I tend to shy away from generalities when it comes to trading, as everyone is different and everyone's trading system and style is different. If someone has a trading system that is consistent over a protracted period of time and it works for them, I would be the last person to tell them to change it.
I traded ES versus NQ this past Spring and early Summer of 2020. It was the first stock index spread trade I took in about 18 months - it has to be a
really compelling setup for me.
I would estimate that over 90 percent of the trades that we are doing are
intra market exchange supported. So that is the same product, different expiries. So there is a singular order book price ladder and the spread leg components are internally matched and filled by the exchange. CME, ICE, Eurex - all the regulated electronic exchanges do this. We also stay away from the prompt months, so we are active in the forward curve. So the slippage and partial fill mistakes are a non-issue for exchange supported spreads.
Having my clients
stay away from the prompt months for intra markets tends to minimize the catastrophic physical distortions like what we saw in May '20 Crude and June '11 Cotton.
I completely agree with you about the dangers of electronic arbitrage.
We are trying to take large chunks of a trading range out of these, so we are hanging on to them as the modeled trading range for a specific spread combination dictates. These are
really cheap to margin overnight so we are modeling them thoroughly and we are holding these (swing trading) and
not arbing them.
Let me give you a couple of examples of trade holding timeframes:
Unleaded Gasoline Butterfly: 90 days, + $1,264.20 on a 1-2-1.
Eurodollar Condor: 63 days, + $1,181.25 on a 1-1-1-1
It would not be unusual for me to hold a Eurodollar Spread for 4 to 6 months - most of them have only a 2-1/2 - 4 tic daily trading range.
On the other hand, a Gasoline Crack is usually held for days. That thing can really move in the late Spring and Summer.
The holding timeframes are dictated by the targets we set on the modeled trading ranges.
Generally speaking, the majority of our
inter market spreads (being the minority) are in the interest rates and energy and they are highly correlated. Examples would be CBOT Fives versus Tens, Eurex Schatz versus Bobl or Unleaded Gasoline versus Crude Oil - that type of thing.
I have some European clients who will spread Dow Jones Euro Stoxx versus FTSE - stuff like that; but even for them it is the exception rather than the rule.
YMMV, and I wish everyone good fortune!
Just my own opinion, but I have quite a bit of personal account, Commercial, and hedge fund experience trading futures spreads and OTC Swaps spreads. I've also worked with a couple hundred clients through the years.
i have what i think is a very extensive well thought out spread system and yet i don't trade it.
because when it gets down to it - it's more difficult and perilous than trading just one market alone.
mistakes, fill's, slippage, partial fills, catastrophic events and much more all have greater impact on spread, arb, pair trading. it's an idea that was spawned in the era of buying options gets you rich with limited risk. it's an idea that is almost as dead and flat as the profit curve of one big money manager that i know who uses arb.
the reason for using it would be if you have big money to manage and you are participating in commissions, in that case it's a way to generate income while basically parking the clients money in the market.
however if you are serious about spread trading no one did it better than goldberg.