Pullback Reversal
Given the abundance of seesaw motions in any chart, it is no wonder that pullback reversal strategies rank high on the list of popular trading techniques. However interesting the general premise involved, aspiring traders would do well not to think of these tactics too lightly. Always a reason for caution is the contrarian element within. For even when the reversal itself is traded in line with the dominant pressure, there is still the pressure of the pullback to reverse.
Pullbacks come in many shapes and sizes, but all can be classified into one oftwo categories: they are either a correction in price or in time. In the time variety, the puliback often resembles a horizontal or slightly tilted flag. In many such cases we can trade the “reversal” as if dealing with a regular pattern break setup (or a variant of it).
The correction in price is usually more diagonally shaped. To qualify as a reversal candidate, this type of pullback should correct about 50 to 60 percent of the dominant swing. (In a very strong trend, 40 percent might do, provided the turn sets up well.)
But rather than measuring the amount of retracement in absolute terms, perhaps an even better marker is to wait for the correction to reach the 25ema first, preferably even perforate it a bit, and then look for a turn. In fact, adhering to this one filter alone may already prevent the majority of pullback reversal traps.
It gets even better if, prior to turning around, the correction manages to put in a test with an element of support or resistance, like a pattern line extension, a ceiling test, a previous high or low in the dominant swing, or even a round number level.
Our standard entry on the pullback reversal is taken on the break of a signal bar, too. This bar may represent the high or low of the pullback itself (one-bar turnaround), but probably more common is for the anticipated turn to show at least a few bars of pushing and pulling before the pullback appears to be ready to “roll over”.
It should be noted that the specifications listed above are best seen as guidelines with which to gauge the ripeness of the correction and the readiness of the turn. Reversal tactics are always a function of the bigger picture and best applied accordingly.
Figure 5.16 In earlier discussions we touched briefly upon the virtues of proportion in the trend/ pullback relation. We can refer to it as the harmony principle. Although somewhat prone to personal interpretation, within it there are quite a few elements present that can be put to good use without much room for debate. A first sign of harmony, and rather easy to detect, is a favorable degree of retracement. Another thing to monitor is the way the bars in the pullback get printed. They do not have to be all white or al black, but their average span should not exceed that of the bars in the dominant swing. On balance, the best reversal candidate is a pullback that appears to travel almost “reluctantly” against the dominant pressure.
To estimate the percentage of reversal, it is best to start out with the point from where the latest dominant swing took off (1). In this manner, pullback 3-6 had retraced about 50 percent of 1-3; and so had pullback 9-10 in relation to the 7-9 swing. This picture demonstrates also why the 25ema is such a great average to plot in the 5-minute chart; in a relatively calm trend, it tends to “lag behind” just enough for a pullback to collide with it prior to reversing.
Take note of the bear trap below bar 5. At least four reasons come to mind why this mini break did not deserve entry signal status: (a) the break was directed against the pressure of a little bull block to the left (4); (b) it went up against the resistance of a pullback line; (c) the 25ema had not been properly tested (no buildup in the turn); and (d) no technical tests were hit upon in the high of the pullback yet.
None of this may have called for aggression on the buy side, but these were valid reasons to lay low on the sell side.
Equally poisonous was the perforation of the pullback line a few pip further south. When aiming for a reversal, it is crucial to realize that the pullback will possibly do everything in its power to postpone, if not fully fend off, the moment of the turn. This is why we see so many false turnarounds on first try; rather than stepping out of the way on the initial break, contrarians tend to come in again to take advantage of the pullback’s temporary weakness. If only for this reason, a broken pullback line, by itself, does not make for the most reliable of entry signals, and even less so when the correction has yet to run into an element of support or resistance.
As to the latter, the more conspicuous this obstructive element the better, but even the high or low of a seemingly insignificant bar could do. For example, in bear swing 1-3, bar 2 represented a tiny one-bar “hiccup” in the downtrend, essentially a bullish reversal attempt that had failed. Insignificant as this may have appeared at the time of occurrence, ultimately it was this bar’s low to which the high of the pullback was drawn (high of 6 tests low of 2). As we know, when a 50/60 percent correction meets up with a trending 25ema and a technical test (triple), things always get interesting. Any moment now, plenty of reversal strategies could spring to life , and as a consequence, contrarian parties may be forced to bail out (double pressure potential) .
