this article relates to the v1 pools so some things are out of date, but the main changes are the new SMA for volatility decay minimisation and pnl compression now uses the Max-Sigmoid function.
https://tracer.finance/radar/skew-farming-explained/

Most interesting data from the above article:
- Pnl Compression returns - as you can see here, Pnl compression **alone** can lead to some great returns (backtested model). Note: this is from v1 pool function.

- Skew-Farming returns under different skews and slippages of hedging - Note: this assumes hedging based on spot, thus leverage will only increase these return. However this is also based on v1 data, since v2 is too new to generate any concrete data on skew-farming returns yet. The Tracer team is working on it currently!

More detailed guide on how to skew-farm on v2 Pools, including a different strategy based on positive exposure to delta:
https://tracer.finance/radar/v2-skew-farming/