I am supposed to keep some ES contracts a bit longer and I will probably have to roll them forward, from Mar to Jun. I can do a combo order, sell Mar and buy June, at once. I can see the price for the combo -.50 / .75 (bid / ask), and changing one or two ticks. Looking for some input from experienced traders: - is it a reasonable expectation to get a fill for zero limit? - if I know for sure that I have to go after Mar 20, when shall I do it, sooner or later? TIA.
You may roll your "position" forward, but you don't "roll the contract". You're "fretting over a tic-or-two". Unlikely you'll be on the right side of anything that small... if such a thing exists at all. Nothing to be gained by trying to do a "combo" order or trying to "time" it. Just close out your Mar and open a Jun.
Got it about rolling the position forward. My bad. About "fretting over a tic-or-two", 10 contracts would cost me $500 plus commission for a $1 fill. I am pretty sure that independent sell / buy orders would have slippage as well. But like I said, I am doing this for the first time and I could use help. Is there anything wrong in doing it as combo?
10 contracts is a $1.5 Million nominal position. Fretting over $500 potential slippage and $40 commission? I don't even know if there is such a thing as a "combo" order to try to save the slippage and commission. Even if there is, it's chicken feed compared to $1.5 Million.
If you have access to spreads - you can use calendar spreads to roll your positions very efficiently.
This isn't dumb, and remember the old adage 'look after the ticks and the points will look after themselves'. If you're trading pretty slowly (like yours truely), then roll costs becomes quite a large part of your total cost (at the limit of no trading you'd have 100% roll costs). Day traders don't need to care about roll costs because they close every night, but their normal trading costs will be much higher. My prefence when rolling is (assuming a 2 tick spread in both the roll and the outright market) from best to worst: - naturally close my March before the roll date (if I need to do a closing trade, do it in March. If I need to do an opening trade, do it in June). Zero roll cost, zero risk. - to do a spread combo order, passively filled (so basically a negative cost of -1 tick, zero risk) - do a spread combo order, crossing the spread (cost of 1 tick, zero risk) - do two seperate trades (in instruments where the spread market isn't liquid enough) (cost of between 2*1 = 2 ticks if I have to do this actively, cost of -2 ticks if I can do both passively, plus execution risk of course). - let the March expire and take the cash settlement. Then open up in June the next day (cost of 1 tick on the opening trade, but considerable overnight execution risk) GAT
The ES is "too noisy" to be concerned about "2 tics". You should have the mind set of "trying to risk a few points (not tics)" to make 3x or 4x of what you risk... something like "risk 4 to make 12 or more".... or even better, "risk 4 for 24".
There's an important difference between pre-cost returns and costs. The former have huge uncertainty, the latter do not. So you should always be trying to save costs, especially roll costs since this will have no bearing on your pre-cost returns (whereas trying to trade more slowly to reduce costs may affect your pre-cost returns). It's fair to say that the risk adjusted costs on ES are not especially large, but on something like Eurodollar they are extremely important. And even on S&P you are going to be throwing away about $50 a year per contract if you pay 0.25 every time you roll. That might not sound much, but it all adds up. Roughly speaking I'd pay 1% of my account value annually in extra costs if I just rolled the way you suggest. Seems stupid to deliberately reduce my returns by 1%... GAT