It took three bars in the high of the pullback to set up the turn in tradable fashion (6-7) . Technically seen, this was also a pattern break combi setup. Even though both the powerbar and inside bar of combi 7 were equally nondescript, their timing and placing were superb; and the equal lows allowed us to short from a double-bar break. Note also the favorable 50-level magnet about 25 pip below. This was a high-odds reversal setup in all respects (could even qualify as a pattern break pullback setup).
Another element that contributed favorably to the prospects of the reversal was the fact that this was the first pullback to defy the 1-3 breakout. Subsequent corrections can still set up a tradable turn (as shown in this chart also), but the first pullback to hit upon the 25ema in a newborn dominant swing is generally the one most eagerly countered.
About an hour later, the second pullback showed strong similarities with the first. Once again the correction had retraced about 50 percent of the dominant swing, and in the top there appeared another technical test with a former “hiccup” on the left (high of bar 10 hit upon bar 8). Perhaps a ceiling test with the lows of combi 7 may have represented a more notable magnet, but that would have required a serious breach of the 25ema and a near full retracement of the 7-9 swing. This pullback was fine as it was, and thus a reversal candidate.
Bar 1 1 was the first bearish bar to show up in the highs of the correction. Was it equally poisonous, as a signal bar, as bar 5 had been earlier on?
Let’s examine the pros first: (a) the pullback had corrected about 50 percent of the 7-9 swing; (b) prices had pierced the 25ema a bit and then were stalling in it, building up pressure while at it; (c) the pullback line was already broken and could not block the path; (d) bar 1 1 was a bearish turnaround bar and set up favorably close to the average (no adverse magnet) ; (e) ; the high of the correction had put in a technical test with bar 8; (f) the market was still in bearish mode.
On the downside, bar 1 1 was a very modest turnaround bar in relation to the relative strength of the bullish pullback, and also the first bearish bar to show up in the turn (not necessarily a deal-breaker, though) . And the pullback itself was the second to appear against the bearish pressure. To a very conservative bear these elements may have raised a little red flag still.
At the risk of missing the turn completely, always an option is to wait for one or more bars to bring more clarity about. Should bar 11 have been discarded as a signal bar, bar 12 would have made for a good replacement. Following the earlier break of bar 11 , this was the second break in the turn to favor the bears. In countless turnaround situations, it is the second break that triggers the desired double-pressure response.
Note: To understand the advantage of a second break over a first, we could look upon the turn in a bullish pullback as an M-pattern variant. (And upon the turn of a bearish pullback as a W-pattern variant.) To visualize this, simply track the price action from the start of the pullback to the start of the second turn. For example, in a bullish pullback, the action first goes up to reach the 25ema, then goes down on a first break, then goes up again in defiance of this break, and then goes down again (second break) . That is an M-pattern turn.
Figure 5.17 Next to the dominant pressure, another thing to assess in pullback reversal tactics, and no less significant, is whether the market is currently trending or ranging. A pullback reversal that needs to surpass the barrier of a range to get to target is likely to meet more opposition. The session above shows reversals of either kind.
The trend variant came about as a consequence of the 3-4 correction. This pullback was diagonal and orderly (black bars only) , had retraced about 50 percent of the 1 -3 swing and was the first to reach the 25ema since the break of the box in the left hand corner. Since the foregoing turn at 2 had been rather swift on our 5-minute, the chart failed to provide a textbook ceiling test (no floor to bounce off) , but the 3-4 pullback had found support nonetheless, which formed a higher bottom. The indisputable pullback line was a welcome bonus too, and so was the lovely combi setup squeezed tight beneath it (5).
A nonfavorable element may have been the adverse magnet of the OO-level; although the lows at 2 had come real close, the actual test was still left open. However, with both the environment and the setup speaking in favor of continuation, this was of minor concern only and certainly not disruptive enough to skip the trade (enter long on the break of combi 5) .
How about the 7-9 retracement. This was another pullback to the trending 25ema, but we can immediately see that the conditions in which it showed up were not nearly as favorable. First of all, the progression was not a neat, one-sided diagonal retracement to the average (the bullish upstroke at 8 had turned the pullback into a two-legged swing) , but more worrisome was the fact that the correction was now part of a head-and-shoulders topping formation (6-7-8). As a matter of fact, the neckline of this progression (low of the box) was broken by the very same bar that treacherously posed itself as a signal bar for a pullback reversal long (9).
