This post deals with some ‘philosophical’ aspect of trading, so for those who don’t have a habit of thinking deeply into why they do things so as to take the most optimal action (which is totally fine), you might want to stop reading.
The Market Does Not Care About Your Entry Price
There is this line of thinking, that the actions that you take for a trade at any point in time, should be the same regardless of your entry price, because the market doesn’t care about your entry price.
For example, this would mean that your stop placement would be placed at the same logical spot regardless of your entry price. Taking it further, it would also seem to imply that whether you should be in or out of a position can purely be determined based on the market action, again regardless of whether you had taken the trade earlier, later, or even if you were not in the trade in the first place.
Note that not having a position is practically the same as having an entry price that is equal to the then current market price, because with today’s low commissions, you can get into or out of a position at negligible cost. The state of “having no position” is economically equivalent to the state of “having a position at the current market price”, so the optimal action that should be taken would be the same regardless of whether you have a position.
To clarify this even further, imagine two scenarios.
- Scenario #1: You had taken a position when the entry price was 3 ticks from your stop. Price moves up another 5 ticks, so now price is 8 ticks away from the stop.
- Scenario #2: You missed the entry when price was 3 ticks from a logical stop. Price has now moved to 8 ticks above the logical stop.
In both scenarios, the market is the same market, the time is exactly the same. Another sort of hybrid hypothetical scenario is that say you suddenly found yourself with a position just like Scenario #1, but you couldn’t for the life of you recall what was your entry price. In this case you would have a position, yet you cannot take actions depending on your entry price because you don’t have it. How would you manage such a position moving forward? The decision faced here would be the same as the decisions in Scenarios #1 and #2.
Now let the potential reward in both scenarios to be xyz. In Scenario #1, your basic choices (let’s not complicate it with pyramiding and stuff) are to
- Exit: Trade expectancy = 0
- Hold on: Reward / Risk = xyz / 8
In Scenario #2, your basic choices are to
- Enter: Reward / Risk = xyz / 8
- Stay out: Trade expectancy = 0
Now with a commission of $1 per contract, you can switch between Scenarios #1 and #2. So effectively, you are not “constrained” by your current state. If you are in Scenario #2, you can get to Scenario #1 by paying $1. If you are in Scenario #1, you can get to Scenario #2 by paying $1. Since the cost of translating between states is negligible, the best course of action would be to choose the best choice among all the choices available to Scenarios #1 and #2.
As you can see above, there are essentially two choices: “Expectancy = 0” vs. “Reward / Risk = xyz / 8”. Now which is the optimal choice is a tough question, but the point here is that whatever the optimal choice, there is only a single optimal choice, and whether or not you had the position earlier does not matter. It will not make sense if you choose “Expectancy = 0” over “Reward / Risk = xyz / 8” in Scenario #2, but choose “Reward / Risk = xyz / 8” over “Expectancy = 0” in Scenario #1, when the comparison is the same.
But Should You Care?
While that line of thinking seems intuitively correct, one thing that always nagged at me was that intuitively, the entry price does seem to matter.
For example, if I entered at a turn when the price turns up and the stop is tight, I might still be holding my long position when price runs up far away from the turning point. However I would not enter at that price after it has run up because it is far away from a support and the risk is large. So in this example, I should be staying in a long position because I entered earlier, but if I didn’t enter earlier, I would have no position. Why is it that in this case, the same market situation, your entry price matters?
Let’s look at what’s going on in our heads in the two scenarios above.
Scenario #1: Entered at the turn, in a good profitable position with a tight trailing stop
= Original balance + unrealized profit + optionality ranging from defined small negative to unknown positive
= Original balance + locked in profit + optionality ranging from 0 to unknown positive
= Original balance + locked in profit + free punt with no downside
Scenario #2: No position
= Original balance + optionality ranging from defined big negative to unknown positive
= Original balance + bad risk/reward trade
So comparing the two scenarios, why is it that Scenario #1 has a free punt trade and Scenario #2 has a bad risk/reward trade? What’s the main difference?
One key difference is obviously the stop. Scenario #1 has a tighter stop in place while Scenario #2 has a far away stop. But wait! Shouldn’t both scenarios have the same stop since the market doesn’t care about your entry price? Does that mean that the stop placement for one of the scenarios is incorrect?
Is it wrong for Scenario #2 to place its stop at a far away but logical spot? No because that is the logical spot where you know that your trade has failed, that your reason for taking the trade is no longer valid. So does that mean that for Scenario #1, the stop should not have been trailed higher than the stop in Scenario #2?
Why did we trail the stop higher in Scenario #1? There are two possible reasons
- We entered the trade because we think a trend will develop. We had a good open profit and we do not want that open profit to turn into a loss. If we kept the initial stop, and price goes all the way back, a very profitable trade would have turned into a loser. In protecting our profits, we also avoid our psychological capital from being depleted because of the distance to the stop, making it easier for us to stay in the position. But in taking this action to protect our open profits, we have raised the stop to a level where even if price breaks below it, the reason for the trade has not been invalidated, so in a way, it is still good for re-entry at the price of the stop.
- We entered the trade because we want to just grab a momentum move. The trailing stop tells us when momentum has stalled and is time to exit. So when price hits the stop, our reason for the trade is over.
So say this is a trend trade, and we follow the trading rule (designed to ride a trend to the end) to only get out when the trend is invalidated, meaning that we don’t protect our open profits, and we keep both stops at the same level. Nonetheless, Scenario #1 would still be in the trade while Scenario #2 would not be in. Why?
These are the likely reasons
- House money effect
- In Scenario #1, we are treating the open profit as ‘house money’ that is a freebie that is ‘losable’, hence placing less value on the ‘house money’ compared to our original capital. This is a very human psychological pitfall.
