Initial Margin Monitoring (Crypto)

[alen.pavlovich]

Hello everyone,

I'm junior trader in a crypto fund. One of the strategies that fund is running is the cross-exchange arbitrage capturing the funding fees between the two swaps using leverage.

At the moment approach is super conservative but the idea for the future is to loosen that a little bit with the proper tools. I was looking into the literature how to monitor inital margin requirements.

Anyone has any ideas, tools or papers that I can use for reference? Best idea I have so far is to calculate volatility every hour or so and apply on the initial margin requirement.

 

[CrossGamma]

Hello everyone,

I'm junior trader in a crypto fund. One of the strategies that fund is running is the cross-exchange arbitrage capturing the funding fees between the two swaps using leverage.

At the moment approach is super conservative but the idea for the future is to loosen that a little bit with the proper tools. I was looking into the literature how to monitor inital margin requirements.

Anyone has any ideas, tools or papers that I can use for reference? Best idea I have so far is to calculate volatility every hour or so and apply on the initial margin requirement.

This is a pretty crowded trade already since it's so easy to execute. Being more aggressive in margining would increase profitability but also introduces significant tail risks. A few things to consider:

1. A short term historical volatility measurement as you suggest will not capture the tails. Your margin needs to suffice in the 99.99% tail event, not the average realization.

2. Looking at individual token distributions is not sufficient but you need to model the joint tail behavior of your portfolio unless you run only isolated margin positions, which would generally require more capital. Especially large down moves are strongly correlated. You can look into e.g. copulas for this.

3. Apart from market ineffiencies / varying demand for leverage, there are a lot of other reasons for the persistent funding differences across exchanges that are globally accessible (i.e. not only open to individuals of certain nationalities). These include e.g.: (i) The risk of exchange suspending deposits or withdrawals temporarily which prevent you from moving collateral. (ii) The exchange going down during large moves and preventing you from trading out. (iii) The books getting very thin during large moves. Each exchange uses different reference prices for their margin calculations, typically including their own spot market when available. Different exchanges will thus see potentially very different highs / lows during liquidation spikes. You’d need to put a price tag in terms of expected losses / returns on these to evaluate whether the trades are still profitable afterwards.

 

[alen.pavlovich]

Would you be kind and suggest any books or materials where I can learn more about this and potentially implement some items?

 

[CrossGamma]

Would you be kind and suggest any books or materials where I can learn more about this and potentially implement some items?

My previous post has plenty of searchable keywords. This is a relatively new market and thus you won’t find any recipes for how to approach these kinds of problems.