When we have both bearish and bullish elements in place (reversal element in a trending market), the general consensus on the situation is likely to be divided. A problem this is not, because we do not have to trade. Nine times out of ten, the best side is neither short nor long, but outside of the market (skip bull entry above bar 9).
But how about pullback 1 1 -12. This retracement wasn’t as gentle as the 3-4 correction earlier on, but when compared to the powerful bear rally before it (10-1 1), the countercharge wasn’t overly aggressive and as such may have qualified as a reversal candidate. However, there was a good reason also to decline the offer below bar 12: the surrounding conditions were unsupportive.
If we solely concentrate on the bear rally preceding the pullback, the trend was unmistakably down and maybe there was more of this to come. Yet if we zoom out to address the bigger picture, we can see that prices were actually residing in the lower part of a much wider range. In fact, the low of bar 1 1 , a spot-on round number test, represented a double bottom with the low of the box on the left. In other words, even though the collapse of the earlier bull trend was beyond dispute, with prices now in the lower region of a range this was not the best environment to look for bearish continuation. To better understand the dangers involved, let us briefly recall the typical forces in play in the lower region of a range: bears take profits, sideline bears stay away and contrarian bulls might come in any moment.
This is not to suggest that we cannot trade a pullback reversal within a range, but we do have to be more picky about it. particularly when prices plummet straight down from the top of the range to the bottom (7-11 , or 10-11, if you wish) , we definitely shouldn’t be too eager to play for more continuation near the lows.
While nimble scalpers may have been able to extract some quick profits in the turn below bar 12, or thereabouts, powerbar 13 depicts a perfect example of the dangers involved. Granted, this trap scenario is not unthinkable even in the best of reversal breakouts, but it is all the more common in a nonfavorable environment, and with the pullback already not the prettiest to begin with.
For future purposes, it is recommended to examine for a moment how this trap came about. Note that contrarian parties didn’t immediately come out to counter the bearish intentions below bar 12. Instead, bears were “allowed” to enter the ring and even enjoy some follow through on the way to their 50-level magnet: so far, all quiet on the southern front. Until suddenly, somewhere around the halfway mark of the 1 1 -12 swing (very typical) , contrarians came in full throttle with a merciless powerbar that singlehandedly set the stage for the upcoming bull-side reversal (1 3). Do realize, though, that such a powerbar is never solely a marker of contrarian prowess; bears, too, will have contributed to the nature of it, as they toppled over one another in their hurried flight for safety (double pressure). Particularly in the lows or high of a range, this flight potential of fellow participants is always a serious concern and thus a good reason to practice caution rather than bravery.
If all of the above still raises some questions as to the validity of the skip below bar 12, in Chapter 7 we will dig into the matter of unfavorable conditions more deeply; and there will be countless more examples of pullback situations, both tradable and non-tradable, in the chapters ahead. For now, as a rule of thumb, if the pullback itself is already not of the most exemplary kind, and the conditions for follow-through at least questionable to some extent, by al means just skip the trade. For all your time and effort in the markets, reward yourself with high-odds wagers only, rather than throwing yourself too eagerly at the mercy of the market’s fickle whims.
An event that shouldn’t have raised any questions at al as to the proper cause of action is the pullback reversal offer above bar 16. You don’t even have to assess whether the market is trending or ranging to decline a trade like this. Whenever these extraordinary tall bars show up (13- 14 and 14-1 5), things are indeed out of the ordinary and for this reason alone, it is no longer safe to venture out with a tiny stop. Don’t even consider trading this. Just skip.
Note: Ifyou regularly find yourself caught in ventures that look okay at the onset yet rather questionable in hindsight, chances are your perception of the odds is affected by your desire to trade. It is therefore crucial to analyze all your past trades, both winners and losers, to find out how much the element of overtrading plays a part in your game. If it does so to an uncomfortable degree, and this turns out to be a habit hard to kick, a possible solution could be to start trading multiple markets simultaneously. Thus, rather than trying to squeeze the life out of one favorite instrument, set up a few more markets on your screens and trade their nondebatable offers only.
In any case, always keep in mind that a savvy trader has no need for a trade, which is why you will not easily catch him searching for one either. Although we may freely use the verb as a figure of speech, the best opportunities in the market are not found, they are spotted. The sooner a student trader understands this subtle distinction, the better his odds to emerge from the learning stages with funds left to trade and morale still intact.
