Run Profits and Cut Losses

Well, how many times have we heard that one? This must be the number one piece of useless advice trotted out repeatedly on trading websites, newsletters and courses. What does it really mean? The trading world is so full of quantitative data - stock prices, indices, moving averages, indicators - that the above advice seems not only such a platitude, but is also rarely backed up with any hard and fast rules.

The opposite of this rule, "Cut Profits and Run Losses" is obviously insane (and yet how many of us have been guilty of doing just that?) The delusional hope that the market will turn in your favour. The look of horror as your profit dwindles to nothing. Praying to your indicators will not appease the market demons. To our knowledge, Dante was no day trader, but his admonishment to "abandon all hope" as he stepped into the darkness should be plastered above the entrance of every stock market. Not much of a unique selling proposition, but nevertheless true. Dante was lucky in having a guided tour, but you're on your own.

As I write this I am increasingly aware that this could get very long, so will break it down into digestible pieces. I first want to look at the idea of running your profits. As I said, all rather obvious, but what does it mean? You sit there, your trade is, say, 50 pts in profit. Your exit indicators (whatever they might be) are nowhere in sight, but the market starts to move against you. You see your profit shrink to 40... to 35... to 30 pts. Your exit indicator tells you to stay calm; the trend is not broken. Your profit keeps dwindling, now standing at 20 pts. Is this noise or a signal? Take your diminished profit or wait for the trend to return in your favour? After all, you were right, weren't you? But were you 50 pts correct, or just 20 pts? We've all been here, and this is exactly where many will "cut their profit". You really don't want to see a winning trade go into the red so take a deep breath and take the 20 pts on the table. Damn! It's just 10 pts now! Take it! Now!

So what does the wise trader do in such a situation? Some courses suggest that before opening a trade you write down not only your stop loss but your exit point. But doesn't that go against the advice of running your profits? Well, yes, however, I'd suggest writing down the exit point just to remind yourself of your original expectation. You may have an overly optimistic expectation but trading should have beaten that out of you by now. The crucial point here is, I think, what instrument you used to initiate the trade. What was the timeframe of the chart you used? This is what I tend to do. Let's look at some numbers.

A 3-minute FTSE chart (using my combined SAR and MACD indicators) will give about 20 signals during a day. Some of these will be mixed signals and I personally would not trade, but that's another article. The average trend on a 3-min chart is about 20 pts. A 30-min chart will give 2 or 3 signals during a day with an average trend of some 50 pts. Already you can see that your expectation - your tentative exit point - should be (indeed must be) different depending on which signal on which chart you were using to trade. To wait for a 50 pt profit on a 3-min signal will doom you to lose most trades. So what should we have done in the above fictional trade?

Firstly, at +50 pts, we would have already moved our stop. Using a 3-min chart, remember that the average trend, and hence average profit is just 20 pts. I would have moved the stop down to +30 pts by this stage, already locking in a more than healthy profit. If the 50 pt trend was very quick the exit indicators on the 3-min chart will lag a bit but will probably be somewhere about the +10 to +20 mark. If the trend was slow those same indicators should be fairly close to +40. On a 3-min chart the strategy seems crystal clear. You have already run your profits as they are well above the average. You have already followed the golden rule. The chances of there being more profit using the same 3-min chart is very very small. Take the +40 pts and pat yourself on the back - save the sweat for another time.

But what if we had used the 30-min chart for the same trade. As I said, depending on how the day goes there may be just one signal, or possibly 2 or 3 - very rare that there are more than 5 signals. Your average expectation is about 50 pts. You've hit that but have in the back of your mind this running your profits rule. You're sure that if you take it now it will go another 50 pts in your favour and you'll regret it. At this point I would have moved my stop to either +10 or even +20. You won't lose anything. So now we watch the market reverse and see our profits dwindle. I think this situation is less clear. I personally like to switch between timeframes as they give different views as to what's going on. Indeed in this 30-min example I will usually flip to the 3-min chart just to micro-manage the trade. The 30-min chart is good to show the trend for a session (or half a session) but is too coarse to find a really good entry (or exit) point. In our example, we will reach the point where the 30-min chart is still showing an unbroken trend, but the 3-min chart has already signalled an exit. What to do?

