A note to readers: While Dr. Tharp’s content is timeless, this article is from our newsletter archive and may contain outdated information, missing links or images.
PART 1
People tend to think it’s hard to develop trading systems. I believe that having a process to follow makes developing trading systems a learnable skill. Let’s list the steps you might take to develop a low-risk idea and then look at each step in detail:
- Knowing the big picture and trading a strategy that will work given that big picture;
- Developing a fictional short-term model for what the market is doing;
- Finding a low-risk trade within the model where your risk is minimal;
- Taking the trade only if your potential profit is at least three times the size of the potential loss;
- Trading it at a risk level that allows you to continue to play the game;
- Continuing to trade the model if the trade works out or
- Adjusting the model if the trade proves that it is no longer valid. Let’s look at each of these seven steps.
1. Know the Big Picture and Develop an Appropriate Strategy
I believe we entered a long-term, or secular, bear market in 2000 and that it should last about another eight years. A secular bear market does not mean that stock prices are in a steady decline over the period; rather, it means that price-earnings ratios will be much lower at the end of the period—in the single-digit range—than they were at the start. A secular bear market may include strong economic cycles, and the market can actually seem to be bullish for long periods. In our current situation, however, the economic fundamentals for the U.S. are perhaps the worst ever on record.
My beliefs about the big picture mean that I will likely not be holding any equity positions for very long. The volatility associated with bear markets means great conditions for shorter-term traders. If inflation appears, commodities and real estate are likely to be very good areas for investing or trading. Because the U.S. dollar may be in for trouble ahead, gold and silver probably make some sense too.
If you understand your big picture and the elements involved in trading it, then you might come up with several types of strategies. You could use a model that predicts when stocks are rising or falling to buy stocks during up markets and short stocks when the market falls. You could develop an inflationary model and apply it to a long-term gold strategy because gold usually does very well in an inflationary market. Another strategy could be to wait for the market to become extremely oversold and then buy stocks that are deeply discounted.
My point is that you need to understand the big picture and what kinds of strategies will work given your view of the big picture. You have to understand the big picture and how it changes in order to make money in the long run.
2. Develop a “Fictional” Model for What the Market is Doing
Part of my big picture is the long-term decline in the value of the dollar. From 2002 until 2008, the U.S. dollar lost about one third of its value. Since then, dollar index has been moving sideways in a range, but for about the last year, it has been rallying.
Let’s say you decide you want to trade a foreign currency pair with a good trend, keeping the big picture in mind. Figure 1 shows a monthly bar chart of the U.S. dollar versus the Japanese Yen. You can easily see the decline in the value of the USD and the appreciation of the Yen since 2006. The trend is very clear. Generally, you would like your trades to be in the direction of the major trend in congruence with the big picture.
Your next step is to simply invent a model that will allow you to generate a low-risk idea within that big picture. Now take a look at Figure 2, which is a daily bar chart of USD/JPY since the beginning of 2012.
The daily prices seem to be moving in a down-trending channel since the highs in mid-March. Two trend lines have been drawn to reflect this, but are the prices really moving in that channel? Have prices really reacted to and stayed within those lines? Who knows? Whatever the case, we can still use this (invented) channel to find a low-risk trade. Currently, it looks like prices are hitting resistance against the upper trend line, and they seem to be rolling over.
So here’s an idea for a low-risk trade using a channel:
- Determine the direction of the longer-term trend (which in this case is down).
- We want to be ready to trade when prices get to the top of the channel. Ideally, we would like to see price go slightly above the top channel and then come back down through it.
- We don’t enter a trade until prices move in our direction. This is a very important principle for all traders: only take trades when prices are moving in your favor.
Notice that it doesn’t really matter whether the lines we drew reflect anything “real” or not. Do candlestick charts and price channels really reflect anything real? No, they’re simply tools or guidelines that help us set up a low-risk idea. In fact, the worst thing you could do is to start believing that your lines are real phenomena rather than an invented model. If you do that, you might be convinced that your model is right and fail to get out when price action proves it wrong. Many people who trade sophisticated models like Gann, Elliot Wave or others with similar complexity may run into trouble if they believe the model more than the market at some critical juncture.
3. Find a Low-Risk Trade That is Set Up By the Model
We’ve already started to do this with our lines on the chart. Our criteria for this low-risk trade include the following ideas:
First, we want the longer-term trend to be in our favor. Consequently, we really only want to take trades that are moving down off the top line. This is happening in the last few bars in Figure 2.
Second, we want to have some point of reference for our trades to set up the low-risk idea. This is the overall purpose of our model. The model suggests that the market is currently moving within the channel we’ve drawn with our imaginary lines. We don’t know that this is truly the case, but the model gives us a frame of reference. If prices leave the channel, the model is no longer useful.
Third, we could be rigid or fairly loose about our frame of reference. If we want to be rigid, we might demand that prices cross (or touch) our upper boundary line of the channel and then cross below it again before we could enter. The problem here is that we’d get very few trades. If we are willing to be more flexible, we might just require that prices approach the upper-line reverse and then we could enter a short position. It really doesn’t matter, because you only trade your beliefs anyway. And since you made up the model, you can make up the rules.
Fourth—and this is very important—we must set up a point at which we know that we are wrong about the trade. We might say that if the price reverses after our entry and moves above the upper trend line, we would cover the position. We might also say that if the price just goes sideways for more than a week, then something else is happening, the model is wrong about the trade and we would exit. You must have a clear point at which you declare yourself wrong about the trade so that you exit. The amount you are willing to risk (lose) is your 1R for the trade.