Let’s check out a couple more pullback reversal examples.
Figure 5.18 In almost all trading literature, the trend is passionately glorified as if it were a treasure trove full of high-odds opportunities. Whether or not this reputation is rightfully earned, it is a fact of trading life that on our 5-minute frame the beloved trend is more a rarity than a common sight; and even if it shows up in full-fledged fashion, it may offer very few pullbacks that retrace deep or lengthy enough to play a reversal with acceptable risk. One benefit of the trend is beyond question, though, and that is that it clearly tells us which side of the market to shun.
Although printed as a continuous curve, the 25ema has a closing price just like any regular bar. Since its slope is a reflection of the average closing price of the last 25 bars, it cannot help but lag behind if suddenly a frantic wave of buying or selling takes the current price bars away from the mean. In order for the average to catch up, so to speak, prices either need to pull back towards it or they need to go sideways for more than a bit and put in the so-called correction in time. A good example of such a sideways correction is illustrated by the bull-flag formation 1-4, hanging from the high of the UK morning bull rally.
In a more rangebound market, it usually doesn’t take very long for a decent pullback to hit upon the 25ema. In a strong trend this may not be the case for many hours on end. Although our core strategy is designed to trade off the 25ema “base”, there is no rule that says we cannot deviate from standard procedure. But for this to be granted, we best make sure to have the chart conditions strongly on our side.
I believe the entry above bar 4 provides a good example of a valid exception. By accepting this offer, we basically imply that it is not our entry that is premature, it is the average that is too late. Of course, trading is never a game of semantics. The only reason to accept a trade at all is simply because the odds favor a positive outcome.
As we know from earlier discussions, to qualify for a breakout it is essential for the flag to express a certain ripeness in harmony with the pole from which it hangs. We also took up that it’s hard to determine the ideal presence of this concept. But maybe we can obtain a better understanding by imagining its absence first. For example, should we picture this flag to already have been broken above, say, bar 2, this would have made for an uncomfortably small flag in relation to the might of the bull rally. (Apart from the fact that such an entry would have been way too far removed from the average to accept.)
It is important to understand, though, that the idea of harmony is never a matter of bars per se, it is a function of the number of sideways bars relative to the strength of the flagpole. Flag progression 5-6, for example, was a nice flag in relation to its pole 4-5, even though it showed only a three-bar body before breaking out.
As already mentioned, popular perception has it that when a flag gets broken, this could induce a new swing more or less equal to the length of the flagpole. There may be some debate on the exact points from which to calculate this, but this doesn’t take away from the poleflag-swing potential itself. Therefore, whenever we see a flag take on “sufficient” body, this could portend the coming of a very interesting breakout.
To determine whether progression 1-4 could boast of such pleasurable characteristics, let’s examine the available hints and clues up to that point. First off, the overall pressure was indisputably up. Second, the flag was the first decent pullback since the start of the bull rally. Third, a premature break at T had been fended off by the bears, but the subsequent correction had not surpassed the earlier low of 3 (the higher low after the tease break indicated bullish resilience) . Fourth, progression 3-4 showed the blocky features of buildup. Fifth, bar 4 had put in a false low with its neighbor before spinning around and setting up as a bullish signal bar.
An element of concern may have been the 50-level adverse magnet which had been broken on the way up but not yet tested back. And then there was a slight issue with the flag line which may not have been broken yet on the break of bar 4 itself (depending on how sharply that line was drawn by our fellow participants). Nevertheless, considering al the elements that favored continuation, and with the flag itself more and more a token of 50-level defiance, I believe we should just bite the bullet here and accept.
If not in position on this flag break trade, or possibly already on target, could we deviate again from standard procedure so as to pick up the flag break entry above bar 6? And what about an entry above bar 8? Conservatively regarded, both are easy skips. In a strong trend it is certainly not unthinkable to see some follow-through on these smaller flag breakouts-and if already in position, it is always a pleasure to see them break-but by themselves both entries would have demanded a serious violation of standard operating tactics. Just look at the distance away from the 25ema base. On the break above bar 4, we deliberately strayed from the regular path on account of the excellent prospects (big flag, first pullback); to do the same on subsequent wagers, of lesser quality also, could be asking for trouble.
