Stop vs Limit order for Entry, a Thought Experiment
I was considering methods of entries and specific order types. Comparing in my mind the benefits of each. I know some people favor the market order, but i never use it at all. I would say 99% of the time I am a limit order kind of guy. However lets give thought to using a stop order. Think about the following Hypothetical Scenario, it can be reversed for a short position. Assume that the instrument we are trading trades in ticks of 1$ for simplicity.
Scenario: You are looking to buy at a particular level 100. Price is currently above your level and moving closer, it is now in your zone for potential execution it is now trading at 101. You can place a Limit Order at 100, or a Stop order at 102. You have a fixed stop and target of 10 Ticks for each.
Outcome A) You end up getting filled on your limit order but price travels lower and never turns back. You take a Limit Stop Loss. However your stop order never gets filled. Limit Filled, Stop No Fill.
Outcome B) Price shoots up dramatically from 101 and never actually trades at 100. Or we trade at 100 but you never get filled and price takes off. Stop gets a Target, Limit gets nothing. Limit No fill, Stop Filled.
Outcome C) We retrace higher to 102 and then shoot lower past 99. Both Stop and Limit are filled. Assuming the fixed 10 tick stop, the Limit order has the advantage of 2 additional ticks of stop. Both Limit and Stop Order were filled. So, there are 3 sub outcomes for this. Both orders go to target. Stop gets stopped out, Limit does not and goes to target. Stop and Limit both get stopped out
Outcome D) We push lower to 99 and fill the limit and take off past 102 to fill the stop. Assuming the fixed 10 tick stop again, the limit order has the advantage of 2 additional ticks. Both Limit and Stop were filled.
Again, there are 3 sub outcomes for this. Both orders go to target. Stop gets stopped out, Limit does not and goes to target. Stop and Limit both get stopped out
I will address Outcomes A and B first. Clearly in A and B the stop order has the advantage, because it will enter every winner, and not enter every loser. The limit order must enter all of the stop orders losers and its own losers. So clearly it is advantageous to use a stop order for these types of uni directional situations.
Outcomes C and D are much more complex as they have sub outcomes. However they are basic mirrors of each other, one order gets filled first then the other because of choppy price movement. So both orders are engaged. At the moment that both orders are filled the Limit order has a 2 tick or 20% stop loss advantage.
For sub outcome 1: there is really no difference, both orders are not stopped out and go on to make a profit. The farther the profit target, the less %age the 2 tick difference makes. This is also the case with the stop but I don't want to make this any murkier than it already is.
For sub outcome 2: This is definitely to the Limit orders advantage. It is able to capitalize on these 2 ticks were the stop order can not due to a better entry price. So the stop order takes a full loser, while the limit is able to take a winner.
For sub outcome 3: Both get stopped out full stop, there is no difference here again. Everyone gets to feel the pain and the situation is equal for both order entry types.
My thought is that it actually makes a case for using stop orders. It clearly wins the A/B category, being able to get 100% fill on the winners and dodge losers is impressive. For outcomes C/D 2 of the 3 sub outcomes are equal, everyone makes or loses the trade. But for sub outcome 2, Limit goes to target but stop gets stopped out due to limit orders additional 2 ticks of stop advantage. So the question becomes, is winning the 2 sub outcomes of C/D worth taking eating all of the losers of Outcome B, and missing all the winners of outcome A?
Back of the envelope comparison the hypothetical value it may not be. For C and D I am only going to use sub outcome 2 because the other 2 are equal. I multiply them by 1/3 because it is 1 of the 3 possible sub outcomes.
Outcome A: Stop Order = 0 Limit Order = -10
Outcome B: Stop Order = +10 Limit Order = 0
Outcome C-2: (Stop order = -10 , Limit Order = +10)*.33
Outcome D-2: (Stop order = -10 , Limit Order = +10)*.33
Total Stop Order Value = +3.34
Total Limit Order Value = -3.34
Now, this is assuming everything has an equal probability. The true expected value difference depends on the probability of outcome C-2 and D-2 occurring more or less often then any of the other 2 sub outcomes. If they occur more than 33% of the time, then Limit orders are worth more, if they occur less than 33% of the time they are worth less. Sub Outcome CD -2 would have to occur 50% of the time for all sub outcomes for limits to be equal to stops.
I would love to hear opinions, comments, or if anyone wants to do a more in depth probability study that would be great.
Last edited by treydog999; January 2nd, 2015 at 07:47 AM.
Reason: Bad math
The following 2 users say Thank You to treydog999 for this post:
I was thinking about this some more. In sub outcome C/D - 2. All things being equal, if the limit order's advantage in stop distance is 20% then 20% of the time the limit order should stay alive while the stop order would be triggered. This would actually be less than the 33% probability i assigned it. It could be 30% as the stop can trigger if we trade at the stop order's trade stop order price level (i hope that makes sense) and it goes to market as it would add an additional tick that we may not have actually traded through. However that is still smaller than the required 50% for limit orders to be more profitable.
The following user says Thank You to treydog999 for this post:
this is exactly what i was thinking about aswell. which is superior- a limit or a stop order?
As i read Al Brooks books (which i can highly recommend) i tried to focus on stop orders to enter my trades. Al says that -at least for beginners- it is better to enter with a stop order, since the trade is already moving in your favor.
I had mixed results trading this live though. But it is my fault not Al's.
What i am doing since i couple of months is, that i am taking notes of all my live and sim trades, one of the notes beeing if it was a limit or a stop order.
My notes show that although Limit or Stop Order Entries have pretty much the same win % (-in my case-), certain entry criterias that i trade like a 50% retracement or a measured move i only enter with limit orders were as other entry criterias like a Low2/High2 or reversals i only enter with a Stop Order.
So maybe depending on a persons trading style or entry criteria Limit Orders or Stop orders is the way to go.
Maybe it is best to start taking notes of your trades to get some statistical data, because everything else is just hypothetical. You can take notes of a trade that you intend to take as both a Limit and a Stop Order Trade (taking into account a different Profit Target and Stop Loss for Limit Order and Stop Order trades ... if you want to adjust them). After a 100 trades you will get fairly relialbe results and know which one is better or if there is any difference at all.
Another thing is slippage. Although this is just my impression (since i just started to take notes of whether i had slippage or not) stop entries tend to more frequently generate slippage than limit orders.
I definitely believe that you need to collect stats on your particular system. But i think from a purely system neutral perspective that stops may still hold advantage over limits. Also limits can never incur slippage due to the fact that limit orders must be filled at that price or better ie price improvement. However stops are really just fancy market orders and thus you always pay the bid ask spread. Yes you can have stop limit with a threshold but still has the mitigation factor of a normal limit that you may not be filled. Also changing targets due to the difference is something thats system specific so i did not address it.