I chose a model based on channel lines because most people can see fairly easily that it’s just made up. However, your model could involve Elliot Wave, Fibonacci numbers, support and resistance, pivot points or anything else you care to invent. In every case, there’s only one purpose for the model: to set up low-risk trades.
PART 2
In Part 1, we discussed the first three steps of a seven-step approach to developing low-risk trading ideas:
- Knowing the big picture and trading a strategy that will work given that big picture
- Developing a fictional short-term model for what the market is doing
- Finding a low-risk trade within the model for which your risk is minimal
This week, we’ll talk about steps 4-7, which address the need to manage risk and adapt your system when necessary.
4. Make Sure Your Trade Has At Least a 3R Potential Gain.
Now that you know what a 1R loss will be, it’s important to determine your profit potential. If your model is correct, the prices should go to the other end of the channel. Ask yourself, “what kind of move is likely, and how big is that move with respect to my initial risk?” If it is at least three times your initial risk, it’s probably a good trade for you to take. If it’s less than three times your initial risk, skip it. You can find a better trade later.
Notice what this step means:
- The smaller the initial risk set up by your model, the more likely you are to get an acceptable trade.
- The model is useful as long as you’re not married to it. As soon as you decide that it might be real, you run the risk of developing a psychological need to be right about the model, and that will influence your decisions.
- It doesn’t matter whether or not the model is accurate or real—just that it gives you low-risk trades.
Consider this chart (Figure 1) from last week’s article, which shows daily bars of the USD/JPY prices for 2012 (through mid-June). If we entered a short trade at 79.30 and placed our initial stop at the upper trend line, 79.50, our risk per unit would be 0.2. Our first exit target might be the swing-low close at 78.00 on June 1, so we’d have the potential for an 8R profit—well above our minimum of 3R.
Figure 1: USD/JPY Daily Bars Jan. 1, 2012 through June 14, 2012
5. Take the Trade With a Position Size That Allows You To Continue to Play the Game.
Essentially, we’ve already talked about this step in our definition of a low-risk idea. You are basically trading beliefs that you made up, but as long as you remember that you made them up and keep from getting attached to them, it doesn’t matter.
Think about it. Let’s say you make money on 50% of your trades and that your average profit is about 2R, while your average loss is 1R. This gives you a pretty good system – even though you invented everything.
Keeping this illusory nature of your system in mind, you must trade it at a low-risk level that will allow you to survive any unexpected occurrence. Although I recommend that generally you risk no more than 1% on each trade, you are probably much safer risking ½% per trade when you start trading, especially if this is a new model with which you don’t have much experience.
By the way, let’s make sure everyone understands what risking 1% means. Suppose you found a stock and decided that your initial risk was $2.50 per share—the difference between your entry price and the initial stop. You would get out of the position if the price dropped by $2.50 after you entered. If you have a $50,000 portfolio, and you want to risk 1% of that, then you’d risk $500 on the trade. Now, if you divide that $500 by the $2.50 unit risk you get 200 shares. If the stock cost $40 per share then you purchase $8,000 worth of stock. You would get out of the trade if the price moved against you by $2.50, so your risk is only $500—not the entire $8,000 position. This is what we mean by risking 1% of your portfolio.
6. Continue to Trade the Model if the Trade Works Out
If the trade works out, you might find that the model you made up will work for four or five more trades. If so, that’s great; continue trading the model.
Even if a trade doesn’t work out, it might not destroy the model. For example, the trade might retrace and go back above the line you drew, stopping you out. However, it if just stays there or then reverses again, it doesn’t invalidate the model. You might still get some more trades out of it. Continue to do that. You might have made up the model, but as long as it works, just continue to use it.
7. Adjust the Model if the Trade Doesn’t Work
Let’s say that the position does reverse and goes way back above the line. In fact, it goes so far above the line that your model no longer makes sense. At this point, you must realize that your model is just made up. Abandon the old one and find a new one.
From our previous example, you may find that you simply need to redraw your channel lines. In Figure 1, the USD/JPY bars were rolling over at the upper trend channel line. They did go lower but then later reversed to go through the upper trend line (see Figure 2 below). You might make up that there’s a new upward channel in which prices are now moving—or at least until they don’t anymore.
Figure 2: USD/JPY Daily Bars Jan. 1, 2012 through June 27, 2012
You might find that nothing you invent makes sense. However, later, after studying the charts some more, you find that another pattern seems to jump out at you. Okay, that seems to be what now fits. Go ahead and use the new model to help you generate new low-risk trades using the same principles we’ve just described.
If you trade complex models (made-up models in my opinion) such as Gann or Elliot Wave, always remember that those are just invented models and you’ll be fine. You are simply following the steps I’ve just outlined. Once traders start believing the model, however, they have become convinced that their model is real and they stop seeing what the market is truly doing. Problems always occur at this point and the net result is usually a disaster.
Just remember that you are simply inventing models that give you the basis for low-risk trades. If those models are based upon commonly accepted trading practices (e.g., trend lines, support and resistance, Fibonacci numbers, etc) then enough traders may believe them to give them some validity. If not, then you still have a low-risk idea and you must act on it and get out. The net result, even though your idea is based on a fictitious model, is still a good trade.
Remember, the key is that you cannot trade the market; you can only trade your beliefs about the market. So why not make up some useful beliefs?