Following on from yesterday’s post, I have been thinking about when to close positions. The reason that this is pertinent is that a couple of trades are just starting to go my way, and one is doing well — everything has moved my way this morning (sugar is looking very sweet, oil is sliding, equities are set fair; natural gas is bottled up but under pressure; I find these so funny that I am sitting chuckling to myself). I could exit all trades now for a MTD return of 2.11% and a YTD return of 1.78%. This question is distracting and is in danger of becoming corrosive.

In my first three months of trading, I wished several times that I had just closed everything. In particular, I have twice seen my equity rise above its starting level, giving me a profit since inception. I keep having the looming idea that I could have become a good trader just by taking profits when they were on offer. But this brings out the point: it wouldn’t have made me a good trader. I was overtrading and only a combination of luck, leverage and the fact that I was sometimes doing the right thing gave me those brief moments of unrealised profit. Even if I had taken the profit, it is likely that my next few trades would have eaten it up. And because my entry method was not good, everything that I think I have learned about exits rests on a shaky foundation — if profits come and go at random, it makes sense to grab them when they are there; if they are better than random, it makes sense to hold out for a bigger move.

Constraints on what a Sell Discipline should Look Like
I still cannot be sure of the worth of my entry method (although my confidence both in the method and in actually using it is growing). From reading around on the internet, I have become convinced of the importance of improving my win rate — I suspect it is possible to have a win rate well over 50%, compared to the 30% that I had in my first three months. If I did achieve this, it would mean that most of my entry signals presaged decent moves. So my sell discipline should be consistent with a majority of trades being decent winners.

Selling Out when you have Made Enough
The idea of closing several trades at once makes sense if you have a lot of trades that start off in the right direction, and then might or might not continue to do well. Why? Because if you sell when you have made enough, you will probably sell when one or two trades have had a decent move, and a few others are just getting going to the extent that they add a meaningful bit extra to the profit.

This would make sense if, say, my entry method had a 60% probability of indicating a move of one risk unit in a given direction, but only a 50% probability of being followed by a further decent move after that. That might turn out to be true of the entry method that I finally come to; but it is not my objective at present.

Selling Out on a Trade-By-Trade Basis
The alternative is to sell out of each trade one-by-one as each has its day and grows old. This is consistent with an entry method with a good probability of heralding a strong move — the benefit of capturing the strong moves will more than outweigh the advantage of taking several small ones at random. I am not saying that I have achieved this, simply that if it is what I am trying to achieve, it makes sense to have an exit method that is consistent with it.

When has a Trade Had Its Day
So how to establish when a trade has had its day? There are three ways that it might be possible to know this: technicals, market story and fundamentals. Technicals and market story are the only ones that can tell you when a trade is in danger of reversing in the short term (and every exit is a short-term trade); fundamentals change slowly and are to some extent mediated by the story.

The following questions therefore seem to make sense for establishing when to exit a trade:


  • Is the trend broken?
  • Has a key level recently been broken?
  • Is a strong level impeding further gains?
  • Is the story getting overdone or old?
  • Has the story changed?
  • Is the market moving to another story?
These questions will not yield a single clear answer, and it will require time and experience to understand how to balance the answers in order to reach a hold/close-out decision.
None of this has allowed me to establish a good sell discipline — that will only come with experience. This piece is about working out what direction I should be heading in as I try to gain that experience. Having done the exercise I am comfortable with the direction, and when I have the capacity to back-test ideas I will be able to try out some of the technical exit indicators on historic data.

Update: Note that I tend to use “sell” and “sell discipline” to talk about closing out of trades. To a trader, to sell is to short, to buy is to go long, and to close out is to do the opposite of either of these in order to negate the position. For the sake of non-specialist readers of this blog, I tend to use “short”, “go long” and “buy” when talking about entering new positions, and “sell”, “exit” and “close out” when talking about closing them out (obviously I avoid saying I have sold my short positions, which would just cause confusion).

Update 2: Should I add to the list: “Economic: have you made enough on this one trade?” If so, how should I define “enough”? I have been working with a mental level of 7 risk units.