- In Scenario #2, our loss would come from our original capital instead of ‘house money’, so that hurts more psychologically. Because of the potential psychological hurt that would come from losing our original capital, we would go for the “no entry” choice where we don’t lose anything.
- In Scenario #1, because we are already in a position, we need a reason to get out. If it market does not show any sign of turning down, there does not appear to be a reason to change the state, hence sticking with the original state means holding on to the position.
- In Scenario #2, again you are looking for a compelling reason to change your state of having no position. You see that the risk is large, you drop that option, and you are left with staying put.
- Confirmation bias
- When you are in a position as in Scenario #1, you think that the market is going up, especially when you have open profits. There is a confirmation bias that makes you stay in the position.
- Trading plan
- In Scenario #1, we need time for the trade to work, for the expected reward to be realized. When the reason for the trade is still valid, there is no reason to sell.
- In Scenario #2, when risk is minimal, we know that those are good bets. When risk is not minimal, those are uncertain bets, which we do not take for good reason.
Does Flipping the Scenario Rules Work?
It seems from the analysis above, the reason why there is disagreement is a matter of logic versus emotions / psychology.
Let’s explore a bit further. Will it work if we simply take the action that we would have taken if we were in the other scenario?
In Scenario #2, we would not enter a trade unless the risk is 4 ticks or less. If we apply that to Scenario #1, meaning that we would exit whenever the risk is more than 4 ticks, does that work? Obviously not, because we would be getting a 1 tick reward for 4 ticks of risk. The reason why we cannot simply flip the entry rule is because the entry rule in Scenario #2 is designed to take trades where the risk is minimal. It does not mean that a 5 tick risk has a bad risk/reward, it just means that 4 ticks and below is a good bargain. It’s like if you estimate the intrinsic value of a company to be $100, you might buy only if the market price $40 or below, but it does not mean that $60 is not worth being in the position.
Similarly price of $60 is worth holding, doesn’t mean that the price of $60 is worth entering. So if we would hold on to an existing position when the price is 8 ticks away from the stop, doesn’t mean that we should enter when price is 8 ticks away a stop.
The Reconciliation: Atomic Bets
What the thought experiment above means is that when we choose a course of action in each scenario, the decision does not mean that we have considered all facts and deemed one action better than the other. This point bears emphasizing:
- When we choose to hold on in Scenario #1, it does not mean that the trade expectancy of holding on is greater than 0 (exit). We are letting the bet that we took earlier work out, so that we know the result of our bet. In a sense, you can think of “finding out whether a bet worked” as an atomic operation (those with a computer science background would know what I mean). Trade management is included in that atomic operation. This does not mean that you cannot perform dynamic risk/reward assessments during the life of a trade. Those dynamic risk/reward assessments would be considered as trade management of a single bet. Hence the decision to hold on is a trade management decision, not because there is a calculation of the trade expectancy of holding on and a comparison with zero. On whether or not the trade management process falls into the ‘playing with house money’ pitfall, I would say that it does. The mitigating reasons would be (i) the difficulty of implementing a system that can dynamically calculate trade expectancy so that we don’t play with house money, and (ii) only adding closed profit to our capital, and progressing that capital, is a perfectly reasonable method of progressing our wealth over the long-run.
- When we choose not to enter in Scenario #2, it does not mean that the trade expectancy of entering is less than 0 (stay put). We do not calculate trade expectancy for a comparison (barring nearby levels) between the choices, instead we take trades based on minimal risk and an unknown positive reward. It’s more of a “it’s tough to get a deal better than this” logic, that’s why we enter. But doesn’t not entering implicitly mean that the trade expectancy of entering is less than 0? No it doesn’t. It just means the trade expectancy of entering is difficult to estimate so it was passed over. It’s just like if an interview candidate was passed over, it does not mean that the candidate is lousy, it can be a case where the hiring manager is just too uncertain.
As detailed above, the actions taken in both cases did not require the comparison of trade (or action) expectancy between the possible actions. We essentially bypassed the need to assess the trade expectancy of the “Reward / Risk = xyz / 8” action.
This means that there is no logical conflict, when we take different actions depending on whether we are in a position or not in position.
Is It a Matter of Trading Strategy Then?
The reconciliation above seems to be tied to a trading strategy. Is that the case? What if we have a trading strategy or method that is able to dynamically calculate trade expectancy at any point in time, compares the expectancy among the different choices, and takes the action with the highest expectancy. In that case, wouldn’t we expect both scenarios to have the same optimal action?
There is a subtle point when assessing trade expectancy and risk / reward. The risk is set by the stop. The reward is set by the trade management and exit strategy applied after you get into a trade. Now to estimate the reward, you have to make an assumption on how / when you would exit the position. In other words, the constantly running algorithm that decides what best to do at any point in time, needs to consider what decisions the algorithm itself would make in the future, so there is a recursive element to it. Conceivably, a straight-forward way might be to just build a tree of probabilities of price movement and calculate the expected P/L at each node, with information on stops, market structure etc changing depending on the path in the tree. With a trade expectancy algorithm in hand, you can then choose the action with the best expectancy, with potentially a buffer (e.g. one action’s expectancy has to be 30% better than another action’s expectancy) to give a margin of safety.
So yes, it seems that it can be a matter of trading strategy. To implement the kind of strategy to calculate trade expectancy on-the-fly though can be pretty complex, and is probably done only by pretty sophisticated groups.
For the large majority of traders, the market doesn’t care about your entry price or when you entered, but you should!
[This has been a pretty twisty thought process. If readers spot any logical errors or have different viewpoints, please feel free to post comments. Thank you.]