A Single Well-Drawn Trend Line Can Be More Reliable Than Just About Any Indicator Reading

moneykeyboard-featured

Welcome to our feature-length blog article for the month of December, 2016, sponsored by the Trendline Mastery video foundation course (www.forexmentor.com/trendline/), and the related Trendline Mastery Trader’s Club membership service (www.forexmentor.com/trendline/club/) available exclusively from Forexmentor. Either of these learning resources alone, and especially both used in combination, can greatly assist you in learning how to successfully apply the many trading tools and techniques that can be derived from multi-timeframe trend line analysis.  If you should have any questions on these learning products, or any query on the topic of trend lines in general, feel free to drop me, the author (Frank Paul), a line anytime via email, at frank@forexmentor.com.

On to the topic at hand: In this latest monthly installment in our blog series, we’re going to consider the relative performance attributes of trend line analysis versus indicator readings, for the purpose of detecting or confirming a viable trend trading bias. As you can probably guess – if by nothing else than the admittedly ‘biased’ wording of this article’s title – our contention here is that a single well-drawn trend line on any chart can usually do a more effective and more timely job for you in identifying a trend bias, than any number of concurrent indicator readings on that same timeframe.

Now, before we dig in to some analysis to show you how and why it is we think that’s the case, and before I inadvertently end up inviting a flurry of emails loudly objecting to the assumed implication – that “indicators stink and should be avoided at all costs” – let me acknowledge the contrary, which is that indicators can and do have an important potential role to play in Technical Analysis.

So, if you interpret what we have to say in the following paragraphs as a broadside attack on a chart tool that you have a long-standing personal preferential familiarity with, let me be clear by stating that is absolutely not our intention.

That said, all indicators do have the potential to generate unreliable readings at any given time, and if you’re not at least warily on guard against that vulnerability, then you can easily be led astray by them, causing otherwise avoidable trading losses. At the very least, a prudent approach to technical trading entails a prohibition against overly mechanical, rules-based interpretations of what indicators seem to be saying, e.g. “If the MACD line crosses down and through its Signal line, then a bearish reversal can be assumed”.  Sometimes that’s true, but often it absolutely is not. And it is only by assessing concurrent price behaviour in relation to the diagonal form of Support/Resistance that is conveyed by trend lines, that you can validate (or correctly refute) any given indicator reading.

Illustration #1: MACD ‘Gap & Angle’ vs. Trendline Analysis

MACD is one of the first so-called ‘momentum’ indicators that most student traders come in contact with, after Moving Averages. (I use the phrase ‘so-called’ here because, given the fact that MACD is constructed from the difference of two underlying Moving Averages, which are lagging indicators themselves, it is arguable that MACD is more like a trend following indicator than a momentum indicator).

The simplest and easiest application of MACD is when a strong trending move aligns with the MACD line itself pointing higher or lower, on a bar-by-bar basis, and specifically from above or below its Signal line accordingly, resulting in a so-called ‘Gap & Angle’ reading.  The idea here is that if you have such a reading in force, then any movement by price in the opposite direction should be read as counter-trend. That information, in turn, can help you (at least in theory) identify a potential opportunity to sell a corrective rally in a downtrend, or buy a corrective dip in an uptrend.

While a crystal-clear MACD directional bias reading can provide a useful visual confirmation of a strong underlying move underway on the price panel, its validity is only good if you can assume that the indicator is not, either in the live edge of price action or in the very near future, vulnerable to lagging a trend reversal of some kind. Unfortunately, the reality is that MACD very often can and does lag at important turns. Because it is constructed from underlying Moving Averages, there will always be a certain duration and magnitude of price action in the opposite direction that is typically required to change that directional bias on MACD. In such instances, the turn will happen on price before it is confirmed on MACD, and by then, it could easily be too late.

Consider the far right-hand edge of our first chart example below:

Chart Example 1A: MACD Gap & Angle Is Bullish

In the above screen capture, MACD is providing a clear and unequivocal bullish ‘Gap & Angle’ reading: the dark blue MACD line is trending higher on a bar-by-bar basis to the right side of the chart, is doing so specifically from above its lagging Signal line reading, the latter in turn also waving higher on a bar-by-bar basis.  This bullish reading seems to be further underscored by the MACD Histogram, which seems to be likewise trending higher to the far right-hand side of the chart. So, a simple, mechanical, rules-based interpretation focused on the MACD indicator in this example is that price is in an uptrend, and therefore, a bullish bias should be pursued.  Right?

