Seasoned Speculator
Expressing Long-Term Views With Less Risk Of Getting Liquidated
Jul 8, 2024
/
Copy link
Traders get access to digital asset prices either via:
1) Spot (no leverage)
2) Perpetuals (w/ leverage)
Perps have been the preferred vehicle to amplify or fully implement your conviction.
However, with wild daily swings in asset prices, traders are limited to either:
a) spot trading and capping their upside
b) running a high risk of liquidation
What’s happening in practice is that up to 98% of traders using leverage get liquidated when trading perps.
But what if you could shield yourself from this liquidation risk? Whilst maximising the potential upside for your view?
Example of Liquidations
Consider the SOL rally over the past 1 year (365 days).
The price has appreciated +611%, despite the recent correction.
Holding $1000 of spot to today would have returned you $6110. But not everyone has the cahones to hold, so lets say average return is $3000 for most who entered and exited spot at some point during the rally.
Clearly, if you were bullish $SOL, it would have made sense for you to use leverage to amplify your views. Let’s say with 10x leverage, and 600% return on SOL, the investor with $1000 stands to make $60,000. Now that’s more like it!
However, the problem here is that you can have the right view, but get rekt. This is depicted in the SOL Liquidation history below.
As per the data, there was >$500 million of long liquidations over the past year, which means lost capital for those who should have been right, and $billions of foregone profit.
For example, on the week of March 11th, just before SOL broke out above $200 in price, these leveraged longs on Binance lost $30.35m of collateral due to some intraday moves to the downside.
Instead of doubling their $30.35m of collateral when the market moved another 10% (with 10x leverage), they lost ALL their margin due to liquidations.
There is nothing more frustrating than having the right view that SOL will outperform, but missing out.
Pairs-trading is a way of hedging your market risk (beta), and shielding yourself from liquidation risk, whilst still holding leveraged upside exposure to SOL.
Liquidations More Generally
We decided to zoom out and look more closely at Hyperliquid traders.
https://stats.hyperliquid.xyz/
Of the $2.4bn of total user deposits over the past 1 year, guess how much of that was liquidated?
They are transparent about this - $3.2bn!
That means >100% of deposits was eventually liquidated. How? Traders win on the way up, they have unrealised gains, and then lose their gains and their collateral in short term wicks during market turmoil.
If we extrapolate out to other CEXes like Binance, Bybit and OKX those numbers run >$100bn.
How can we keep our gains when you have the right view?
Let’s take a deeper look.
Pair Trading
Pair trading is a new financial primitive conceived to enable investors to leverage their views while minimizing the risk of short-term liquidation due to systematic market movements.
These short-lived but impactful blips frequently trigger waves of liquidation that further snowball across the crypto market.
The short-term protection arises because of the greater short-term correlation between crypto-assets: over short horizons, most risk assets move in tandem (aggresively down), while over the long-term, one asset tends to outperform the other - leading to opportunities to capture the relative value. One way to achieve this is to simultaneously short a correlated asset like $ETH, which is a good proxy for market risk. This is denoted as being long SOL / ETH.
On one of these down days, if SOL is -5%, and ETH is -10%, then you’ve made a net return of +5%. With 10x leverage thats a +50% move in a single day. Your collateral is cross-margined, meaning that the positive PnL on the short eth leg is used to offset the negative PnL on the long sol eth. This way we get to stay long SOL even during the downturns, whilst still maintaining 10x leverage.
Understanding What the Short-and-Long-Term Dynamics of Asset Prices
The price of cryptocurrencies can be dissected into two main components:
- Systematic Component: Market-wide influences affecting all coins.
- Idiosyncratic Component: Specific factors affecting individual coins.
Over the long term, idiosyncratic factors dominate — companies outperform because of their “fundamentals”.
Conversely, in the short term, systematic factors prevail, with markets seemingly overreacting to news. This apparent overreaction is not due to irrational investor radically altering their assessment of fundamental value, but to changing risk aversion, and to the impact of flows (which can snowball into liquidations)
The Time Structure of Correlations
Empirical data shows that correlations between coin prices vary with time frames:
- At lower frequencies, correlations are typically weak (less than 40%).
- At higher frequencies, correlations surge (often exceeding 90% for 1-minute ticks).
Systematic jumps, particularly those impacting Ethereum (the crypto benchmark), propagate quickly across the market. However, the reverse is less true: there is a greater probability that a shock affecting a single
As is well known (see e.g. Pindyck, 1984), correlations increase with volatility. This also explains why over the recent highly volatile period, high-frequency correlations have been particularly high.
Graphs
Image 1: one thousand weekly returns
At low frequency, correlations between blue-chip crypto assets are typically less than 50%; they are close to zero between blue-chips and meme coins
Image 2: one thousand daily returns
At medium frequency, correlations between blue-chip crypto assets are typically around 70%; they are close to 50% between blue-chips and meme coins
Image 3: 90 days of 1 minute returns
At higher frequency, correlations become quite high across the board between similar assets
Image 4: 30 days of 1 minute returns
Overly volatile periods such as the last 30 days (June 2024), one-minute correlations are very high across all assets. It is during these volatile periods that the most dramatic liquidations occur, and that pair trading offer the greatest protection.
Image 5: 30 days of conditional jump corelations (over 1 minute returns)
One can also measure correlations conditional to a jump happening on any given asset — this table measures correlation only when a jump ( one minute return greater than 3 standard deviations of returns ) occurs for the asset the row index.
During volatile periods, these correlations are extreme too. This is called scam-wick protection.
Conclusion: Scam-Wick Protection from Pair Trading
Pair trading mainly enables investors to shed themselves from short-term liquidation risk that arises from the frequent extreme downside movements in crypto-assets.
During times of volatility, correlation typically goes to 1:1, allowing gains on the short leg to offset losses on the long leg. Whilst in the good times, your long leg outperforms the other asset because they are sufficiently de-correlated on longer time horizons. Pair trading leads to a more controlled trading environment, providing some degree of liquidation protection. We call this scam-wick protection, and it’s something that is available through our dynamic cross-margin pair trading engine powered by SYMM.
-----
Article written by mazett and huf
-----
Academic References
Andersen, T.G., Bollerslev, T., Diebold, F.X., & Ebens, H. (2001). Journal of Financial Economics, 61(1). The Distribution of Realized Stock Return Volatility.
Andersen, T.G., Bollerslev, T., Diebold, F.X., & Wu, J. (2005). American Economic Review, 95(2). A Framework for Exploring the Macroeconomic Determinants of Systematic Risk.
Barndorff-Nielsen, E., and ,Shephard N., (2006). Journal of Financial Econometrics 4(1).
Econometrics of Testing for Jumps in Financial Economics Using Bipower Variation
Bollerslev, T., Tauchen, G., & Zhou, H. (2009). Review of Financial Studies, 22(11). Expected Stock Returns and Variance Risk Premia
Bandi, F.M., & Perron, B. (2006). Journal of Financial Econometrics, 4(1). Long Memory and the Term Structure of Volatility: New Evidence from the French Stock Market.
Pindyck, R.S. (1984). National Bureau of Economic Research. Risk Aversion and Determinants of Stock Price Volatility.
Schwert, G.W. (1989). Journal of Finance, 44. Why Does Stock Market Volatility Change Over Time?