In the 9- 10 correction, the intermediate lows of 6 and 8 were taken out, but in the light of the bullish dominance these were minor breaks that weren’t likely to harm the prevalent pressure. More interesting, therefore, was the fact the bar 10 had put in a technical test with the 3-4 cluster, thereby taking care of the only adverse magnet of significance in the area. Once this test was put in, it took only a couple of bars for the market to set up another pullback reversal (enter long above combi 1 1 ).
Prior to taking any position there are always two type of dangers to assess: adverse magnets on the way to the stop and obstructive elements on the way to the target. While adverse magnets deserve caution by default, if only as a consequence of a tight stop strategy, obstructive elements on the way to target are best judged in relation to the market’s current strength. For example, in a more rangebound market, the 7-9 double top in the round number area may have called for more caution on the bull side. Yet it is the very nature of a trend to sooner take out such resistance than be obstructed by it.
Note: Since our entry bar (above the arrow) did not shoot off right away but instead closed bullishly a few pip above the average, this basically set up another pullback reversal (for those not yet in position). But here we touch upon a provocative issue: if the second break with the higher entry was also playable for 20 pip, should we then not adjust the target level of the first trade, if in position on it, to meet the higher objective of the second? That’s a valid question for sure, but I leave it up to the reader to answer it. Personally, I am not a fan of meddling with an open trade in an attempt to gun for more profits than originally intended. But it’s hard to argue with anyone taking a different view on this matter. By the same token, if position was taken on the first entry, the stop could then be lifted to the level of the stop on the higher entry. Yet all this is very much dependent on personal style and management technique. (Do recall that our 20/ 10 bracket, though highly effective, is merely a suggestion, too.)
Amazingly, just a few bars later, this bull trend came up with yet another flag pattern (12-1 3low) hanging sideways from the pole 1 1 -12 . Although the break of this flag did not set up in tradable fashion, the powerful pop in bar 13 shows us once again that when the market is in trending mode, contrarians have little business defying the obvious. More than learning about specific entry techniques, probably the biggest lesson to take away from this chart is to always acknowledge the presence of a trend. When it comes to playing it for continuation, it may very well be that no valid entries are provided or that they feel too uncomfortable or too borderline to accept. If that is the case, just let these sessions be. It is always better to remain empty-handed on the sidelines than to chase the market up or down out of fear of missing out.
Figure 5.19 While it is impossible to predict the exact level for a trend to shift into a ranging phase, the shift from a sideways market into a trending phase can often be anticipated with the utmost precision. If only for this reason, every trader owes it to himself to get up to speed with range break tactics. But the barrier fights are not the sole determinants of participation. When showing sizable width, it is not unthinkable to trade within the range itself. In many an instance, the outer barriers may even work as magnets, pleasantly sucking an inside-range trade in the plus, or even to target.
A good indication of a ranging market is a rather flat 25ema with prices meandering above and below it without making much headway on either side. Always handy is to wrap a box around such action and extend it to the right for future purposes. The moment prices edged down from the high of bar 5, the box could already have been deployed as depicted. When it comes to the degree of its extension, just plot whatever looks harmonious in terms of width and length; a ratio of about 1 :3 wil usually do fine. You can always extend the box some more if necessary.
This range eventually got broken not from a buildup situation directly underneath the top barrier, but from a level well below it. This implies that we couldn’t have traded the breakout in regular pattern break fashion. It is interesting to see, though, that the range barrier still played a role in both trades of the session: first it showed its magnetic powers on the break above combi 11 , and later on it provided the base for a bull-flag reversal (13-1 5). Before we take up the details of these ventures, let us quickly run through some of the earlier action first.
On the far left there appeared a bearish combi at the end of a diagonal pullback to the 25ema (2), but from what is shown we cannot tell whether this set up a reversal that called for action.
A pullback reversal to have skipped for sure was the short below bar Not only was there zero buildup in this turn, there was also a strong risk of the 50-level adverse magnet.
Bulls bravely marched on beyond the round number and soon hit upon the high of bar 1 at 5. As is quite common, the former high offered resistance and although it took some pulling and pushing in the area, ultimately prices were forced to retreat, leaving a triple top in their wake (1 -5-6). Not long after, the 6-7 correction hit upon a supportive triple of the 25ema, the round number and what looked to be a 40 percent retracement of the 3-5 bull swing. As bears laid low in this tricky area, bulls were given ample chance to regroup and plan another charge on the top barrier defense.