Well, not exactly. The first clue that the above MACD reading is possibly not actionable is conveyed by something on the price panel that actually doesn’t have anything to do with trend lines, and that is a series of four consecutive rejection wick highs all lining up at about the same level, and overlapping very closely a Swing Point High pattern in the background.  In chart example #1B below, we draw a horizontal line segment rightwards from the old high (the first red arrow to the left) to the more recent highs on the right-hand side of the chart.

This is very clearly a resistance barrier, and the prudent trader would tend to avoid a taking a fresh long position from close by, due to the possibility of anything from sideways consolidation to a minor retracement to a major retracement to an outright trend reversal unfolding thereafter.  Without being able to predict with certainty which (if any) of those possible scenarios will ensue, the prudent trader will wait for a candle close above that resistance barrier before pursuing the long bias further. And yet no such cautionary warning can be gleaned from MACD, which remains clearly bullish as price itself hazards a warning of a nearby danger.

Chart Example 1B: Despite a Bullish MACD, Resistance Lies Close At Hand

As we follow this chart example several bars into the future, we can see that, after a few more hits of resistance at the horizontal resistance line identified, price breaks the most recently in-force Demand trend line, with the small (yet very real) down close on the bar second to the left from the dashed vertical line in chart example 1C below.  Though some MACD aficionados might object that the MACD line, at this point, is starting to wave lower (as is the Histogram), it is nevertheless generally assumed that only a crossover of MACD through its Signal line from above on a candle close, constitutes a sufficient confirmation of a trend reversal, from up to down. However, in the following chart example, that does not actually occur until after the very sharp down close two bars later. In the very best case scenario, a late entry is achieved in a still viable short move, resulting in significant opportunity loss (at least as compared to the reversal signal generated by the TL break referenced in Chart Example 1B). In the worst case scenario, taking action on the MACD crossover is simply too late, as the spike bar in question is enough to trigger a retracement against it that is more than sufficient to trigger any but a possibly very wide Initial Protective Stop.

Chart Example 1C: MACD Crossover Lags At A Sharp Reversal

Though the above chart example was obviously selected to make a certain point, and it is not specifically the case that any strong MACD ‘Gap & Angle’ reading will always likewise fail so dramatically, it does nevertheless illustrate the vulnerability that such readings can convey.  In the above example, the only way to anticipate the sharp reversal ahead of time – not too later after the fact – was to note the bearish trend line break occurring two bars prior.  As a short trade entry signal, its timing was vastly superior, in terms of allowing the trader to get in closer to the actual top, with less drawdown potential, and much greater profit potential, in relation to the ensuing follow-through.  (FYI: Trend line break entry points are covered extensively in both the Trendline Mastery foundation course, and in the archival contents posted to the Trendline Mastery Trader’s Club).

Illustration #2: MACD Divergence Readings vs. Trendline Analysis

Another application of MACD (as well as other ‘momentum’ indicators, such as Stochastics, Williams %R, CCI and RSI) is to generate ‘divergence’ readings, which tend to imply reversals against the preceding trend.  Divergence readings can either be ‘negative’  (indicating a potential downside reversal against a preceding uptrend), or ‘positive’ (indicating a potential upside reversal against a preceding downtrend).  Negative Divergence readings, in turn, can be sub-categorized as either Standard (where higher highs on price are contradicted by lower highs on MACD) or Hidden/Inverted (where higher highs on either price or MACD are contradicted by flat highs on the other panel). Similarly,  Positive Divergence readings can be sub-categorized as either Standard (where lower lows on price are contradicted by higher lows on MACD) or Hidden/Inverted (where lower lows on either price or MACD are contradicted by flat lows on the other panel).

While these divergence events can be helpful in identifying tradable turning points in the market, particularly when aligned with other elements of Technical Analysis in a multiple timeframe ‘Confluence of Events’ scenario, the shortcoming of this approach is that you cannot assume that any and every divergence is going to result in a significant turn that allows price to travel far enough to generate sufficient profit potential and Reward/Risk potential.

Sometimes a divergence can occur at a profound turning point, after which price travels a great distance in the opposite direction. Other times – very often, in fact – a divergence precedes nothing more than a low-level retracement or consolidation which not long thereafter resolves sharply in the direction of the trend that preceded the divergence. In these cases, all you’re doing by entering into a divergence trade is unwittingly jumping in to a counter-trend trade. If you’re not specifically aware of that fact ahead of time, it might be almost impossible to correctly nail an appropriate scalping take-profit target before price does a sharp U-turn against you. And then you’re left on the sidelines wondering why it is that this supposedly wonderful entry signal failed you so badly.

