1 month performance analysis of 3L-BTC and 3S-BTC

I recently completed analysis on 1 month of 3L-BTC and 3S-BTC token performance and compared to both FTX and Traditional Leverage rebalanced daily.

TCR rewards are ignored for this analysis. TCR rewards have added roughly 8% to the PnL of every Perpetual Pool user. If you want to read the below data with the TCR rewards included, add 8% to Perpetual Pools performance.

The last analysis posted only to discord was with the first 2 weeks of data where the average Tracer Perpetual Pools user ended up with 17% more money than FTX users. In this second analysis with a month of data FTX outperformed Perpetual Pools leaving the average FTX BTCBULL/BTCBEAR user with 6% more money than the average Perpetual Pool user.

Why did things reverse from having Perpetual Pools outperforming both FTX and Traditional Leverage to the opposite in the last 2 weeks if Perpetual Pools has less volatility decay?

I suspect the answer to this lies in how volatility decay works. For funds that must buy and sell the underlying asset or a derivative of it to rebalance to target leverage, these funds tend to sell low and buy high. This tendency leads to a loss over time. In contrast, Perpetual Pools do not have to sell low and buy high because it is not built on top of another asset, and is a new financial primitive.

Still our simulations show that Perpetual Pools do have a decay about half that of traditional leverage. This decay likely comes from how price movements affect the timing of dilution of shares. When the underlying price rises, money is transferred from the short pool to the long pool, incentivizing money to enter the now undercapitalized short pool. This dilutes shares of the short pool and now if the price reverses and falls, the profits from the fall are shared among more share holders.

If the underlying price were to have sideways volatility, both traditional leverage and Perpetual Pools would decay, but the latter would have less decay because dilution has less impact than buying/selling. This is a good thing overall because sideways volatility is the norm.

However, the feature of volatility decay can sometimes be beneficial. In an up only market where the large portion of moves are up, traditional leverage is consistently able to buy at what would be a high price only for that high price to become a good deal shortly after. This is what happened on the latest BTC bull run to all time high.

As you can see here, during the 2 week bull run for BTC prices, there is almost no price to buy in at that was a bad price even if you had to sell 24 hours later. What is normally a downside for traditional leveraged assets is actually positive in this case leading to exponential returns. During this timespan the underlying price increased by 30%, but a traditional 3x leverage trader rebalancing once per day ended up with gains of 109%. Sometimes habitually buying high gives returns higher than the leverage used. In this case the traditional leverage user gained as if they were on 3.63 leverage despite staying at 3 leverage due to exponential growth.

Overall, Perpetual Pools is expected to outperform traditional leverage according to simulations, and did indeed do that for the first 2 weeks of operation. In some short term samples though, traditional leverage will outperform Perpetual Pools when buying at high prices actually pays off.

In total for this month selection, FTX outperformed Perpetual Pools by 6% as measured by total holdings from buying equal amounts of long and short on day 1. Traditional Leverage users outperformed Perpetual Pools by 7%. The TCR rewards more than covered relative losses, but as stated earlier, this is being ignored for this analysis.

Now for the actual numbers:
Timeframe: September 15 @16:00 CST – October 18 @18:00 CST

FTX leveraged BTC price source:

Traditional 3x Long BTC starting price: 48139
Traditional 3x Long BTC ending price: 88540 (+84%)
FTX BTCBULL starting price: 9454
FTX BTCBULL ending price 16868 (+78%)
Tracer 3L-BTC starting price: 1
Tracer 3L-BTC ending price: 1.464 (+46%)
Simply holding 1 BTC starting price: 48139
Simply holding 1 BTC ending price: 62039 (+29%)

Traditional 3x Short BTC starting price: 48139
Traditional 3x Short BTC ending price: 16109 (-67%)
FTX BTCBEAR starting price: 0.00024061
FTX BTCBEAR ending price 0.00008527 (-65%)
Tracer 3S-BTC starting price: 1
Tracer 3S-BTC ending price: 0.552 (-45%)
Simply borrowing 1 BTC starting price: -48139
Simply borrowing 1 BTC ending price: -62039 (-29%)

Outcomes when starting with $100 invested in both short and long position:
Traditional Leverage Dual Hold: 217
FTX 3BTC Tokens: 214 (-1.4%)
Tracer 3BTC Tokens: 202 (-6.9%)
Spot 1 BTC + Spot -1 BTC: 200 (-7.8%)


So, why did the 3S-BTC perform so well relative to the 3L-BTC tokens? The short answer is that the effective interest payments from collateral imbalance were very high during the first month of Perpetual Pool operation.

Why were they high? A few factors played into this, millions of dollars a day were flowing into the protocol mostly going long causing high demand for market makers, meanwhile, market makers (those who buy short tokens and hedge with spot exposure) were unable to keep up. They were not able to keep up in part because this is a new financial primitive and whales prefer to see empirical proof of high yield not just simulations. Market makers also do not have bots setup and the overwhelming demand for long was being met with people manually farming the interest rates in their spare time. Large demand to go long for a market approaching all time high coupled with a supply shortage led to not enough 3 short tokens being minted despite the incredibly high interest that they were earning. With not enough people competing to keep the interest rates low, they ended up quite high.

Relative to traditional leverage users, 3L-BTC holders lost about 21% of their position. Relative to traditional leverage users, 3S-BTC holders gained about 67%. A large part of this bias in favor of shorts is due to the effective interest rate that comes from polarized leverage.

While I have above pointed out that when measured by dual hold that is never rebalanced back to delta neutral Perpetual Pool 3BTC users ended up with 6.9% less money than traditional leverage users. You may be quick to notice that this is not a fair metric though almost no one uses leverage this way. They pick short or long only. If they want to be delta neutral they hedge elsewhere and rebalance to remain delta neutral.

While Perpetual Pools definitely had a higher rate of effective payments from long to short in the first month, you should be able to tell at a glance that one group of users ending up with 21% less money than they would have had with traditional leverage and another group ending up with 67% more money than if they had used traditional leverage is overall a good deal.

If you would like to be a part of earning the high interest rates that have been paid from long to short, the chart below from a 2.6 year simulation of the estimated APRs that come from various skews will be helpful. You still have time to collect on these high APRs, but they will come to an end once whales get the empirical data they need to trust the returns and setup bots.

8 Likes

This is why I am LONG Tracer

1 Like