TF focuses on closed-trade equity. I don't think TF sees a beginning cocoa trend as somehow inferior to a winning, older trend in coffee. There's really no way for a pure trend follower to sit there and say, "I'm going to take some of this winning position off today." As a result, closed-trade equity ends up being the closest thing to order you can think about on a daily basis as a TF. If the system rings the bell and says place the trade, then, at that point, all trades are equal. Take every trade, or something like that.
The market's been swinging back and forth since April highs, so I now trade swings rather than "sit on" positions. Some examples are I bought IRR and NLY when they were sold off after the May crash, collected the dividends, then sold them when the market was extremely overbought not long ago. I will look to buy these again when the market establishes signs of beginning a new move up. I bought PWRD when it sold off to major lows in late June and sold it several days later when price moved up to an overbought condition. Also bought CIM at 3.58 when it sold off to major lows in early July, and will likely hold this one longer term, as it pays a very high dividend. This is all IRA stuff. My trading account is used strictly for day trading and an occasional overnight hold. I would think it's difficult to hold longer term positions now because of the large market range and no specific news to spark either a major selloff or major bull run.
There could be some mixing of timeframes (the trade timeframe, and smaller timeframes - for example daily chart vs hourly chart, weekly chart vs daily chart etc) if you want to optimize this a bit more. In other words, have longer term indicators of when the trade is over - trendlines, hitting or violating major support/resistance, crossing major moving averages, etc. If any of those are violated, you are done with the trade. However, the same indicators on the shorter timeframe are cues of when to lighten up the position, add to it, or reinstate the position. Say for example with a long position we buy 5 units. Put the original position on and ride the trend as planned on trade timeframe. Sell stop is placed where the trend will be considered over. Meanwhile, monitor the faster timeframe. If signs of exhaustion, rolling over, loss of momentum, breaking short-term trend occur, then sell off 1 or 2 units of the original position. You can even sell off all of them if desired, but have a definite way to reinstate the position on further trending behavior (such as breaking out over consolidation to new highs, etc). Only re-add the units once the consolidation has shown signs of ending and trending behavior is evident again. The advantages to "trading around the position" is that if there is a quick reversal to the downside and the trend dies, you will get out faster if you have front-run the exit on the shorter timeframe. If a trend dies and the market goes listless and trendless for a long time, your money isn't tied up riding back and forth without generating returns. Also, if a normal looking consolidation suddenly accelerates to the downside and turns into a trend reversal (ie flash crash), you don't ride a fully loaded long position into a crash.
How you enter is key with holding for bigger moves. Need to get in near the start of the move so that the swings don't take you out. I hold TF , ES, and NQ overnight. I Mostly use 2B and 1-2-3 reversal patterns. Your mind set has to be focused on weekly and monthly goals. My mind set is that I will be probably be wrong 7 out 10 times but the size of the 1 or 2 swings I catch each week more than pay for the losses. Some people focus on win rate, I focus on high R multiple trades where I don't have to try to be right a lot. Most the entry techniques win rates are less than 60% which is random, hence is why I focus on exits and catching some moves to pay for all the small losses. You can control the emotions of holding also with your position sizing. Size for your worst case drawdown which for me is 7 losses in a row. I keep the risk size to less than .5% of total equity which would be 3.5% drawdown after 7 losses in a row. Once you can get good trade location, then it's just a new mindset you have to program yourself with. Hence we all trade our beliefs about the market and not the market. To change any trading behaviors you have to get to the root of them which are your beliefs. That's why it's not what good traders do that makes them the money, it's how they think when they do what they do.
Forgot to add that you have to really understand market context and the phases of the market to hold the trades. Need to be able to spot accumulation and distributions phases. Also need to spot the swings from the different time frames so you can understand what's a pullback and what's a reversal.
If it's any consolation. It was either Morgan or Lehman (can't remember which book) described the same situation. The position was short and went up, every day they had a meeting, the analysis was correct they all agreed. This went on for 3 weeks, no choice they sold at a loss. Next order of business was when to re enter the short. When they did again, same results. Sell at a loss weeks later. Fundamental were stronger at this point to confirm their position. The trade eventually worked out in their favor but not without much anguish, and gruff from those who worked there but disagreed but not there dept or money. I think the lesson was: it was prudent to sell. ------------------------ On the flip side, if your long and it is going down, and you need to hold, you'd like to think it would go down.. and down ... and down far enough to make it worthwhile to add. Pffttt.. how often does that happen?
I think the key is identifying major long term S/R levels and getting short or long at those points, based on your longer term view of the market. For example 1000 on the SPY is probably the most important S/R level at the moment. If you're longer term bearish, and the market breaks below 1000, you'd be looking to sell the 3 to 4 day highs below this figure. The Dax has just broken 6000 which is an important support. I'm looking for a rally back to nearly 6000 and some obvious loss of momentum so I can start putting on shorts for a retest of yesterdays lows. Also you need to be aware of the average swing length in the market you're trading. At present, the down moves on the Dax are roughly 5-7 days so once the move begins, you need to psychologically commit to sitting in the trade for at least five days before looking to take profits.
I came to the conclusion that even if I foresee a move correctly it is impossible for me to foresee how exactly the move will unfold. This was a most crucial understanding for me. It pointed me towards my present solution and simplified things for me very much. It is always a work in progress with room for improvement. I use different systems to trade the move. Some may get stopped out along the way with re-entries, others stay in. In genuine trending periods I will have my maximum position on from all my systems with rare stop outs. My systems have different sensitivities or speeds and entry/exit rules, from limousine to supertanker, so to say. I risk little on each system. When they all have kicked in I have a substantial position on, sometimes uncomfortably large, but I also look at a trend that ticks all boxes. When studying past trends one can see that no single system handles all of them correctly. One can also see that most TF methods work reasonably well during such moves. As has been already mentioned, when a market goes vertical and the volatility increases dramatically exiting before the systems do is often wise. An observation that may be helpful: My best periods are not special situation markets that make some nice move in isolation but themes that play out over many markets. This can accumulate large open profits but there is then no diversification to speak of. After I had my systems worked out I began to concentrate on recognizing such themes but this is harder than the systems part!
The following pretty much explains (and with great detail) what I had in mind: The idea is you don't wait till retracement actually happens in force. Instead you wait for signal of a short-term top (such as congestion area or sharp move in your favour) and take profit on a part of the position. If it indded happens to be the short-term top then whatever level you re-enter at will likely be better. If the price keeps moving in your favour... Well, you have locked some profit. You don't necessarily have to scale out continuously till the remaining position is negligibly small. Think of it as holding half of your maximum position and actively trading the other half in the direction of the trend. I know it can be quite frustrating if you unwound part of a position at the beginning of a large move but think about it in a different way. Suppose you enter a position with $5 stop. If you unwind half of the position when the price has moved in your favour by 5%, then the original stop on the other half is a break-even point for the trade as a whole. Now you don't need to be concerned regarding whether your conviction is correct. You have a "free" trade. Regarding the trend in 1995-1997, I know people hwo held a dotcom stock through this period. They waited for the stock price to increse in value 10-fold... and ultimately closed the position at a loss to the original entry... because their conviction that such a strong market can't go down forever was too strong.