Let’s look at a real-world illustration of the above scenario. In chart example 2 below, we see a MACD Positive Divergence confirmed by no later than the open of the candle through which the dashed vertical line is drawn.  At this point, we can see MACD crossing up and through its Signal line from below, which in turn signifies a quasi Swing Point Low (second green arrow) that is specifically higher than the one preceding it to the left (first green arrow).  This second higher low on MACD aligns with a progression of lower lows on the price panel, in relation to where price was at the point of the first MACD low to the left.

Chart Example 2: MACD Divergence That Doesn’t Amount To Much

In relation to price action that unfolded subsequently, it’s certainly safe to say that any attempt to enter a long trade on this MACD Positive Divergence was doomed to failure literally from the moment the buy button was clicked – price simply headed down straight away.  Fortunately, a MACD Divergence wasn’t the only thing to be paying attention to at this juncture. The price panel, by contrast, allowed for a trend line analysis that was specifically bearish, not only up to the point of, but specifically some time beyond when the Divergence reading materialized. To put it simply, trendline analysis on the price panel trumped the MACD divergence.

As denoted by the diagonal Supply trend line drawn from the first red arrow towards the left of the price panel, down to the second red arrow and extended rightwards from there, the trend on price was still clearly down. If we want to be accurate trend traders executing on the correct side of a trending market, then at worst, one should have chosen to remain neutral at the point where the conflicting divergence reading materialized. Better yet, trading instead short on the appearance of an ensuing Fibonacci-based Downtrend Continuation pattern setup (not shown) would have enabled a much easier, not to mention actually profitable trade attempt.

What we can conclude from  a chart example like this, and literally thousands upon thousands of examples you can easily find that show much the same thing, is that a clear trend metric on the price panel – in this case, a valid trend line – should always be presumed to trump any conflicting event or pattern on an indicator, such as MACD divergence. The latter only becomes more bankable when it aligns with other elements of Technical Analysis, particularly in context of multiple timeframe analysis.

Illustration #3: Stochastics Overbought/Oversold vs. Trendline Analysis

Stochastics is another popular indicator that’s been around for decades, and quoted widely in trading literature. The two main uses of this indicator are: 1) To help identify divergence readings (similar to the examples we considered above, in relation to MACD); and 2) To help identify tops and bottoms by way of crossover signals occurring from the so-called ‘overbought’ and ‘oversold’ ranges of the indicator panel. The former is usually defined as any instance where the Stochastics %K crosses down and through %D from above the 80 line (some traders tweak this value to 70). The latter is defined as any instance where %K crosses up and through %D from below the 20 line (likewise tweaked by some traders to a value of 30 instead). The idea here is that these crossover events from ‘peak’ readings in relation to the indicator panel tend to signal directional reversals.

In chart example 3A below, we can see how the Stochastics %K line (black solid) crossed down and through the %D line (red dashed) from above the 80 line, which constitutes a bearish crossover.  Consider, though, that the distance traveled from the open of the first bar (denoted by the first dashed vertical line from the left) occurring after the crossover could be confirmed, to the lowest price low thereafter (before price changed direction and headed higher to take out the preceding high) was only 9 pips.  We can conclude from this example – and many thousands upon thousands like it which would be easy to compile – that sometimes, the apparently visually ‘dramatic’ nature of such signals can be totally misleading. If you were to simply defer to a mechanical interpretation of such signals and execute on them accordingly, it can be assumed with complete justification that you would lose far more often than you’d need to.

Chart Example 3A: Stochastics Bearish Crossover That Doesn’t Amount To Much

Sticking with example 3A below for a moment, we can see that several bars later, at the juncture denoted by the second vertical line, something akin to a bearish ‘overbought’ crossover occurs; namely, a lower peak on Stochastics against a higher high on price, resulting in a Negative Divergence. When %K then crosses down through %D from above, we have the confirmed bearish reversal event.  So from here, we should presumably be able to expect price to finally start moving lower…right?

Well, not so fast!  As shown in example 3B below, we can see that price actually headed up – rather than down – almost immediately after the divergence signal could be confirmed, traveling a maximum extent of 60 pips before turning down again to any level close to breakeven, let alone any kind of significant profit on an intended short trade.  In fact, if you follow the readings through to the top occurring on the highest high on the right-hand side of the chart, you will notice more than a half-dozen bearish Stochastics signals, whether crossovers from the ‘overbought’ zone or Negative Divergence. Unfortunately, not a single one of them did a particularly useful or effective job in either confirming a trend reversal, or an actionable entry trigger. Any attempted sell execution occurring as a result of any of these signals would have resulted in a loss, unless using an ultra-wide stop.

