Futures Basis spreads across exchanges as market makers reassess risk and implied default premia.
Credit risk (as well as counterparty risk, etc.) is a tough topic in general, and in crypto even more so. In traditional finance we have many years of history and clearer regulation that make it “easier” to evaluate some tail risk events.
In crypto we have none of the above.
Indeed, the FTX drama highlights – among other things – two main topics:
- Not your keys, not your coins
- Futures (and derivatives more generally) are non-fungible
While real assets (i.e. BTC, common shares, etc…) are fungible assets, derivatives are not. Derivatives, by design, are financial contracts set between two or more parties.
To clarify: two standardised futures contracts with the same specifications (underlying assets, maturity, margining, etc.) but with different counterparties (i.e. Binance and Kraken) are NOT the same instruments: you cannot sell (buy) the futures on Kraken and buy (sell) them on Binance.
Historically, the futures basis across exchanges were almost identical (+/- some spread due to withdrawal/deposit costs and different margining systems) as market makers and arbitrageurs were playing the spreads. Today, this is no longer the case, and futures prices across exchanges are diverging.
Chart 1: BTC 3m Futures Annualised Rolling Basis – source: Laevitas.ch
To keep a position open (whether long or short) you need “some” collateral on the exchange (Initial Margin + Maintenance Margin); if the exchange halts withdrawals your cash balance is gone (i.e. 100% loss).
The futures prices are deviating according to the counterparty risk involved. Specifically, riskier counterparties will have higher prices for two reasons:
- Market makers take more spread for their inventory risk (risker = expensive)
- Traders play the counterparty/default risk by entering long positions on riskier counterparties, and short positions on the safer ones.
Let me deep dive a little on the 2nd point:
Initially, by entering the “same” future contract on different exchanges, you are playing the basis spread.
If your bias is a default event of an exchange – but you still want to have some hedge – it is reasonable to go long on the risky exchange and short on the safer one. In case of default of the risky exchange, the loss is the cash balance on the exchange. Expecting a general bearish sentiment following the bankruptcy of a big player, your short position on the other exchange will gain at least as much your losses.
The trade is profitable if then the profit on the credit risk free exchange is greater or equal to the cash balance on the defaulting exchange. Assuming a leverage of 10x, the short position should earn at least 10%.
The current 3m rolling annualised basis for BTC futures across different venues is as follows:
- CME: -7.7%
- Kraken: -7.25%
- Bitmex: -5.87%
- Deribit: -1.75%
- Bybit: -1.72%
- Okex: -1.7%
- Binance: -1%
There are multiple factors that weigh on these spreads (i.e. index constituents, margin requirements).
But, as in the past, the spreads were almost 0, or way tighter, so we can assume that the premia are the implied probabilities of default. Adapting the futures pricing formula, given a risk-free of 4% and assuming the CME will not default, the implied probability of defaults within 3 months are the following:
(Formula and reasoning in the attached file)
- Kraken: 0.45%
- Bitmex: 1.83%
- Deribit: 5.95%
- Bybit: 5.98%
- Okex: 6%
- Binance: 6.7%
Whether we will see other exchanges defaulting or not, I believe that we will not see futures spreads close to zero anymore, and to me, this makes perfect sense. And it is very likely that the dust raised by the FTX drama will take some time to settle.
Nevertheless, I believe there is some space for some trades here (not risk free for sure).
My bias is that Kraken is not much better positioned compared to the other exchanges (also from a technological point of view). So, long Kraken and short Binance or Deribit makes sense to me.