The break of bar 7, however, was not a good place from where to launch this new attack. Prices may have pulled back to the average, and bar 7 may have been a turnaround bar in it, and even a false low with its direct neighbor, there was neither trend nor buildup to back up this charge. Instead, there was the 5-6 bear block directly overhead (M-pattern middle-part) .
About an hour later, the 50-level defense was much more established. Despite the overall thinness of the “correction in time”, both bar 9 and combi 10 had managed to put in a higher low, and bar 8 a false low with 7. Furthermore, if we zoom out a little, we can see that the 7-11 action was actually part of a much bigger pattern, the bull-flag 5- 1 1 hanging from the pole 3-5.
Although the current development favored a bull break more than a bear break, the offer above combi 10 can be qualified as premature still. Considering the subdued action of the moment, and with the bigger flag line yet to be broken (dotted line), it just wasn’t very likely that prices would shoot off like a rocket here. So why not grant the market a little extra time to set up the break in a stronger fashion.
In all fairness, the span of combi 1 1 , too, was terribly nondescript, but with the dotted line now cleared, at least the. issue of immediate resistance was solved. Looking closely, this pullback reversal harbored all the makings of a pattern break combi setup as well.
If we compare the break of combi 1 1 with the break of bar 7, both at the same price level, this nicely demonstrates that skipping or accepting an offer never comes down to price, it is always a function of pressure, buildup and conditions. And another thing: whereas the top barrier of the range was likely to act as resistance to a break above bar 7, on the break above combi 1 1 it stood to work as a favorable magnet. Simply as a consequence of the better buildup.
This example suggests also that, given certain conditions, a barrier breakout can indeed be anticipated from a level well within the range. Even though prices may still have to surpass the outer barrier to get to target, at least the position will be pleasantly in the plus on impact. Furthermore, there is usually less need also for a substantial break to reach destination. And not uninteresting either, should the barrier level prove too hard to crack, the current profits on the trade provide reasonable cushioning for the position to be scratched with minimal damage, or even for a profit still. (Intervention tactics will be taken up in Chapter 6.) In earlier discussions we came to regard the pullback to a broken barrier as a continuation setup (13-1 4). But we also addressed the fact that, particularly in a tease break scenario, the pullback is not necessarily halted by the barrier level itself; prices are often seen to break back into the pattern to put in a ceiling test of sorts.
Regardless of whether a tease break correction remains outside of a broken pattern or travels back in, things usually do not differ much when it comes to assessing a pullback reversal or pattern break pullback setup. In all cases we have to monitor the behavior of the pullback, the amount of retracement, the harmony aspect and the potential level for a ceiling test bounce. And in all cases we have to locate a crucial bar from which break to play the reversal.
This leaves us to examine the options on the 13- 14 correction. How about bar 14 itself? Although this bar had found support in bar 12 on the left, and also touched upon the 50 percent retracement of the 1 1 – 13 breakout swing, it didn’t really set up as a very dependable signal bar. At that moment in time, the features of the pullback were still quite blocky and thus offered too much resistance for this breakout to qualify as a high-odds wager. Best to relax and see if the market once again would be kind enough to set up a better offer.
Depending on preference, there were several ways to have taken position on the bull-side breakout with reasonable odds attached. The first was to fire long in bar 15 the moment it took out its predecessor, which nicely coincided with the re-break of the range barrier. However, with the pullback still quite blocky overhead, and its angular barrier yet to be broken, this option, too , can be deemed quite aggressive. An immediate alternative here was to wait for bar 15 to take out the pullback line first, a few pip higher up, and then fire long. Evidently, this is a deviation from standard routine because the entry is not taken directly upon the break of a signal bar. Depending on the “ripeness” of the buildup, this can be a defensible tactic at times, but it should be taken into account that the bracket stop will be lifted along with a higher entry. Not necessarily a problem, but it may pay to check if this doesn’t offer a very awkward stop in the situation at hand.
Standard procedure, although at the inevitable risk of missing the break, is to simply wait for a signal bar to set up either against the barrier of interest or slightly through, and then enter upon its break. Bar 15 set up perfectly in this respect, with its high just peeking through the flag line. With both barriers now cleared, the only obstruction of technical relevance on the way to the round number magnet was the high of the flag itself (13) . Seldom a major concern.