Chart Example 3B: Price Continues Trending Higher Despite Consecutive Bearish Stochastics Signals

Here again is where Top-Down Trendline Analysis could definitely have played the hero role, in terms of negating a sell bias based on conflicting technicals.  In example 3C below, we can see that the 4hr chart sported a still intact Demand trend line occurring by the time we get to the dashed vertical line towards the right-hand side of the chart, which is the juncture on this timeframe where the supposedly bearish signals generated by Stochastics on the 15m chart materialized.  In fact, it doesn’t take any great deal of rocket science in this case, to conclude that instead of trying to sell bearish Stochastics crossover signals on the 15m chart, it would have actually made potentially a lot more sense to do the opposite – to buy bullish crossover signals – as long as those signals aligned with still intact bullish trendline based technicals apparent at the time on the 4hr chart (and similarly, the Daily chart, though not shown here).

Chart Example 3C: 4hr Chart Sports a Demand Trend Line, Which Negates The Bearish Stochastics Signals Seen on the 15m Chart

Illustration #4: Bollinger Band ‘Tag’ vs. Trendline Analysis

Entirely unrelated to the Moving Average-derived MACD trend following indicator and the purer ‘momentum’ gauge provided by Stochastics, is the volatility indicator known as Bollinger Bands. This form of analysis is derived from a sub-category of technical indicators called ‘envelopes’, whereby an upper and a lower band are each plotted directly on the price panel. As volatility picks up, or so the theory goes, the distance between the bands comprising the envelope increases; while, conversely, when volatility lessens, the distance between the bands likewise contracts.

The funny thing about B-Bands is that you can come across trading authorities espousing diametrically opposite uses for them. One school of thought suggests that when price makes several up closes to ‘tag’ the upper band, or several down closes to tag the lower band, you have a phenomenon called ‘walking the bands’, and as long as price remains above (if bullish) or below (if bearish) the 20-period Simple Moving Average around which the statistical norms comprising the outer bands are derived, then you stay with the implied trading bias.

The diametrically opposite strategy involves pairing the kind of ‘overbought’ or ‘oversold’ indicator readings on a momentum oscillator such as we discussed in the above section on Stochastics (although in this chart example we’re going to substitute the Commodity Channel Index in its place) with a B-band tag, for either a buy or sell signal. With this approach, we would pair a tag of the upper B-band with a CCI ‘overbought’ reading (turning down from above the +100 line) for a sell; and conversely, for a buy, we would pair a tag of the lower B-band with a CCI ‘oversold’ reading (turning up from below the -100 line).

There is at least one very major problem with overly mechanical strategies such as these (in addition to the obvious issue that the B-band signals mentioned above are completely contradictory to one another; surely they can’t both be right at any given time!), and that is, that they do not reference Top-Down Analytics or price panel-based trend readings. Because they do not, they are unable to help you specify whether a supposedly bearish or bullish signal is indicative of the onset of a retracement, or a trend-friendly continuation.  That distinction is critically important because scalping a counter-trend, corrective move is an entirely different ball of wax than letting profits run on a longer lasting move for higher Reward/Risk intervals.

Consider chart example # 4 below, in which we’ve plotted Bollinger Bands on the price panel (the solid formatted lines are the upper and lower bands, while the dashed line is the 20 SMA, from which the volatility-derived outer bands are derived), and the Commodity Channel Index (or CCI) as the indicator panel at bottom.

Chart Example 4: Bollinger Band Tags With or Without CCI Signals: Confusing At Best When Taken At Face Value

At the juncture denoted by the first dashed interval on the left, we see a tag of the lower B-band, which corresponds to an ‘oversold’ reading of CCI concurrently.  On the live edge of price action, this would have satisfied the definition of a buy signal using the B-band/CCI combination mentioned above. And, as it turned out, a decent buy opportunity did prove itself as price subsequently rallied by a measurable interval.  Conversely, the more basic ‘walking the bands’ interpretation would have suggested a short bias at this point, one which would have been stopped out for a loss in pretty short order.

But look what happens just five bars later, at the juncture denoted by the second dashed vertical line. At this point, we have a bullish ‘walking the bands’ reading whereby price tags the upper band from below, and stays above the B-band 20 SMA line for long enough to generate roughly 20 pips profit before price commenced the next pullback apparent on this timeframe. Going long with this simple signal would have worked out fine this time (last instance it would have resulted in a stop loss). Conversely, the B-band/CCI combo occurring at the same interval actually implied the exact opposite action: a sell as price tagged the upper band with CCI clearly now ‘overbought’.  Trying to go short on this combined signal would have resulted in a pretty quick trip to a stop loss, as price continued trending higher in direct contradiction to what this format of entry signal told us to do.

