Dec 23, 2021

7 min read

Temple AMM resumes

A negative feedback loop in the unlocking process has been identified and trading was paused to resolve it.

Fortunately, the exit queue did its job slowing down actions, providing time to find a resolution. We inadvertently introduced a dynamic we could not let continue as it promoted a zero-sum selling behavior that was not aligned with the sustainability of the Protocol. The code works for templars, not the other way around. As such we are treating this as a mechanic bug and one we are working diligently to resolve

We have three perspectives in our mind as we listened to all feedback from the community and advisors to evaluate the pathway forward as a team:

  1. Any change we make will disadvantage some people and advantage others.
  2. Some changes are better in alignment with our principles, our stated intent for the AMM launch, and the long-term health of the protocol. We are attempting to prioritize the long-term benefit of the Temple.
  3. We were conscious of changing the rules of the game for participants e.g. the changing the order of unlocking in the middle of the process.

The post outlines the problem and the solution.

The problem

In short: The current dynamic will strongly disincentive buying $TEMPLE for over a month, while the Treasury value is drained at unnaturally high prices.

This is our miscalculation. We set the parameters of the system up for an equilibrium state and didn’t correctly take into account the unique dynamics at play during this unlocking period.

This is how it occurred:

First, we designed the Opening Ceremony such that the first buyers would get both a better price and unlock first. We should have done this in reverse order like the Fire Ritual — buy first for the best price but unlock last.

Second, we delayed the launch of liquidity until the custom AMM was ready. This meant the buy pressure immediately after the Opening Ceremony wasn’t captured, so the price remained at the level of the Opening Ceremony.

Third, we were too quiet over that time building, and in the meantime, the market grew more bearish.

As a result, early selling pressure overwhelmed the buying pressure.

Once the selling pressure took hold, everyone could see the queue growing. It was doing its job, slowing down the selling pressure and giving us time to react. In more stable conditions, it may have been appropriate to slow things down and disincentivise unstaking by stopping APY while in the queue. However in this case it mainly provided a very tangible sense of the future selling pressure because everyone could see the exit queue getting longer and longer.

Once people could see a large queue, there was no reason to buy until the selling pressure in the queue emptied. If you think the price is going to go down, you’re better off waiting until it does than jump in and get dumped on.

Therefore, the dominant strategy became ‘join the queue as soon as you can to preserve your order, and sell as soon as you’re able.’

This strategy is best for those wanting to sell or stay. If you were a believer but willing to trade: queue, sell, and once the queue empties, buy back at a much lower price.

The problem is that this became a negative feedback loop. As more people took this strategy, the queue became longer, encouraging more people to take up the strategy.

If we did not intervene and everyone joined the queue, it would have grown to 75 days in length. Many were choosing to stay staked and chill. But the queue would have likely grown to over a month, and during this time, it would make no sense to buy — only wait until the queue drained.

To make matters worse, we launched with a very large amount of liquidity. Post this unlocking phase, a large LP would be a real benefit for stability.

However, during this negative feedback loop, it kept prices unexpectedly high making it profitable for everyone to join the queue and plan to sell.

Over the course of a month, there would have been no buying pressure, and FRAX slowly drained from a vast pool.

We have no problem allowing people to take exit liquidity, and we expected that the price would be volatile during this price discovery phase. But this specific outcome was unintended by all, misaligned with Temple’s principles of fairness and stability, and frankly unsustainable. We had run into the classic “bank-run” scenario.

We wish that we had the foresight to exclude or quickly identify at least one of the five ingredients above, so the negative loop did not take hold.

The team will always do its best to act boldly in alignment with the Temple principles and serve all Temple holders.

The solution

A solution needs to include the following elements:

First, Preserve the Deal Made for the Opening Ceremony.

We have heard this mentioned a lot in the last 24 hours and we agree — this is absolutely fair. For this reason, it is better if ordering prevails.

The Temple honors its commitments, and even if we think in hindsight a different unlocking order would have been better — we don’t want to change terms this late in the game.

Therefore, trading will shortly resume, and the exit queue will preserve ordering.

