“Intermediate-term” market cycles – which I define as 6 to 14 months in length in most cases, with an average of 10 months – are the “sweet spot” of trading. I make the most money trading these cycles, and you can too.
Shorter-term trends and cycles are more prone to unpredictable random noise. Longer-term ones can be accurately forecasted, but because long-term cycles average around 8 years in length, it takes a long time to extract profits from them. Good for a long-term investor, but not enough if you are serious about trading and you want to make a living from it or generate enough trading profits to meaningfully supplement your annual income. Active traders must be aware of the long-term trends but will place trades according to the intermediate-term trends instead, even when those intermediate-term trends oppose a slower-moving long-term trend, because active traders have a shorter time horizon for holding positions.
Don’t Trade Without an Intermediate-Term Forecast
The five components are the blue boxes in this flowchart, taken from that post (with the blue circle added to mark today’s focus):
In the earlier post, I briefly touched on how I make my intermediate-term forecasts. Again, when I refer to the “intermediate-term”, I mean cycles that generally last between 6 and 14 months, with an average of 10 months. In another post, entitled “Want to Conquer the Investment Universe? First, Make a Map!”, I listed the asset classes and subclasses I forecast. They include: the US dollar, major commodities, bonds, and each of the ten stock sectors. Now, I’ll explain my forecasting methodology in more detail.
Start with Your Long-Term Forecast
Always consult a long-term chart (30+ years of data, if possible) before placing an intermediate-term trade. In my post titled “Fearless Forecasting for Long-Term Investors”, I introduced my methodology for making long-term forecasts. Here’s a high-level summary:
- Begin with a baseline trend, which you expect to observe over several long-term cycles.
- Establish a range around the baseline trend, based on the depth and length of prior long-term cycles.
- Determine the current long-term trend direction by looking for a series of higher lows (an uptrend) or a series of lower highs (a downtrend). Also, determine whether this trend is strong or weak. If there isn’t any clearly discernable trend, look at the prior trend direction. Most likely, what is occurring is a flat correction and we would expect the prior trend to resume once the correction is over.
- Forecast the long-term trend direction and strength based on: The current long-term trend direction and strength, momentum, intermarket analysis, and fundamental factors. Look for more details on this in future posts, podcasts, and videos.
- Take the actual and forecasted long-term trends from steps 3 and 4 and translate them into a single percentage between -100% and +100%. This percentage determines where your long-term forecast will fall within the range of possible forecasts we defined in step 2.
Let’s use the technology sector as our example for this post. My long-term forecast for tech stocks is +1.4%. Here’s how I got there:
The range of possible long-term forecasts is -15.3% to +18.0% with a baseline of +7.0%. Technical note: The positive skew exists so we have symmetry across up and down cycles. (1+.180) * (1-.153) = 1.000. The percentage I apply to this range is -25%, which reflects that the current long-term trend is “Up” and the forecasted downtrend is “Down (?)”. In other words, there is a strong uptrend now, but I expect a weak downtrend to emerge so I select a long-term forecast that is below the midpoint of my forecast range by 25%. Calculation: 1.4% = 7.0% + (-25%) * (7.0% – (-15.3%)).
Study Prior Intermediate-Term Cycles
Using a daily or weekly chart with at least 10 years of history, identify past uptrends and downtrends. Let the price be your guide – don’t assume that every single cycle will follow the same length of time. It rarely works out that way. But, although the duration of any two cycles may vary, their average duration is usually stable over time. The half-cycles will average approximately 10.5 months from low-to-high or high-to-low. Thus, the full cycles will average approximately 21 months (10.5 * 2) from low-to-low or high-t0-high.
Here’s how it shakes out for the tech sector (as tracked by the XLK exchange-traded fund):
Lots of variation in cycle length, indeed – but averaging 23 months apiece when examining the last 10 years of history. Pretty close to our 21-month expectation. It’s very helpful to know that the cycle lengths tend to revert to a mean over the long-term. This knowledge helped us anticipate a couple quick turns in late 2015 and early 2016 following the extremely long uptrend that ran from August 2011 to July 2015.
The fourth column in the chart above shows the lag between the tech sector and the S&P 500. This is part of the intermarket analysis framework I’ll mention later in the post.
Identifying the past trends is pretty simple and is done in just the same way as the long-term trends. Look for a series of higher lows (an uptrend) or a series of lower highs (a downtrend), in alternating fashion. If the price action looks choppy or flat, look at the prior trend direction. Most likely, what is occurring is a flat correction and we would expect the prior trend to resume once the correction is over.
There are other nuances that make this both an art and a science. Sometimes I select a cycle low or high that is not the precise low or high that was reached. For example, the XLK hit 11.44 on 11/20/08, which is below the low of 11.53 I marked on 3/9/09. But, I chose 3/9/09 instead because it coincided with the overall low in the S&P 500 and because the points immediately before and after the 3/9/09 extreme were lower, indicating the 11/20/08 low may have been a fluke.