What the two highlighted junctures on chart example clearly prove, is that you could pretty well flip a coin as to whether a B-band tag is either a continuation signal or top/bottom reversal signal, since half the time it will tend to have significance one way, the other half, the other way.  But is flipping a coin as to whether an indicator signal means what it looks like it means really the best we can do to align the probabilities in our favour?  Hopefully you get the idea that this is a strictly rhetorical question, because in my mind, there is absolutely no doubt at all that the answer to that question is a resounding ‘No!’, and the way we resolve the impasse is with recourse to Top-Down Trendline Analysis.

To put it simply, the best and only way to use indicator-driven reversal signals for the purpose of entering a trade is to line them up on the right side of a clear, unambiguous, multi-timeframe trend reading.  If the trend is up (i.e. Demand Trend Lines evident), then you simply ignore sell signals on any one timeframe, whether MACD bearish Gap & Angle readings, MACD Negative Divergence readings, Stochastics bearish crossovers (whether simple or divergence-related), or Bollinger Band tags of the upper band with concurrent overbought CCI. Or whatever other forms of bearish entry signal you care to name.  If the trend is up, then to trade with it, the only types of indicator-driven trade entry signals you would focus your attention on are the bullish ones; and even then, subject to careful scrutiny, to further filter out potentially bad ones from good, on the basis of nearby Support/Resistance barriers, proximity to scheduled Market-Moving news events, and so on.

In the case of the final chart example above, which is actually for the exact same currency pair, date and chart timeframe as we used in examples 3A and 3B above (AUD/USD 15m chart for Wednesday, December 7th, 2016), we already know that Top-Down Trendline Analysis established for us that the bias on the Daily and 4hr charts at the time was up, not down.  Therefore, the only type of B-band signal that should have interested us was either a bullish ‘walking the band’ reading, or a tag of the lower band with a concurrent ‘oversold’ reading on CCI.  While many traders would voice a “well, duh…” objection in response to being reminded of such simple logic, you might be amazed to see how many chart examples profiling avoidable losses I’ve received over the years by traders who fail to implement such trade entry filters in the logic of their own trading decisions.

Simply put, you cannot ever hope to realize optimal long-term Net Profit results by using indicator readings or signals in the isolation of a single chart timeframe as the primary decision point for any trade execution.  The best and only way to sort the good from the bad, the wheat from the chaff, is to derive a trading bias from application of trendline analysis on the price panel, and then line up your indicator-based reversal signals in the appropriate direction: looking for bullish reversal signals in an uptrend, and bearish reversal signals in a downtrend.  If you can train yourself to do all of that with equally diligent focus on understanding and properly responding to the Support/Resistance environment relevant in the live edge of price action, you’re on your way to becoming a whiz-bang market technician.

And that’s pretty much it!

As this is the final monthly blog article for calendar 2016, I’d like to sign off by wishing you Happy Holidays (if you happen to observe them) and thanking you for following us over the past years.  Your interest is very much appreciated, whether you access only our free materials, or if you are a paying subscriber of our Forex learning products and services. I hope you’ve learned much from us that is useful, and here’s looking forward to working with you further in 2017.

Good Trading,

Frank Paul

Frank Paul at Forexmentor.com

Frank Paul at Forexmentor.com

I am well known for my ability to deliver clear and concise explanations of complex trading topics.   I am the force behind a series of comprehensive, yet practical, forex courses and training programs at Forexmentor.  (Forexmentor links to www.ForexMentor.com.)

“…purchased at least two of your courses and I must say that I greatly appreciate your methodology and willingness to share with others. One of the things I appreciate most about you is that you are not like some of those money grabbers out there who is so quick to sell you a system for huge bucks. I believe you are a very genuine person who would want to see me or anybody succeed in their Forex Trading career. Thanks again, Omar.”

Sharing This Post:

facebooktwittergoogle_plusredditpinterestlinkedinmail

After you share this post, please let me know what you think of this post in the comments section below.  Don’t know where to start?  Let me know what I can do for you.  How can I help you?  Thank you in advance! I’ll be around to give personal advice and guidance, and answer your questions.  So, please feel free to leave a comment with your feedback below right now.

Please keep in mind that all comments are moderated according to our privacy policy (Forexmentor’s policy on the site), and all links are nofollow. Please do NOT use keywords in the name field. Let’s keep our conversation meaningful and personal. I would appreciate it if you would provide the information marked *.
Your e-mail address will never be published. Thank you for sharing it. I will NEVER spam you – 100% privacy assured.