Executing Futures Spreads As Block Trades With Paradigm and Deribit

Published On

March 25, 2021

Article by
Cryptarbitrage
CryptArbitrage is a well known personality in the crypto derivatives space. He is a Deribit Insights content creator and has experience trading all kinds derivatives in both crypto and traditional markets.

Deribit recently introduced a weekly and bi-weekly futures contract for BTC and ETH, in addition to the existing three quarterly futures (and soon to be released Monthly Futures). This means there are now up to 5 (6) different futures expiries available at any one time. Not only does this give option traders more choice when hedging, but it also means there are more futures spread trades now possible.

A futures spread trade is where a trader buys one future, and sells another future with a different expiry date. Rather than trading the price of the underlying asset directionally, the trader is betting on the price difference between these two contracts. This action of buying one future and selling another is also how to roll a position from one future to another.

Once you’re set up on both Deribit and Paradigm, with your Deribit API key added to your Paradigm account, the quote and execution process for futures spreads is fast and simple. Both legs are executed simultaneously as a single order.

My usual Cash & Carry Trades

If you follow me on Twitter @cryptarbitrage you’ll be aware I like cash and carry trades. Briefly, a cash and carry is a trade that captures the premium of a futures contract over the spot price, with no directional preference. For more on this, you can see my post on the Deribit blog here: insights.deribit.com/education/cash-and-carry-trades/

When rolling these trades to the next contract, the trade executes exactly like a futures spread in that it’s two legs, one buy and one sell. I’m typically buying back the front month that I’m already short (that now has little or no premium remaining), and simultaneously shorting the next future that still has plenty of premium to capture.

The combination of these two legs moves (or rolls) my short position to the next futures contract. The current minimum size for futures block trades on Deribit is $200,000. When executing a two-legged spread though, this of course means it can be as low as $100,000 per leg.


Rolling The Cash & Carry To The Next Contract

So let’s say I have a cash and carry with a $125k short on the current front month of March. March is about to expire and so has almost no premium, but the next quarterly (June) has a nice rich premium I would like to capture next. So I want to roll from March to June. To execute this roll I need to buy back my $125k short on March, and simultaneously sell $125k on June. (In practice I would actually short a little more on June to include the extra premium I’m capturing, but we’ll keep things simple for this example)

If I’m feeling patient I can stick a limit order for one side into the orderbook, wait for it to fill and hope the other side hasn’t moved too much and market into the second side. But this can be time-consuming and I don’t know how good the fill will end up being.

Alternatively, I could just market order into both sides straight away. The fills will be quick and I’ll be able to see roughly what prices I’ll be getting in the orderbook, but the prices obviously won’t be optimal. The larger the order, the less optimal the prices will be.

Put simply, I can execute this roll on screen, but not in a single order and not for a guaranteed price. By using block trades via Paradigm though, I can. And for block trade sizes, this will almost always result in a better price as well.


Live Price Example Of Rolling A Cash & Carry

To illustrate why using Paradigm for this is so attractive, let’s look at a snapshot of how rolling from March to June is priced, both on screen and via Paradigm.

Remember we have a $125k short on March, and want to roll it to June. So we want to buy $125k of March, and sell $125k of June. Market Orders for this will get us average prices of:

  • BUY 125k March Future: $56,979
  • SELL 125k June Future: $60,355
  • Future Spread Difference: $3,376

Executing on screen then moves our short from March to June, and our effective entry up to $3,376. The on screen bid/ask for this size is effectively $3,376/$3,540. How does a Paradigm RFQ compare? Here is the RFQ form with details of the futures spread entered:

1 — Deribit account selected; 2 — Liquidity providers selected; 3 — Futures Spread strategy selected; 4 — Leg details entered

We’ve entered a March/June futures spread with a quantity of 12,500x contracts for each leg. The front month is always displayed first, and we’ve left the default setting with the futures spread preset of selling the closest month. (We actually want to do this the other way around so we’ll be selling into the bid on the next screen instead of buying the ask)

After clicking Send RFQ we receive the quote:

The bid/ask we receive for our future spread is $3,464/$3,470 ($6 spread). Compare this to the on screen bid/ask of $3,376/$3,540 ($164 spread), and you can see this results in a price improvement in both directions of $88 and $70 respectively.  This means the bid/ask spread is reduced by about 96% by using Paradigm in this example.

As mentioned we’re selling into the bid here, which results in us moving our short position from March to June, and our entry price up to $3,464. This is a price improvement of $88 compared to on screen. And this is just above the minimum size for a futures block trade. With larger size this improvement is often even greater due to the extra slippage you would get on screen.

If you wanted to execute larger size, you would see price improvements for either side of:

Futures Spread: BUY 26MAR21 / SELL 25JUN21

Futures Spread: SELL 26MAR21 / BUY 25JUN21

This price improvement is not the only advantage. We have:

  • Price improvement compared to on screen
  • Guaranteed price/no slippage concerns
  • Execution of both legs in a single order

There is also no extra fee from Paradigm for executing in this way, so there’s really no reason not to use it if you can.