A note on trading being both an art and a science: I’m always skeptical of trading schemes that claim to be purely quantitative, or algorithm-based. There is simply no way to get around the fact that anytime you develop a trading system, whether it explicitly allows for the use of educated judgment or not, you ARE using judgment. All mathematical models rely upon some sort of assumptions, such as: market returns following a given distribution, a random process being stationary, or the model having an error structure that follows a normal probability distribution with a mean of zero. By using a mathematical model to make trading decisions, you’re making the judgment that those assumptions are valid.
Why do we record and study the prior cycle highs and lows? Because, we need the following info for our forecast:
- The depth of a typical low-to-high or high-to-low cycle: the percentage gain or loss that usually occurs. This is one piece of evidence we use to determine whether the current trend is nearly exhausted or has more room to run. First, select a depth of a low-to-high cycle, expressed as an annual return. For the tech sector, I chose +40.0% as the annual return during an upward cycle after examining the data. This seems like a very high return, but remember that it’s the return from the exact day of a cycle low to the exact day of a cycle high, with hindsight. It’s not meant to be a real-life return. For a downward cycle, I solved for -26.5% because this is the annual return that produces an annual return of +1.4% across an upward cycle and a downward cycle. Recall that +1.4% is our long-term forecast for tech stocks from the section above.
- How long the typical cycles are: do they average close to 21 months between highs and between lows? Have recent cycles slowed, suggesting the current and next cycles will accelerate? Or vice versa?
- Is there another sector, market, or asset class that typically lags or leads this one? In our example, the tech sector usually turns within a month or less of the S&P 500, so we’d look at any divergence between the two with great skepticism. This is intermarket analysis – drawing connections between different sectors, markets, or asset classes. The S&P 500 highs and lows should themselves be determined by considering highs and lows in bonds, for example. Most traders fail to consider this big-picture view, focusing on narrow indicators or chart patterns instead. Many of the best opportunities to rake in serious profits are found in scenarios where the asset you’re trading is out-of-line with similar stocks, sectors, or asset classes. If you learn to spot these situations, you’ll start to capture those profits and avoid traps.
We will choose an actual trend (direction and strength) and a forecasted trend (direction and strength) for the intermediate-term, as we did for the long-term. We’ll choose from one of the five options below:
- “Up” – representing a clear upward trend
- “Up (?)” – representing a weak upward trend
- “Down” – representing a clear downward trend
- “Down (?)” – representing a weak downward trend
- “??” – representing a situation where the trend cannot be determined
Now that we’ve recorded and studied past cycles, the actual trend for tech falls right out of that analysis. The last low was 2/11/16, so an uptrend is in place and it’s a clear one, as the below chart shows. So we pick “Up” as the actual trend.
As far as the forecasted trend goes, we’ve already gathered some info that will help us make a prediction. Yet there’s more we can and must consider.
Forecast the Intermediate-Term Trend
So far we’ve looked at the depth and the duration of the current cycle and we’ve seen how it compares with past cycles. We’ve also brought in intermarket analysis as part of this process.
Next, let’s join this insight with some more technical analysis to select a trend forecast and to pinpoint entry and exit points for a potential short-term trading position.
I’ve organized this into categories of: Pattern, Price, Momentum, and Time in homage to Robert Miner’s outstanding book High Probability Trading Strategies: Entry to Exit Tactics for the Forex, Futures, and Stock Markets, which I selected as one of my ten most essential books for traders and investors. Each of these categories deserves its own follow-up posts, so stay tuned for those. I’ll simply introduce each for now.
In general, a “trend” usually consists of five distinct sections or “waves”. During a trend, as “trend” is defined here, the market is reaching successively higher or lower extremes. A “correction”, during which the price moves in the opposite direction from the trend, usually consists of three distinct sections or “waves”. The corrective waves will also usually be shorter in time than the trend waves. These principles come from Elliott Wave theory. Many well-known traders have integrated these principles into successful trading strategies.
How does this fit in with the framework we’ve defined? I’ve called every cycle a “trend”, and haven’t used the term “correction.” It’s simple! Since we’re talking about the intermediate-term timeframe here, the trend (5-wave) direction is the direction of the long-term trend. The corrective (3-wave) direction is the opposite of the long-term trend.
For the tech sector, the actual trend we’ve chosen is “Up” in both timeframes. Therefore, we expect upward intermediate-term cycles like the one we’re currently in to have 5 waves. We expect downward intermediate-term cycles to have only 3 waves, and to be shorter in time than the upward cycles. This will hold until the long-term trend changes direction.