Second, Speed up the Process of Letting Sellers Exit.

The exit queue has done its job of slowing things down so TempleDAO can respond. However it is not good to extend the process of price discovery out by over a month.

Due to the nature of the smart contract, we cannot accelerate people’s unlock times to claim their OGT tokens. The policy lever we do have is to speed up the exit queue to reach equilibrium shortly after the last people unlock — rather than many weeks later as the exit queue continues to process.

One of the mechanisms we looked deeply into was an OTC market on the exit queue, which captured the imagination of many. Unfortunately, it became very challenging to allow this to occur and accelerate the queue when people sold in our current queue design without prohibitively expensive gas. Without the queue accelerating as people sold OTC, the feature did not achieve its needed aims. Instead, we will just go with the fast queue.

Third, Implement the Faith / Buy the Dip Game.

This is near implementation and will be released shortly.

We don’t expect it will achieve its full impact for incentivizing buy pressure until everyone is unstaked and we reach equilibrium. Still, we will launch it soon regardless so we can start seeing how it works, marketing its function, and calibrating its settings.

And be ready for it to do its job once we reach market equilibrium.

Fourth, Temporarily Reduce the Depth of the Liquidity Pool.

No one expected Temple to launch with a liquidity pool roughly equal to our market cap. At $300m it was 5–20x the proportion of market cap to other serious launches (and 100x those that launch with tiny liquidity).

The thesis behind such a large liquidity pool was that by providing superior price stability we would attract buyers interested in the unique Temple offer: far more stability than you would expect for a coin with high yield.

This attraction would drive buy pressure, which grows Treasury through our custom AMM, and supports protocol health and runway. This may be a sustainable revenue strategy, earning income for treasury as a result of providing price stability (similar to a successful stablecoin protocol, though with a different risk/return).

This strategy may provide far more revenue than investing the FRAX in yield farming directly, and could be boosted with staking the LP tokens for FXS emissions.

This is a new strategy in defi on a few fronts, and remains untested because of the problems identified earlier in this article. We were not yet able to get fair market participation because of the exit queue, unlocking dynamics, and early sell pressure. Once we reach equilibrium, we can put this strategy to work.

However, to get through these next two weeks of unlocking and distorted incentives, the proposal is to significantly reduce the liquidity pool down to more typical levels to accelerate price discovery and resume growth.

We will relaunch the AMM at the current price of ~$3.48 with liquidity depth of $30m.

We believe this is fair for three reasons:

  1. The price on launch nor the liquidity depth was part of our Opening Ceremony contract, there has been no commitment to our current levels of liquidity
  2. We acknowledge that some people were able to participate in the high liquidity, but believe it is better to stop a dynamic that was detrimental to the long-term health of the Protocol as fast as we can rather than leave it
  3. During this time of volatility and unlocking, lower liquidity will allow faster price discovery. Once we find the right level, liquidity can be added if we choose.

The best action here for the protocol is to return the liquidity pool to a more typical level for the next two weeks to let unlocking and price discovery occur, add it back over a period of time when price stability is desired.

We have engaged in a wide range of discussions to assess the appropriate level to set liquidity. The smaller the liquidity, the fewer people that are about to unlock will be able to do so at higher prices. This preserves the Treasury value without which the Protocol cannot grow and protects stakeholders with a long-term outlook.

This preserves Treasury value, though it will lead to greater price volatility in the next two weeks. This in turn will put to the test the other mechanisms in place within the TempleDAO protocol, in both directions.

Next steps

We will be immediately reducing liquidity and resuming trading so as to return to the commitment made during the Opening Ceremony.

This ensures ordering is best preserved, and an expected amount of liquidity is there to support any exits people choose to make.

Over the next few days, we will complete testing on the Faith/Buy The Dip Game, as well as an accelerated exit queue, and will launch both as soon as possible.

We truly appreciate the community and all of the feedback from stakeholders. Templars have been constructive, thoughtful, and supportive in a time of rapid change. The Temple wouldn’t exist without you.

We have taken on many lessons here, and will rebuild stronger.

Thank you for your faith and patience.

Temple Masters