There is, in my mind, one crucial question behind these judgments that we have not as yet explicitly stated. How can we tell a valid signal from background noise? We use all kinds of analytical tools to somehow smooth out the noise but only with hindsight can we tell which reversals were genuine signals and which were the noise of a complex system. However, we can use statistical methods to give us an edge. In the above examples, such statistical data will save us a great deal of sweat. Although, like anything, it will not be 100% right, it will provide a benchmark for us to distinguish a signal from noise.

So let me introduce two quantites that we have actually already been discussing but haven't as yet defined: our Expectation; and our Signal Strength. We have already met our Expectation values - for the 3-min chart it is 20 points and for the 30-min chart it is 50 points. These are our average maximum profits using the open signals from the given charts but closing at hghest possible profit before the close signal. These values will be different for different chart timescales and for different indicators. You will have to calculate your own. The Signal Strength I calculate as the difference between your average profits (I hope they are profits, otherwise change indicators) and your Expectation value. This gives us a statistical number that defines how much a market pulls back from potentially maximum profits before it hits your close indicator. I find this helpful in precisely the above scenario, where the close indicator is still a long way from being triggered but the market moves quickly. For the 3-min chart the SS=10 points and for the 30-min chart SS=25 points. I am not sure if it is significant that these are half the expectation, but perhaps you can save yourself a lot of calculations and take half the expectation as the signal strength.

So what use are these new numbers? To me, they give me an indication as to when any market reversal is a signal rather than noise. It also gives me a new exit point if my usual indicators have not yet been triggered. There is a case to possibly change exit indicators but that's for another time. The other option is to adjust the parameters of the indicators to be more effective as market volatility changes, and yet again this is another topic for discussion. What I have set myself in this article is what to do in the scenario where the market starts going against you, your profit is dwindling, and there is nothing in your trading armoury to tell you to bail out. Remember, we were trying to define what "run your profits" actually means, apart from being a platitude.

So what would we have done, using our new tools? In the 3-min example, we had reached a profit of 50 points. With a Signal Strength of 10 points, I would have closed at +40, which is exactly what we did. Using the 30-min chart the situation was less clear; we were sitting at our Expectation value so would have been a good trade but it was possible for the market to move in either direction. Even if the market retraced 20 points, on a 30-min chart this could still just be noise. With our new parameter of SS=25 we would set our exit point at 50-25=25 points. We would thus close the trade at +25 points profit. It is unusual for a 3-min trade to yield a higher profit than a 30-min trade, but the latter signal is much slower to respond to rapid changes, hence the need for a profit-taking signal. The SS value can also be used to set your initial stop loss, indeed you can remove yourself from the close altogether by setting a trailing stop at these values. Sometimes the market will then resume in your favour and you will have stopped-out with only a modest profit. As always, hindsight is really of no use whatsoever in deciding what to do in the here and now. With just fairly modest statistics we now have a handle of how large a movement can be counted as a real signal rather than random noise. If your closing indicator is already ahead of your trailing stop then use your indicator - you have already run your profits. If your new trailing stop gets hit then it is done with the knowledge that the retracement was large enough to be considered a signal - yet again, I think you have done your best to run your profits - nobody is going to tell you the precise turning points.

Just some closing comments. I have developed these numbers myself and have no idea if they already have a name or if they are quoted elsewhere. I really can't be bothered to search this out as I find them useful and that's all that matters. If anybody cares to enlighten me on such links then feel free to leave a comment. I may refine these in the light of further data. For example, when I started analysing the data I thought the standard deviation of profits would prove useful, but then found that it was too large a number; mainly because there are a small number of very profitable trades that skew the distribution. Those of you who wish to do your own analysis, it is very simple. You just need to input three numbers for each trade: the opening value, the closing value and the best possible closing value (the mythical turning point that we can only see in the rear-view mirror). Then calculate your average profit/loss using just your signals and your maximum average profit/loss using your entry point and the best possible close. The more data you have, the more realistic your averages will be. Even if you don't use my ideas above, these really are the minimum requirements to test whether your indicators are useful or not. Once you have back-tested everything and you are happy with your profit expectations it is then time to forward-test your system in real time doing paper trades. If you're still feeling happy, then it's time to open your wallet.