In the tech uptrend that began 2/11/16, there is a 5-wave pattern:
Consider also checking for other, more complex patterns. I recently began incorporating harmonic patterns, as found in Scott Carney’s book Harmonic Trading, Volume One: Profiting from the Natural Order of the Financial Markets, into my technical analysis framework. Simpler patterns like a head-and-shoulders top, double bottom, or wedge/triangle formations are also useful.
The introduction of a wave structure also allows for the projection of price target points using Fibonacci principles. This is a deep subject that I’ll tackle in a later series of posts. Here, the most likely endpoints of Wave 5 are between 51.72 and 52.51, so we should expect a little further rally before this upward cycle ends.
Many traders use momentum oscillators like the MACD, Stochastics, or RSI to measure how rapidly price is changing. This is what momentum oscillators do well, but most new traders go too far with them. They trade using very simple rules like MACD crossovers, and they aren’t successful. Never, ever, trade based on one indicator or one simple rule in isolation! A crossover, in which a more responsive momentum indicator crosses over a smoothed momentum indicator, may indeed signal an important reversal in momentum and may be a sign of an important trend change. But, you need to look at all the other market conditions first before executing a trade based on this single factor.
I recommend looking at momentum across multiple time frames, as Miner teaches in his book. As a general rule, you want to trade in the direction of the longer-timeframe momentum and execute your trades when the shorter-timeframe momentum reverses in the direction of the longer-timeframe momentum. For instance, when momentum on a weekly chart is rising but not yet in the overbought area, you would want to look for an entry point to buy when the momentum on a daily chart makes a bullish crossover. Again, don’t execute the trade unless you’ve first studied the overall market conditions.
In our tech example, the momentum on a monthly chart is positive but very overbought and starting to decline. This reinforces my view that the high between 51.72 and 52.51 will represent a long-term high in tech stocks. The momentum on a daily chart, shown in the screen capture above below the price chart, has risen into the overbought zone but has not yet reversed. This reinforces the view that there is just a little more room to run in this intermediate-term cycle.
Volume (the number of shares traded) can also be a useful indicator of momentum, but has become far less reliable than it was in the past because many trades take place off the major market exchanges so they are not part of volume statistics. Many of these are large block trades between institutions.
The only thing I’ll add on time, besides what I already mentioned in the context of intermediate-term cycles, is that you can layer time projections onto a 3-wave or 5-wave price structure to identify dates where a wave is more likely to end. Like Fibonacci price projections, it’s a deep topic that I’ll save for future posts.
I added fundamental analysis as another crucial element to consider, although some traders will consider it anathema. Fundamental analysis is a broad term for factors that aren’t viewable on a price chart. It includes: political trends, demographics, social patterns, economic indices, and other factors that may impact a single sector or all financial markets over a very long period of time.
Many technical analysts believe all information is reflected in the market price, so they’d say it’s merely a distraction to look at fundamentals. I agree that much of the “market news and commentary” out there is distraction, but I strongly disagree that you can ignore the fundamentals. They can provide clues that you’ll never find on a price chart.
I believe fundamental analysis is most valuable for spotting major extremes in price or bubbles. In 2000, 19 online startups bought very expensive Super Bowl advertisements. Certainly, this would add some evidence in favor of the dot-com bubble being about to burst, which it did shortly after.
Considering all the above info, I’ve chosen an intermediate-term trend of “Up (?)”, signifying that the uptrend is still in place but has weakened and is almost done.
Translate the Trend Into A Fearless Forecast
Like we did with the long-term trend, next we take the actual and forecasted long-term trends (“Up” and “Up (?)”, respectively) and translate them into a single percentage between -100% and +100%. I have a grid I use to define the percentage for every possible combination of actual and forecasted trend (5 x 5 = 25 combinations in total). Here, the percentage is +20%. This indicates that although we still want to hold tech stocks (the percentage is positive), we are only slightly positive on the sector and we certainly don’t want to buy more of it.
When the XLK gets above $52, or breaks down from its current levels, I’ll mark the forecasted trend “Down”, the percentage will go negative to -50%, and I’ll be going short the tech sector.
We want to go long the sectors, markets, and asset classes that have the highest percentages (closest to 100%) and go short the ones that have the lowest percentages (closest to -100%).
If you have a baseline asset allocation, you’ll want to adjust it up or down based on the percentages in each asset class, but staying within a minimum and maximum for each asset class. This is what I do to ensure that while I act on opportunities, I don’t get too carried away with a single asset class. More on this in future posts.
Also, look forward to many more examples of this framework in action throughout future live streams, videos, and of course our free trading podcast.
I know this approach takes more time than the easy single-indicator “strategies” (advertised as such, but really a total fraud!). But it works. This is a winning approach for managing a large portfolio of assets or for swing trading a smaller account. It is a comprehensive, time-tested, and successful strategic framework when it is used in conjunction with good risk management and a sound mental approach to trading.