Exits are, in my opinion, both among the most important and least important things to discuss in a trade. Most important because they are often caused the most angst and emotional reaction. Least important because if your other analysis has been done well then exiting a trade just affects how much money you make and not whether you are profitable or not.

In other words you should have already found situations where you think you can make money and then it is up to you to get into those situations in a way that does not mean your positions gets stopped out so often that you’re unprofitable.

While this seems blindingly obvious many Traders often make a concession to themselves in private that they couldn’t have known x would have happened or that y was unusual or Z had a reason which changed the situation. They are all reasons but none of them matter because they all just add up to stats which show you are either profitable or unprofitable.

You can never know whether x will happen or not but you need to know how many times it has happened historically in this setup in order to make a judgement about whether you can use this entry technique to make money.

In order to be able to say why was unusual it’s only meaningful when you can put it into numbers by saying it only happens 5% of the time or one in 6 times after this setup the price makes a new high and then retreats 10%. The remaining five times it only retreats a maximum of 4%, as an example of the specificity you need to have about your trades.

What exactly are the specifics of this particular set up that prompted you to enter the trade? Only once you write down exactly what your criteria were for this particular trade can you go back and check historically on this product or market what your “chances are”. For example if you knew that in soybean futures when the 200 day moving average was sloping strongly down (slope > 0.001%) that a false break occurs 12 out of 15 times, then you either need to add a second filter which can help identify those three times or cut down the 15 to 6 or something like that.

Alternatively if the odds are really terrible for one thing working then it should mean the odds for the opposite working are very good so then that becomes your new trade idea. Then you need to do some trade setup research to see how many times the false break results in a particular outcome such as a 5% move down, which you can then trade.

now you have the idea of getting statistics on how many times a particular outcome occurs as a result of a trade setup. This means you know how many times you might be right and how many you will probably lose money. This becomes very powerful because in effect it tells you if using the previous example 12 trades are losers and 3 are winners then you will need to make 4 times your stop loss on the 12 trades in order to break even .

So the first thing to investigate is what do those 3 winners look like and what is their average return from the period after* the trade set up occurs?

If they are fairly evenly distributed, meaning the returns are all roughly the same, that’s more reliable than two small Winners and one huge winner.

Aside from the fact it doesn’t seem smart to bet everything on 1 huge winner it is also very emotionally draining to have all the losses that you were expecting and then also have your when is be small waiting for a huge winner that may never come.

This may lead you to take a detour from your plan and again make all your historical statistics irrelevant once you start deviating from the plan. Just remember the hardest thing about trading is waiting. Take that into account when you are telling yourself that you are going to take all these losses and eventually smash that home run because chances are, if your method has a high percentage of losses and it’s relying on you to pull the trigger for all of those losses then the process might be so demoralizing you just never wait long enough for that massive winner. In the meantime you’ve burned a good chunk of your capital.

* I say the period after because you won’t know the trade setup has occurred until after that data point has finished so your entry point will be at the start of the next session (at the earliest). If you are thinking that you can just take the closing price on the day the set up occurs then you are leaving yourself open to a mis-calculation, because the set up at the close might be different to what it was 5 or 10 minutes before and therefore your historical statistics are potentially irrelevant. Secondly bear in mind that historical statistics are just that , a record keeping of what happened in the past and no guarantee of what will happen.

So once you have an average profile of what your winners look like then you can start planning what sort of stocks are required to ensure you stay in those winners to the expected result, EG a move of 5 or 10 or 30% in your favour.

If the tradesetup occurs and on one of your trades the lowest price between your date of entry and the minimum profitable result you need is 5% below your entry you can’t say well I would just get stopped out on that one because it changes the odds from 12 to 3 previously to 13 to 2 and means those two winners are shouldering even more of the load dramatically increasing how much they need to return in order for the whole system to be profitable. The other two trades might only go 1% below the entry price before kicking on to Glory and a fantastic 35% return but if you are counting the results of all three winners then you need to make sure your stop loss for every winner is the worst of the 3 and additionally that same 5% stop needs to be applied to all of your losses as well because at the time you will only know the trade setup has occurred and not whether it will be one of your winners or losses and of the winners whether it will be the one that has only a small draw down or one that has a large draw down.

For me personally this has meant I’ve had to say goodbye to many great looking trade setups because the drawdown required to achieve one or a few of the big winners was so big that the potential losses on the large number of losses would make the system barely profitable or operating at a loss.